Abstracts from the 37th Annual Meeting of the Society of General Internal Medicine

Journal of General Internal Medicine, Apr 2014

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://link.springer.com/content/pdf/10.1007%2Fs11606-014-2834-9.pdf

Abstracts from the 37th Annual Meeting of the Society of General Internal Medicine

SCIENTIFIC ABSTRACTS 1 4 9 11 16 19 28 33 36 0 Montefiore Medical Center , Bronx, NY. (Tracking ID 1 A PLACE AT THE TABLE: EVALUATION OF COMMUNITY MEMBERS' EXPERIENCES AND EXPECTATIONS FOR ACADEMIC-COMMUNITY PARTNERSHIPS IN HIV/AIDS RESEARCH Stella Safo 2 Albert Einstein College of Medicine , Bronx, NY 3 Montefiore Medical Center , Bronx, NY 4 A ONE-STOP SHOP: PERCEIVED BENEFITS OF DIABETES GROUP VISITS IN THE SAFETY NET CLINIC SETTING Arshiya A. Baig 5 North Dakota State University , Fargo, ND. (Tracking ID 6 Riverside County Regional Medical Center , Moreno Valley, CA. (Tracking ID 7 Loma Linda University Medical Center , Loma Linda, CA 8 University of California , Riverside, Riverside, CA 9 CODE R: INTRODUCTION OF A HOSPITAL-WIDE PEER REVIEW PROCESS TO ASSESS RESIDENTS' POTENTIAL IMPACT ON PATIENT SAFETY AND QUALITY Daniel I. Kim 10 University of California, San Francisco , San Francisco, CA 11 HAVE YOU THOUGHT ABOUT SENDING THAT AS AN ECONSULT?: PRIMARY CARE PROVIDERS' EXPERIENCES WITH ELECTRONIC CONSULTATIONS AT AN ACADEMIC MEDICAL CENTER Sara Ackerman 12 University of California, San Francisco , San Francisco, CA. (Tracking ID 13 Perelman School of Medicine at the University of Pennsylvania , Philadelphia, PA. (Tracking ID 14 Uniformed Services University , Bethseda, MD 15 University of Chicago, Pritzker School of Medicine , Chicago, IL 16 WHAT DOES A GOOD LIFESTYLE MEAN TO YOU? A QUALITATIVE ANALYSIS OF FIRST YEAR MEDICAL STUDENTS' VIEWS Kimberly Clinite 17 New York University Medical Center , New York, NY 18 Hennepin County Medical Center , Minneapolis, MN 19 A CLUSTER RANDOMIZED CONTROLLED TRIAL OF INTERVENTIONS TO IMPROVE WORK CONDITIONS AND CLINICIAN STRESS: RESULTS FROM THE HEALTHY WORK PLACE (HWP) STUDY Mark Linzer 20 University of Missouri , Columbia, MO 21 University of Alabama , Tuscaloosa, AL 22 Loyola University Medical Center , Chicago, IL 23 Marshfield Clinic , Marshfield, WI 24 Forward Health Group, Inc, , Madison, WI. (Tracking ID 25 University of Wisconsin , Madison, WI 26 University of Miami Miller School of Medicine , Miami, FL 27 Sylvester Comprehensive Cancer Center , Miami, FL 28 ACCEPTABILITY AND FEASIBILITY OF HPV SELF-SAMPLING FOR CERVICAL CANCER SCREENING AMONG PATIENTS AND PROVIDERS IN TWO SAFETY-NET INSTITUTIONS IN MIAMI Kumar Ilangovan 29 University of Miami Miller School of Medicine , Miami, FL. (Tracking ID 30 Yale Univsersity , New Haven, CT. (Tracking ID 31 Yale School of Public Health , New Haven, CT 32 Harvard Pilgrim Health Care Institute/Department of Population Medicine at Harvard Medical School , Cambridge, MA 33 SIGUIENDO ADELANTE: A MEDICAL SPANISH CURRICULUM FOR RESIDENTS Avik Chatterjee 34 Creighton University School of Medicine , Phoenix, AZ. (Tracking ID 35 St Joseph's Hospital & Medical Center , Phoenix, AZ 36 STORY TELLING: MY MOST MEMORABLE PATIENT - LESSONS IN HUMANISM, REFLECTION, AND THE DEVELOPMENT OF EXPERTISE Priya Radhakrishnan S1 - BACKGROUND: Diabetes group visits, shared appointments in which patients with diabetes receive self-management education in a group setting and have a medical visit, are an innovative and promising way to deliver diabetes care. The group visit model may be especially promising in safety net clinics where many patients with diabetes in underserved settings receive their care. However, little is known about safety net providers perceptions of diabetes group visits in the community health center setting. METHODS: The research team conducted site visits at community health centers across the Midwest to assess their experiences in providing diabetes group visits. Site visits were conducted at centers that had implemented diabetes group visits in the past or currently had them. Two members of the research team interviewed health center personnel at each site regarding their diabetes group visit program. Health center personnel included chief executive officers, medical and nursing directors, care coordinators, physicians, nurses, physician assistants, diabetes educators, mental health professionals, dietitians, medical assistants, pharmacists, and other recommended staff members who were familiar with the diabetes group visits at the site. Interviewees were asked about the benefits of having diabetes group visits at their site. All interviews were audio-recorded and transcribed. Systematic qualitative analysis techniques were used to identify the range and consistency of opinions and experiences across interviewees. RESULTS: The research team visited five health center sites across four states in the Midwest and conducted a total of 26 interviews with health center personnel. One site was rural, three were urban, and one was suburban. Health center personnel noted many benefits of group visits to the health center, providers, and patients. The benefits of group visits to the health center included an alignment of the group visit model with the mission of becoming a patient centered medical home, the ability to bill for group visits, and an efficient way to improve guideline-driven care for patients. Provider benefits from group visits included boost to provider morale, opportunities to collaborate with multidisciplinary colleagues, and having more time to focus on other medical concerns during patients routine follow-up visits. Patients derived many benefits, including receiving education and medical care in a single appointment, e.g. the one-stop shop or best bang for your buck, the opportunity to obtain social support and enhanced motivation through peers, and the potential to improve their clinical outcomes by attending group visits. CONCLUSIONS: Diabetes group visits can offer many unique benefits to safety net clinics by providing patient-centered care, boosting provider morale, increasing multidisciplinary collaboration, and offering patients the convenience of combining an educational session with a primary care appointment. Further studies need to assess best practices in implementing group visits in health centers and assess their impact on patient outcomes. BACKGROUND: The foundation of community based participatory research (CBPR) is collaboration between academic researchers and members of a given community. Community advisory boards (CABs) are one mechanism through which academic-community partnerships are formed, but current research about CAB members opinions on barriers to collaborations with academics is limited. This qualitative study examined CAB members expectations and experiences in working with academic researchers in the field of HIV/AIDS. METHODS: We conducted 10 semi-structured one-on-one interviews with individuals serving on a CAB for HIV-related research at an urban academic medical center. Participating CAB members were leaders of HIV/AIDS community organizations in Bronx, NY and had at least 5 years of experience working in the field of HIV/AIDS. Interview questions focused on participants current and previous experiences with academic research and researchers, trust in these relationships, and best practices for collaboration. Interviews were professionally transcribed and data was analyzed using a grounded theory approach. Transcript data were coded by two independent researchers using NVivo 10 software and analyzed in an iterative process to identify emergent themes. RESULTS: CAB members described positive aspects of inclusion on a CAB, including improved access to information about current HIV topics and the opportunity to help shape HIV research in their communities. However, CAB members also described negative previous interactions with researchers, and a lack of trust in researchers themselves or in the process of conducting research with academic institutions. A major reason for distrust was that power was perceived as unequal, often resulting in miscommunications that led to lasting negative impressions. For example, several members recounted researchers approaching subjects directly, without informing the home organization about the intended study focus. In addition, CAB members felt that expectations for collaboration were not clearly established. CAB members also shared suggestions for improving the collaborative process, including outlining expectations from the outset of the partnership and holding CAB meetings in the community rather than at the academic center. CONCLUSIONS: We identified complex interpersonal issues in relationships between CAB members and academic partners, particularly around trust and powersharing in conducting HIV/AIDS research. Our findings suggest that academiccommunity collaborations could be enhanced by improving transparency, explicitly addressing issues of distrust from previous negative experiences, and ensuring equitable power-sharing among researchers and community members. Further research should be conducted to determine the best ways to engage and maximize the impact of CABs in the field of HIV/AIDS. ARE YOU AFRAID OF PNEUMONIA? PATIENTS FEEDBACK ON THE DEVELOPMENT OF AN EDUCATIONAL VIDEO ABOUT PNEUMOCOC CAL VACCINATION Kenzie A. Cameron; Tiffany Brown; Shira N. Goldman. Northwestern University, Chicago, IL. (Tracking ID #1926320) BACKGROUND: Invasive pneumococcal disease causes a mortality burden among people age 65 and older. Consequently, the Advisory Committee on Immunization Practices recommends pneumococcal vaccination (PnVx) for these adults. However, national rates of PnVx for those age 65 and older are just over 60 %. Further, racial and ethnic disparities in vaccination remain: 2011 estimates reported 66.5 % vaccination among whites age 65 and older, as compared to only 47.6 % among blacks. Informed by our previous research on perceptions toward PnVx among unvaccinated adults, we developed a 2 min patient education video targeted at unvaccinated 65 and 66 year olds to encourage uptake of PnVx. Our objective was to discuss patient attitudes toward vaccination and beliefs about pneumonia, and obtain feedback regarding suggested revisions for the video. METHODS: At an academic general internal medicine clinic, we queried the electronic health record to identify 65 and 66 year old black and white patients who had no record of receipt or refusal of PnVx. In early 2013, patients were recruited to participate in race-concordant focus groups. Prior to the start of each group, participants completed a brief demographic survey, which included questions regarding past experience with pneumonia and PnVx. During the audio-recorded focus group, we asked participants to describe pneumonia, tell us what they knew about PnVx, and specify what they would want included in educational messages about the vaccine. RESULTS: Four white and two black patient focus groups were conducted, with a total of 26 participants. Mean age was 65.9 (3.1), 81 % were female, 88 % had attended college, and 39 % reported receipt of PnVx. Based on our previous research, the video included a strong focus on individual susceptibility to pneumonia, as well as personal testimonial from a black man about his late wife who did not think she needed PnVx. Black patients in particular noted that the testimonial was a persuasive part of the video. Participant response to the video was positive, with some individuals indicating a newly found motivation to consider vaccination (e.g., Even though I refused the pneumonia vaccine already, next time I see my provider, Im going to tell her I want to get shot, and Im going to go get my shot now. I didnt realize how potent and prevalent it is.). However, some participants felt that the video focused too heavily on severity and fear (e.g., I find it dire. Very uncomfortable watching it.). The video included an appeal to consistency and the need for vaccination across the lifespan (Just as you got vaccinated as a child, you need to receive important lifesaving vaccines as an adult), which was met with favorable responses from participants. Several participants highlighted the importance of discussing PnVx with their providers and indicated that viewing this video prior to an appointment could provide an opportunity for such discussion (e.g., I might talk to my doctor. I dont know if she could persuade me to get it. Maybe she could explain it to me differently). CONCLUSIONS: Feedback from the focus group participants led to revisions to the patient education video, specifically with regard to reducing the number of times the severity of pneumonia was stressed. The personal narrative was retained as numerous individuals spoke of its persuasiveness, as was a focus on personal susceptibility to pneumonia and the need for vaccination across the lifespan. The revised video is under 2 min long, has been integrated into our electronic health record system, and is being tested at the point-of-care to assess its effect on receipt of PnVx among patients eligible for vaccination. BEYOND OUR CONTROL: HOW ORGANIZATIONAL CONTEXT IMPACTS PERFORMANCE MEASUREMENT Molly Harrod1; Jane Forman1; Claire Robinson1; Adam Tremblay2,3; Leo Greenstone2,3; Eve A. Kerr1,3. 1VA Ann Arbor Health Care System, Ann Arbor, MI; 2VA Ann Arbor Health Care System, Ann Arbor, MI; 3University of Michigan, Ann Arbor, MI. (Tracking ID #1937968) BACKGROUND: The Patient Centered Medical Home (PCMH) model requires a shift from a physician centric to a team-based approach to care, as well as improvement in process measures that reflect PCMH goals. As a result, as primary care (PC) providers are learning to work in teams, they also must learn how to change their clinic processes to meet PCMH-specific performance measures. The Veterans Health Administration (VHA) transitioned over 900 PC sites to a PCMH model beginning in 2010 and has implemented national metrics of PCMH success that include same day access and continuity with the patients usual provider. One large VHA health system implemented a coaching model to help newly formed PC teams redesign their delivery processes to improve access and continuity measures. This study examines how these PC teams discussed and responded to these measures during their coaching sessions. METHODS: Nine of 20 PC teams were longitudinally observed during coaching sessions (25+ hours) that entailed discussion of both performance measures and processes to improve measure results. Conversations and interactions were recorded via hand-written field notes. We analyzed data using a grounded approach. RESULTS: As PC teams reviewed their access and continuity measures, two predominant, and often overlapping, themes emerged: lack of control and lack of consistency across the organization. PC teams felt they did not have control over processes both within and outside PC. For example, a centralized call center with variable understandings of PC redesign continued to refer patients to the Emergency Department even though PC providers had open appointments within their schedules, thus impacting continuity. A lack of consistency across the organization as a whole was apparent given that many of the other departments did not change their patient care processes to align with PCMH goals. For example, inpatient physicians followed a rule that patients discharged from the inpatient setting needed follow-up with primary care physician in 710 days. Because this 710 day mandate meant fewer open slots in a PC providers clinic, it decreased same day access. Thus, organizational processes were often working against one another resulting measures that were not reflective of all the changes. CONCLUSIONS: Improving even straightforward performance measures, like same-day access, requires an understanding of the entire practice context. In our study, measurement results that fell short of goals reflected the organizational inconsistency of processes that impacted individual measures. It is important to develop processes designed to meet PACT goals across the organization rather than within each team or department. As more PC practices become part of Accountable Care Organizations and are assessed using these and other performance measures, a better understanding of what is being measured, how these measures reflect patientcentered care, unintended consequences of measures and other organizational actions, and how to align organizational performance goals is needed so that organizational performance reflects true quality of care and patient-centeredness, and not just performance on individual metrics. BACKGROUND: The hospital physician peer review process is a valuable method of reviewing physician competency and safety, but it is rarely used to evaluate residents, who frequently work under indirect supervision. In 2008, our institution introduced Code R to our peer review process to highlight the potential resident involvement in cases undergoing peer review. To our knowledge, there are no published studies that utilize a hospital-wide peer review process to evaluate and track residents in training. Our study sought to characterize resident errors using the ACGMEs core competencies to determine whether interns are more prone to errors than senior residents and to see if there were patterns of resident errors by time of day or month of year. METHODS: The study is a retrospective review of all peer review cases recorded from 2008 to 2011 by the Quality Management department at Riverside County Regional Medical Center, an academic safety net institution. The institutional peer review process spans all the departments within the hospital and reviews the cases involving patient deaths, unexpected adverse outcomes, incident reports, quality measure fallouts and random physician quality reviews. Cases designated a Code R were identified and the following information was collected using a standardized form: peer review level assigned, area of deficiency, resident post-graduate year, originating department, month, and time of day. Care issues were characterized into the five core competencies of patient care, medical knowledge, professionalism, communication/interpersonal skills and systems-based practice. The chi-square and Fishers exact tests were used to make comparisons and an a priori statistical level of significance of p value 0.01 was used to adjust for multiple comparisons. RESULTS: Between 2008 and 2011, 8052 individual cases were peer reviewed by the medical staff and 258 (or 3.2 %) had Code R assigned. Of the total charts initially identified, nine cases were determined to not be resident related, nine charts were unavailable and two cases had incorrect medical record numbers. Two hundred thirty-eight charts qualified for final data analysis. The peer reviewed designations and prevalence in each category were Level 1 (Meets standard)- 13 (5.5 %), Level 2 (Meets standard - opportunity for improvement) 196 (82.4 %), Level 3 (Deviation from standard) 26 (10.9 %), and Level 4 (Unacceptable care) 3 (1.3 %). Compared with the non-Code R, or attending physician cases, there were significantly more errors in all the competencies (5.5 % vs. 89.3 % Level 1 assignments for resident vs. attending physicians). In the Code R cases, 60 % (143/238) of the charts had documentation issues. One-third (79/238) of the charts had deficiencies in patient care issues, which include clinical judgment, decision-making and lack of follow-up. The next most common errors were in the competencies of Systems-based Practice and Interpersonal Communication, 9.2 % and 8.0 % respectively. We did not find any significant differences in resident errors by time of day or calendar year. However, interns (PGY-1) were more prone to errors than senior residents (PGY 2) (p = 0.002). CONCLUSIONS: The introduction of Code R helped identify cases in which resident involvement potentially impacted patient care and safety. Our analysis of Code R cases suggest specific areas for focused resident supervision and education are needed in the areas of documentation and patient care. This simple peer review process can be readily adopted by other institutions to help improve resident education, facilitate attending supervision, and potentially improve patient safety. Use of this instrument can be used to track the individual resident and attending performance that has not been previously described. BACKGROUND: Electronic consultations (eConsults) are a recent innovation enabling primary care clinicians to request advice from specialists, via a shared electronic health record (EHR), for lower-complexity clinical problems that do not require an in-person evaluation of the patient by the specialist. Prior studies suggest that eConsults can improve the time-to-access to specialty care and reduce specialty care office visit utilization. Using in-depth interviews, we sought to elucidate primary care providers (PCPs) experiences with a recently adopted eConsult program at an academic medical center. METHODS: We recruited a purposive sample of PCPs from an academic medical center with a recently implemented EHR-based program in which specialists and PCPs receive reimbursement for completed eConsults. Semi-structured interviews were conducted to explore the benefits and limitations of eConsults and their impact on clinical work - including information transfer and clinician-patient and PCPspecialist interactions. Interviews were audiotaped, professionally transcribed, and checked for accuracy. Each interview transcript was reviewed by at least two authors in addition to the interviewer. The constant comparative method was used by the team to collaboratively and iteratively develop themes related to the study objectives. RESULTS: Nine interviews have been completed to date, with participants representing a range of primary care professions and levels of experience, including faculty physicians, medical residents and nurse practitioners. A total of 20 interviews will be completed by March 2014. We identified three key thematic domains related to eConsult benefits, limitations, and influence on clinical work. 1) Access to expert advice: Most participants reported that eConsults were preferable to the use of informal, or curbside, consultations because they offer timely and standardized access to specialist advice without requiring PCPs to call on social networks or ask for favors, and because the consultation is stored in the electronic medical record and is conveniently linked to information about related tests and treatments. As one PCP explained, I like that it is more formal and that it gets into the patients chart and it is recorded and I can refer to it later. Nonetheless, the occasional use of curbside consultations persisted even among PCPs who routinely use eConsults, and was not seen as incompatible with eConsult use. The perception that an eConsult renders the specialist part of a patients care team, while simultaneously obviating the need for an office visit, was also reported as a source of increased patient and PCP satisfaction. 2) Providing better care: The use of eConsults was associated with increased clarity about specialists and PCPs roles, more transparent and efficient care coordination procedures, and PCPs enhanced participation in determining a patients follow-up plan. I just feel like Im providing better primary care, said one PCP. Some participants reported that patient management using an eConsult is more timeconsuming, and entails greater responsibility on the part of the PCP, than a traditional consultation, but most felt that this was an acceptable tradeoff for the satisfying learning opportunities and increased sense of mastery accompanying the incorporation of eConsults in their workflow. 3) Uneven adoption patterns: A minority of participants was less enthusiastic about eConsults and the frequency of use varied widely among individual providers. Some PCPs described a reluctance to initiate an eConsult because of difficulty with technical aspects of the program, frustration with slower-than-promised responses from consultants, or concern about how and when to communicate with patients regarding a consultants recommendations. Others reported pushback from patients who wanted in-person advice from a specialist. Most participants reported weighing both clinical and non-clinical factors when deciding when to request an eConsult. That is why I choose the right patients, said a PCP about the need to balance clinical appropriateness and patient preference when deciding between an eConsult and a conventional referral. CONCLUSIONS: Our results suggest that eConsults present an effective strategy for improving PCPs ability to obtain specialist expertise, and that they increase PCPs engagement and satisfaction with their clinical work. However, eConsults do not uniformly replace informal curbside consultation practices, and decisions about their use are influenced by non-clinical factors such as convenience, usability of the electronic interface, and patient preference. eConsult program designers should elicit continual feedback from providers in order to increase the ease of use and appeal of this mode of consultation. IN THE MILITARY, YOUR BODY AND YOUR LIFE ARENT YOUR OWN: UNIQUE FACTORS INFLUENCING HEALTH BEHAVIOR CHANGE IN OVERWEIGHT AND OBESE VETERANS Melanie Jay1,3; Katrina F. Mateo1,2; Molly Horne2,1; Allison Squires4,2; Adina Kalet2; Scott Sherman1,3. 1VA NY Harbor Healthcare System, New York, NY; 2NYU School of Medicine, New York, NY; 3NYU School of Medicine, New York, NY; 4NYU College of Nursing, New York, NY. (Tracking ID #1920967) BACKGROUND: Obesity affects 36 % of the Veteran population. The Veterans Affairs (VA) Healthcare System offers an intensive weight management program called MOVE!, but only approximately 10 % of eligible patients attend even one session. Since veterans see their PCPs 3.6 times per year, developing primary carebased weight management interventions may improve MOVE! attendance. To inform intervention development, we conducted a single-center qualitative study to explore unique VA patient 1) attitudes, barriers, and facilitators to healthy behavior change; 2) uses and understandings of goal-setting; and 3) weight management-related experiences with health care providers, technology, and MOVE!. METHODS: We sent recruitment letters to overweight and obese patients at the New York Harbor VA. Eligibility criteria included Veterans aged 1875, having a BMI 30 or 25 with at least one co-morbidity, and self-reported ability to read and participate in focus groups. We organized patients into focus groups of 611 participants, grouped by gender and MOVE! attendance. The interview guide was informed by the Theory of Planned Behavior (TPB). Sessions were audio-recorded, contextually supplemented with field notes, and transcribed by a professional company. Initial codes were determined based on session recollection and field notes. Using an iterative coding approach, two coders separately reviewed segmented transcripts, modified the codebook as new codes emerged, and met to negotiate codes. We utilized NVivo software to assist with coding and subsequent thematic analysis. RESULTS: We screened 161 patients for eligibility, 77 were scheduled to attend, and 54 attended one of six focus groups (two female and four male). Participants were predominantly male (63 %), African-American (46 %) or White/Caucasian (32 %), college-educated or higher (74 %), and reported having attended MOVE! (61 %). We found that although we attempted to stratify focus groups by MOVE! attendance, the VA electronic health record did not accurately capture this information, and some study participants were confused or unaware of their attendance. Thematic analyses revealed how military service informed participants attitudes, social norms, and perceived behavioral control with regards to weight management. Participants described how the structured environment and strict standards of the military motivated them to maintain weight control while enlisted, but that lack of autonomy was a barrier to learning how to manage their health when they left service. The military also influenced their perceived need for more personalized and tailored health counseling. Physical activity was largely considered the predominant method to lose weight, and participants felt the MOVE! program lacked a physical activity-related component. While participants acknowledged the potential usefulness of technology to help identify and achieve weight management goals, they expressed the need to have a knowledgeable person (i.e. health coach, PCP, dietician, etc.) provide guidance and support. CONCLUSIONS: This study revealed how military service may influence veterans intentions to lose weight and their weight management-related experiences. This will guide the development of a computer-assisted intervention to treat obesity within primary care and improve access and quality of weight management services. The implications from this study may be limited as it was conducted at a single site within a specific and a mostly self-selected patient population. NONE OF THEM APPLY TO ME: A USABILITY STUDY OF THE VAS MOVE!23 ONLINE WEIGHT MANAGEMENT SOFTWARE IN LATINA WOMEN Hector R. Perez4; Michael W. Nick5; Katrina F. Mateo1,2; Scott Sherman1,3; Adina Kalet2; Melanie Jay1,3. 1VA NY Harbor Healthcare System, New York, NY; 2NYU School of Medicine, New York, NY; 3NYU School of Medicine, New York, NY; 4NYU School of Medicine, New York, NY; 5NYU School of Medicine, New York, NY. (Tracking ID #1920971) BACKGROUND: The obesity epidemic disproportionately affects Latina women, but there are few interventions in this population. The MOVE!23 software program is a publicly available, English-only, online questionnaire used by the Veterans Affairs (VA) Administration to facilitate weight management. It evaluates patients current behaviors and stage of change and provides computer-generated, individualized advice. Completion of MOVE!23 during an intensive weight management program is associated with greater 6-month weight loss outcomes. Thus, adapting this tool for use in Latina women may be beneficial. However, the MOVE!23 is not specifically tailored for use in non-VA populations and its usability in this population is uncertain. The International Organization for Standardization (ISO) defines usability as the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use. In order to determine the utility of tailoring the MOVE!23 for Latina women and translating it into Spanish, we conducted usability testing of this software. METHODS: We recruited English-speaking, adult Latina women. A researcher observed each participant individually as they used the MOVE!23 . Participants were asked to verbalize their thoughts and reactions as part of a Think-Aloud protocol. A semi-structured interview followed that elicited reflections on this experience and how it might influence lifestyle behavior change. We utilized ScreenFlow (Version 4.0.3) to capture on-screen interactions, and each session was digitally recorded and professionally transcribed. A coding guide was derived from the ISO definition of usability and focused on positive and negative interactions between four domains (Tool, Task, User, Context). Two researchers reviewed and coded the transcripts independently using the guide and secondary codes were derived inductively. The researchers met to review the codes and resolve discrepancies. NVivo software was used to facilitate data analysis. RESULTS: Six female Latina participants completed the study (mean age 39, SD = 17); all had lived in the US for more than 20 years. Analysis indicated general satisfaction with tool navigation and the content, sequencing and segmentation of materials. Participants wanted individualized patient handouts to be more accessible and easily available on paper and suggested that images and icons might enhance visual appeal. Significant tool-user interactions included problems with navigation, difficulty understanding questions and answer choices, and sensitivity to the personal nature of the topics. User-context interactions included lack of experience with technology and requests for more support during completion of the questionnaire. In addition, subjects described personal experiences or beliefs while interacting with the materials. The extent to which questions or advice approximated their own experiences influenced the ease with which they were understood. CONCLUSIONS: This study supports the utility of translating and tailoring the MOVE!23 for use in Latina women and identified several areas for improvement. Findings suggest that task execution may be enhanced when users are able to compare and reconcile the online materials with personal experiences and beliefs, SOMEDAY IT WILL BE THE NORM: PHYSICIAN PERCEPTIONS OF THE CLINICAL UTILITY OF WHOLE GENOME SEQUENCING Jason L. Vassy1,2; Kurt D. Christensen2; Melody J. Slashinski3; Denise Lautenbach2; Jill Oliver Robinson3; Jennifer A. Blumenthal-Barby3; Lindsay Feuerman3; Robert C. Green2; Amy L. McGuire3. 1VA Boston Healthcare System, Boston, MA; 2Brigham and Womens Hospital, Boston, MA; 3Baylor College of Medicine, Houston, TX. (Tracking ID #1934488) BACKGROUND: There is excitement that medicine is entering a genomic era, in which knowledge of ones genetic makeup can guide medical decision-making. Whole genome sequencing (WGS) can identify 35 million genetic variants for an individual. However, the clinical significance of most of these variants is currently unknown, and the specific clinical contexts in which WGS might add value to current standards of care have not been identified. The uptake of WGS into medical practice will largely depend on whether and how physicians perceive it to improve patient care. METHODS: We have enrolled primary care physicians (PCPs) and cardiologists to participate with their patients in the MedSeq Project: a randomized trial of WGS in the clinical care of healthy adult primary care patients and cardiomyopathy patients. We used mixed methods to describe the clinical utility that physicians perceive WGS to have. In a baseline survey on enrollment, we asked physicians to rate on a scale of 110 (Not at all useful to Extremely useful) how useful they thought WGS would be for managing [their] patients health at two times: now and in the future. We categorized responses as not useful (15) or useful (610) for analyses. We then conducted in-depth semi-structured interviews to explore whether and in what circumstances physicians might perceive WGS to have clinical utility. Transcribed interviews were consensus-coded with thematic content analysis to identify emergent themes. RESULTS: Of 17 physicians (mean age 52 years), nine were women, and five were of non-white race. Nine were PCPs, and 11 reported no genetics training beyond the typical medical school curriculum. Only 4 (24 %) said that WGS had utility now, but 15 (88 %) predicted it would have future utility for their patients care. We identified three major themes influencing physician perceptions of utility. 1) Inevitability: Physicians expected WGS to be widely integrated into medical practice in the near future. When asked his opinion about the routine use of WGS in clinical care, one responded, Thats like saying, What do I think about getting cholesterol tests on people? I mean, its a reality, and I think its going to happen. 2) Uncertainty: Physicians perceived WGS results to have too little analytic and clinical validity at present to be useful for patient care. One commented, Its like doing a CT scan and finding a nodule and not knowing what it means. And the next thing you know, you have a biopsy, and then you have a complication from the biopsy. 3) Contexts of utility: In particular, physicians anticipated that WGS may 1 day be useful for nudging patient behavior change, tailoring drug therapy (pharmacogenomics), and efficiently targeting preventive health maintenance. One said, If it turns out that we dont have to do colonoscopies every 10 years on 100 % of the people, maybe that will save a lot of money. If we know who we really should be doing PSAs or mammograms on, instead of doing them on everybody, we can save a lot of money, stop creating a lot of harm and disability. Physicians often used analogies to current medical testing to describe their perceptions of WGS. CONCLUSIONS: Physicians are skeptical about the current clinical utility of WGS for general patient care but anticipate a near future when the clinical integration of WGS will be widespread. Evidence that WGS improves health outcomes will likely be necessary to persuade physicians of its utility in patient care. WE TORTURE THEM BEFORE THEY DIE, EVEN THOUGH WE KNOW THAT THEY ARE GOING TO DIE: THE INFLUENCE OF HOSPITAL POLICIES AND CULTURE ON ETHICAL DNR DECISION MAKING IN THE US AND UK Elizabeth Dzeng1,2; Michael P. Kelly2; Sydney M. Dy1; Thomas J. Smith1; Martin Roland2; Stephen Barclay2; David M. Levine1. 1Johns Hopkins School of Medicine, Baltimore, MD; 2University of Cambridge, Cambridge, United Kingdom. (Tracking ID #1939234) BACKGROUND: Physicians face ethical challenges when there are conflicts between autonomy and beneficence, which is defined as acting in the best interest of the patient. Shared decision-making, where the patient and doctor form a partnership to decide on treatment plan, has emerged as a model on the spectrum in between medical paternalism and patient autonomy. In the United States, there are significant variations in hospital policies regarding Do Not Resuscitate (DNR) decision-making. In the United Kingdom, it is standard practice for doctors to make unilateral decisions to institute a DNR order in situations where survival is unlikely. The aim of this study is to explore how doctors in the US and UK balance autonomy and beneficence in decisions regarding DNR status at the end of life and how their ethical values and clinical approach are shaped by their institutions culture and legal frameworks. METHODS: Semi-structured in-depth interviews were conducted with 58 internal medicine doctors across four sites (New York, Baltimore, Seattle and Cambridge, England), who were routinely involved in DNR conversations with end of life patients. Participants were purposively sampled by stage of training and medical subspecialty to provide a wide range of perspectives and contribute to understanding emerging patterns and themes. Interviews lasted on average 60 min and were audiotaped and transcribed. Transcripts were analyzed and double coded using thematic analysis with an interpretive approach. RESULTS: 1316 doctors participated at each site. Approximately equal numbers of senior faculty, junior faculty, fellows and residents were interviewed. Experienced doctors at all sites generally felt comfortable engaging in shared decision-making and when clinically appropriate, making more paternalistic recommendations against resuscitation. However, there was greater variation amongst residents and fellows that was site specific. At each hospital, there appeared to be a dominant mode of decision-making that impacted trainees perspectives on DNR decision-making. Although hospital policies primarily dictate whether patient preferences should be honored, trainees at hospitals with policies that prioritize patient autonomy often interpreted them to mean that they should not provide clinical recommendations. They felt compelled to offer choice even if the chance of successful resuscitation was negligible. In contrast, trainees at hospitals that had policies that encouraged paternalism or shared decision-making, which included the hospital in the UK and one in the US, felt more comfortable expressing clinical judgment against resuscitation when appropriate. Trainees at hospitals that prioritized autonomy more frequently perceived conflict between honoring autonomy and acting with beneficence and recalled more conflicts with patients regarding preferred decision. CONCLUSIONS: Institutional norms and policies influence how doctors develop their professional attitudes and practices regarding DNR decisionmaking. Doctors at the American hospital that encouraged shared decisionmaking, had attitudes and beliefs about DNR decision-making that were more similar to British doctors than doctors at the other American sites. Hospitals characterized by higher levels of shared decision-making and paternalism provided an environment where trainees felt more comfortable withholding resuscitation when success was unlikely and in doing so, experienced fewer patient conflicts. BACKGROUND: Medical students consider a specialtys lifestyle when selecting a career; meeting the physician workforce shortage requires an understanding of how students define a good lifestyle. Understanding the lifestyle expectations of entering students may help medical educators guide students towards careers that will fulfill their ambitions and serve the health needs of our society. The purpose of this study was to determine how first year medical students define a good lifestyle. METHODS: First year medical students at 11 schools were surveyed at matriculation in 2012. The survey included the open-ended question, When someone says That specialty has a good lifestyle, what does that mean to you? Two investigators performed a directed content analysis of responses. Descriptive statistics of the themes were calculated. RESULTS: 60 % (1020/1704) of students completed the survey; 934 answered the open-ended question. Responses encompassed four major themes: schedule control (66 % of students), off-time (54.5 % of students), financial aspects (35 % of students), and work-life (21 % of students). Schedule control was described with multiple subthemes (reasonable hours, flexible hours, predictable hours, and limited call). Nine student responses were categorized as disparaging. CONCLUSIONS: This study has helped establish how matriculating medical students define lifestyle, a known factor in career selection. Matriculating medical students definition of lifestyle is largely defined by schedule control, though students concept of schedule control varies. Salary associated with a good lifestyle ranged from $175,000$400,000. Understanding early perceptions of career decisionmaking may provide a context for medical educators to begin conversations with students about career choice. Additionally, specialties experiencing physician shortages should consider how their specialty will evolve to meet the needs of society while addressing the concerns of trainees. Policies that hope to successfully address the medical workforce shortages must account for students desire to have control of their work schedule. WHEN YOU ARE OLD, HAVE CANCER AND A STORM IS APPROACHING -THE EFFECTS OF HURRICANE SANDY ON CANCER PATIENTS AND PROPOSALS ON POTENTIAL INTERVENTIONS Misako Nagasaka1; Hassan Alsabbak1; Zaid Aljuboori2; Koji Sasaki3; Alfred Burger1; Benjamin Levy1. 1Beth Israel Medical Center, New York, NY; 2Bronx Lebanon Hospital, New York, NY; 3MD Anderson Cancer Center, Houston, TX. (Tracking ID #1939110) BACKGROUND: The estimated loss to the nations economy from Hurricane Sandy has climbed to as much as $50 billion, making it one of the nations most costly disasters. While data on the economic impact of natural disasters are often reported in a timely manner, very little data exists on the impact of natural disasters to health care, especially in cancer care. METHODS: A retrospective chart review was performed to examine the impact of Hurricane Sandy and the subsequent blackout in southern Manhattan on emergency room (ER) visits of patients with a diagnosis of cancer. The records of cancer patients who had visited ER during the week of Hurricane Sandy (10/ 29/2012 to 11/4/2012) were reviewed. For control, ER visits of cancer patients of the same dates 1 year prior (10/29/2011 to 11/4/2011) were reviewed and compared. All data were extracted from the ER and/or inpatient electronic medical records. Collected data included age group, sex, cancer type, chief complaint, ER diagnosis and admission status. The diagnosis of cancer was extracted from the History of Present Illness or Past Medical History and included solid as well as hematologic malignancies. For patients who were admitted, additional data including metastatic disease, comorbidities, length of stay and mortality were extracted. RESULTS: During the week of Sandy, 144 ER visits of cancer patients were identified. One hundred eighteen were identified the prior year. In patients over age 70, 84 visits were observed during Sandy, while only 41 visited the previous year (p value: 0.000143). The chief complaint of shortness of breath (SOB) or asthma was significant, with 21 ER visits in 2012 and only eight in 2011 (p value: 0.045). Out of the 21 visits of SOB, 10 required continuation of their current therapy (8 requested home oxygen therapy and two requested dialysis), while these requests were unobserved the prior year. During the Sandy week, 77 ER visits resulted in admission, while 64 were admitted the year prior (p value: 0.902). Out of 77, seven expired and from 64, four expired (p value: 0.759). The ER admitting diagnoses of the seven who expired during Sandy included a septic patient with fungating head and neck cancer, a brain cancer patient with respiratory failure, a lung cancer patient with pleural effusion, a sarcoma patient with pneumonia, an ovarian cancer patient with anemia, a melanoma patient with acute coronary syndrome and a lung cancer patient with SOB requesting home oxygen therapy. Corresponding data from the year prior included a lung cancer patient with pancytopenia, a CLL patient with pneumonia, a septic patient with breast cancer and a lung cancer patient with suicidal ideation. CONCLUSIONS: Although there were no significant increase in the total ER visits of cancer patients during Sandy, there were significantly more ER visits during Sandy in cancer patients over age 70, suggesting increased vulnerability to this cohort. To our knowledge, this is one of the first few studies to evaluate cancer care related to natural disasters. In retrospect, the admission and death of a lung cancer patient with SOB requesting continuation of home oxygen therapy may have been the one potentially preventable. Cancer patients requiring additional medical assistance such as home oxygen therapy and dialysis may benefit from a natural disaster back-up plan. House calls, early and prioritized evacuation of these patient groups may be some options to minimize the effects of an approaching natural disaster. Top 10 Chief Complaints 30 DAY READMISSION RATE FOR PATIENTS DISCHARGED WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD): ANALYSIS OF 1,858,618 ADMISSIONS. WHICH PATIENTS ARE MOST AT RISK? Devin B. Malik; Sourabh Aggarwal. Western Michigan University School of Medicine, Kalamazoo, MI. (Tracking ID #1939043) BACKGROUND: COPD is a major cause of morbidity and is associated with a high 30-day re-admission rate (RR) and economic burden on health care. This study was done to determine demographic parameters associated with high 30-day RR secondary to COPD METHODS: Nationwide Inpatient Sample data was used to extract data of patients discharged with DM with complications for years 20092011 using clinical classification software (CCS). NIS represents 20 % of all hospital data in US. All the patients who were discharged with primary diagnosis of COPD and readmitted within 30 days were identified and categorized based on admitting diagnosis. Patients were classified as readmissions secondary to COPD as a primary cause, readmissions with COPD as a secondary cause and non-COPD associated readmissions. Statistical analysis was done using SPSS. RESULTS: We identified a total of 1,858,618 admissions for COPD nationwide during the study period with total 30-day RR of 21.06 % and RR of 7.22 % secondary to COPD as a primary cause (Table 1). Age group (4564), males, patients under Medicaid, patients in low median income for zip code and in metropolitan areas had higher 30-day re-admissions secondary to COPD as a primary cause (P < 0.001). Young patients (1844 years of age), females, patients covered under Medicaid and living in Metropolitan areas had higher 30-day RR secondary to non-COPD related causes. CONCLUSIONS: Strategies to reduce 30-day readmissions secondary to COPD should be focused on more susceptible population including males, aged 4564 years old, under Medicaid, having low median income for zip code and staying in metropolitan areas. BACKGROUND: Work conditions in primary care are associated with stress, burnout, intent to leave the practice and quality of care. Time pressured office visits, lack of control of the workplace, a chaotic work pace, and lack of values alignment between clinicians and leaders contribute to stress and burnout. To determine the best ways to organize clinical practice, we measured the effects of work condition interventions on clinician and patient outcomes. Our goal was to assess if a healthy workplace could decrease clinician stress and improve care quality. This abstract reports on clinician outcomes. METHODS: We performed a cluster randomized controlled trial in 34 rural and urban clinics in the upper Midwest and New York City. Primary care clinicians (general internists, family physicians, nurse practitioners and physician assistants) were studied. Seventeen clinics were randomized to participate in quality improvement interventions (QIIs) to improve worklife. Time pressure, work control, pace of work (chaos) and values alignment with leaders were measured at baseline and approximately 1 year later, along with clinician stress, burnout, satisfaction and intent to leave the practice in 2 years. The 17 clinics randomized to the intervention arm received their baseline data in an Office and Work Life (OWL) summary measure. QIIs were chosen by clinics based upon baseline OWL data, and focused predominantly in three areas: 1) improved communication, 2) changes in workflow and 3) clinical QI projects (e.g., a new prescription-refill telephone line, or a medication reconciliation project). Two level multilevel regression analysis was performed to assess the impact on clinician outcomes of being in the intervention arm. When these analyses did not disclose an overall intervention effect, we then explored whether clinicians who showed improvement in burnout, satisfaction or intent to leave the practice were more likely to be found in intervention clinics. Multivariate modeling determined which types of interventions were associated with improvement in these clinician outcomes. RESULTS: Within 34 clinics, 166 clinicians were initially studied, and 139 subjects completed the study. Data were analyzable from 32 clinics and 135 clinicians. Burnout was present at baseline in 41 % of clinicians in intervention clinics and in 33 % of controls. While there was no overall effect of all interventions considered together, there were more clinicians with improvement in burnout in intervention vs. control clinics (21.8 % versus 7.1 %, Odds Ratio (OR) of burnout improvement 4.0, p = 0.014). Burnout was almost six times more likely to improve within clinics performing workflow interventions than in control sites (OR 5.9, p = 0.015), and was also more likely to improve in clinics performing clinical QI projects (OR 4.8; p = 0.021). Clinician satisfaction improved more often in the 10 clinics with interventions in the areas of communication and workflow than in control sites (OR 3.1; p = 0.042). CONCLUSIONS: While these results should be confirmed in future studies, they suggest that organizations can address high rates of burnout by 1) measuring clinician outcomes (satisfaction, stress and burnout), 2) providing feedback on these outcomes to clinicians, clinic managers and staff, and 3) intervening by improving communication and workflow and initiating clinical QI projects. A COMMUNITY PARTNERED APPROACH TO ENGAGE DIVERSE COMMUNITIES IN STROKE DISPARITIES RESEARCH Nazleen Bharmal1; Aziza L. Wright3,2; Stefanie D. Vassar1; Felica U. Jones3; Loretta Jones3,2; Rebekah Wells4; Jason Cienega2; Arleen Brown1. 1University of California, Los Angeles, Los Angeles, CA; 2Charles R. Drew University of Medicine & Science, Los Angeles, CA; 3Healthy African American Families II, Los Angeles, CA; 4University of North Carolina, Greensboro, NC. (Tracking ID #1929958) BACKGROUND: Racial/ethnic minorities have a higher burden of stroke and worse post-stroke outcomes. Contributors to this disparity include lower awareness of stroke signs and treatment, differences in access to care, mistrust in the healthcare system, and low participation in stroke clinical research studies. Despite emerging evidence that community input into interventions may enhance feasibility and sustainability, few stroke studies have incorporated the community perspective. The Los Angeles Stroke Prevention & Intervention Research Program to Reduce Health Disparities (LA SPIRP) is a center to reduce stroke disparities. The Community Engagement Core of LA SPIRP consists of community and academic co-leaders that review projects and work with investigators to understand community research priorities. To enhance public engagement in stroke research, our community-academic collaborative hosted a one-day symposium in South Los Angeles. The objectives of the symposium were to increase awareness of stroke, share information about and receive feedback on the LA SPRIP research projects, and promote collaboration between researchers and community partners to address stroke disparities. METHODS: We used a community-partnered participatory research process to conceptualize the symposium. The symposium started with talks that provided information on stroke risk factors, stroke disparities, and the impact of stroke from the perspective of patients and caregivers. Simultaneous translation was available for monolingual Spanish speakers, and printed materials were available in English, Spanish, Korean, and Mandarin. We surveyed participants about stroke knowledge, stroke disparities, and trust in medical researchers. In the afternoon, attendees participated in six facilitated, small group breakout sessions (one conducted in Spanish) that addressed two questions: 1) What can we do as a community to prevent stroke?, and 2) How can we recruit more diverse populations in stroke research studies? Breakout session notes were analyzed using qualitative data analysis: content analysis used to code notes from group and pile-sorting by community and academic attendees to identify themes. RESULTS: The symposium was attended by 236 community residents; staff from local clinical, social service, and faith-based organizations; and researchers. The survey response rate was 54 %. Over 90 % of the attendees were African American, Latino, or Asian; 40 % had hypertension. The majority of respondents had limited prior knowledge of stroke, and most had learned about stroke from family members, friends, or media. The average increase in knowledge about stroke pre- and postdidactic sessions was 15 %. On average, 21 % of respondents did not trust medical researchers to provide full explanations of research studies or protect participants from harm. Community recommendations for community-based strategies to prevent stroke focused on types of educational efforts (n = 128 quotes), educational content and cultural considerations (79), venues to share information (65), healthy eating efforts (50), partnerships with trusted stakeholders and institutions (34), types of audiences to receive stroke awareness information (30), physical activity efforts (26), and blood pressure screening (14). Community recommendations to enhance minority recruitment into clinical research studies included types of outreach methods (116), communication strategies (72) and partnerships with trusted stakeholders (51) to gain trust, types of incentives for participation (48), and considerations about the relevance and location of the research study for minorities (11). CONCLUSIONS: Community-academic partnered events are effective ways to obtain community input into stroke research and may build trust and foster collaborations with community members for stroke research. Recommendations from the community stroke symposium will be used to frame future annual symposia, inform current LA SPIRP research, and develop smaller, culturally targeted stroke disparities programs in African-American, Latino, Korean, and Chinese-American communities in Los Angeles. Recommendations from Breakout Sessions Commnity strategies to increase diversity in stroke research studies Provide non-monetary incentives (e.g., blood pressure monitors) Research takes place in the community A COMPARISON OF ACTIVE ADVERSE EVENT SURVEILLANCE SYSTEMS WORLDWIDE Jodi B. Segal1,2; Yulin Huang1; Jinhee Moon1. 1Johns Hopkins Bloomberg School of Public Health, Baltimroe, MD; 2Johns Hopkins University School of Medicine, Baltimore, MD. (Tracking ID #1930394) BACKGROUND: Typically, fewer than 5,000 people have been exposed to a drug before marketing, so adverse drug events (ADEs) at low frequency will not have been observed. Drug regulatory agencies and industry are vigorously developing active surveillance systems for drugs, vaccines, and medical devices in order to quickly identify therapies causing harm to exposed patients. We conducted an environmental scan in order to describe the status of active surveillance systems for the detection of ADEs, worldwide. METHODS: We developed a working definition of active surveillance systems. The key components of the systems that we sought to capture were: Structurewhat is the structure, partnership relationships, and funding of existing active surveillance systems? Datawhat are the data used in existing surveillance systems, what patient populations are included, and what is the content of these data? Functionwhat are examples of adverse drug events that have been identified with the active surveillance systems? Paired researchers reviewed published literature, as well as online sources, to identify existing systems. Data was extracted into tables designed to capture the key components of the active surveillance systems. We reviewed and synthesized the information across systems to identify common, and presumably, essential features of active surveillance systems. RESULTS: We identified 12 active surveillance systems meeting our criteria, meaning that they were designed for post-marketing surveillance of pharmaceutical products (including vaccines), had a goal of generating post-marketing drug safety information, do not require personnel to initiate safety reports (describing individual cases), use real-world data which are generated from routine practice, and involve data from more than a single institution. Three systems are U.S. based - the FDA Sentinel Initiative, the Federal Partners Collaboration, and the Vaccine Safety Datalink; two are Canadian systemsThe Canadian Network for Observational Drug Effect Studies (CNODES) and the Vaccine and Immunization Surveillance in Ontario (VISION); and two are Europeanthe EU-Adverse Drug Reaction (ADR) Alliance and the Vaccine Adverse Event Surveillance and Communication (VAESCO). Additionally, there is a Japanese surveillance system called the Medical Information for Risk Assessment Initiative (MIHARI), a trans-Asian system called the Asian Pharmacoepidemiology Network (AsPEN), and one in Shanghai called the Shanghai Drug Monitoring and Evaluative System (SDMES). Two additional systems in the U.K. met our criteria: the Tayside Medicines Monitoring Unit (MEMO) and the Drug Safety Research Unit (DSRU), an independent academic unit that uses diverse European data for pharmacovigilance. The surveillance systems all use administrative claims or electronic medical records for data; some conduct pharmacovigilance explicitly on behalf of a regulatory agency, while others function more independently. CONCLUSIONS: Active surveillance systems are rare. North America and Europe have the most developed systems and most coverage of the population; although some Asian countries are making good advances. There is no coverage of the patients exposed to medications in India, the Eastern European countries, or the South American and African continents. With the ever increasing availability of electronic data, active ADE surveillance should be increasing feasible. This should translate to more rapid signal detection, timely validation, and then steps to assure that drugs and vaccines are used safely. A DESCRIPTION OF NEEDS FOR CARE COORDINATION AT A COMMUNITY-BASED GENERAL HOSPITAL IN JAPANS SUPER-AGED SOCIETY Yukio Tsugihashi. 1Tenri Hospital, Tenri, Japan; 2Tenri Hospital, Tenri, Japan. (Tracking ID #1920518) BACKGROUND: Japans population is aging rapidly because of long life expectancy and a low birth rate. In 2025, it is estimated that the percentage of aged population will increase to 25 %. Furthermore, the standout features of the Japanese healthcare system are universal insurance scheme and unrestricted access to any doctor which add strain to an already overburdened healthcare system. Therefore, a well-designed collaborative system is essential, one that supports the community healthcare and enables functional differentiation and the establishment of referral/ transferal networks among a variety of institutions. The purpose of this study is to investigate 1) a medical service area in a community-based general hospital in Japan, 2) the places where the patients were discharged with professional discharge planning and 3) to estimate the future demographic changes in the number of inpatients. METHODS: Participants: Inpatients discharged from one community-based general hospital in 2012. Setting: Tenri hospital, Tenri city, Nara, Japan (a total of 1001 beds, 26 departments and over 230 medical doctors) Study design: Retrospective cohort study Outcomes: 1) Mapping the medical service area 2) Specifying the places where the participants were discharged from the hospital that provided professional discharge planning 3) Estimating the changes of inpatient numbers and their age groups in 2012, 2025 and 2035. Analysis: First, we described the medical service area by connecting the participants address code and Geographical Information System (GIS). Second, a focus group interview lead by professional physicians was performed to find out which diseases frequently required professional discharge planning. Then, we evaluated the places where these patients were discharged from the hospital. Lastly, we projected the annual change in the number of inpatients and their respective age groups in the major medical service area for 2025 and 2035 respectively. RESULTS: A total of 17,196 patients were included in this study. The medical service area was wide and included some cities outside Nara prefecture, the established government-designated administrative medical area (Figure). In addition, we highlighted five disease categories that needed professional discharge planning: cerebral stroke, fracture, cancer, infectious disease and heart failure categories. The results of professional discharge planning showed that the patients homes were the main place to be discharged to. There were some notable trends found between the places where patients were discharged to and the disease categories (Table). Furthermore, we estimated the change in the number of inpatients living in the top 10 areas to be 12,242 in 2025 (41.2 % of patients being 75 years old or over) and 11,337 in 2035 (42.2 %) compared with 12,088 (30.9 %) in 2012. CONCLUSIONS: We investigated the medical service area of a community-based hospital in Japan highly reflective of its super-aged society and the needs for care coordination based on the actual practical performance. GIS analysis would help the general hospital to foresee the medical service area more concisely while preparing for the collaborative healthcare network. Additionally, to prepare for the rapid aging of inpatients, general hospitals should develop a collaborative healthcare system and a system for discharge planning appropriate to the medical service area and diseasespecific needs for care coordination. Table. Places where the participants were discharged from that provided professional discharge planning 4 (13) 16 (53) 1 (3) *mean SD, **n (%), the most frequent place in each disease category Figure. The medical service area of Tenri Hospital. Each dot represents the density of inpatients number living in each specified address code. Thick lines indicate the borders of cities in Nara Prefecture. *Tenri City. A FAILURE TO COMMUNICATE: CHALLENGES AND SOLUTIONS TO CARE COORDINATION BETWEEN HOSPITALISTS AND PRIMARY CARE PROVIDERS AROUND HOSPITALIZATIONS Christine D. Jones1; Maihan B. Vu2; Christopher ODonnell3; Mary Anderson1; Snehal Patel3; Eric A. Coleman4; Darren A. DeWalt5. 1University of Colorado, Denver, Aurora, CO; 2University of North Carolina at Chapel Hill, Chapel Hill, NC; 3Duke University School of Medicine, Durham, NC; 4University of Colorado, Denver, Aurora, CO; 5University of North Carolina at Chapel Hill, Chapel Hill, NC. (Tracking ID #1933333) BACKGROUND: Care coordination (e.g., information exchange, direct communication) between hospitalists and primary care providers (PCPs) is critical to successful care transitions from hospital to home, yet such coordination is frequently lacking. We sought to understand care coordination challenges and potential solutions from the perspective of both hospitalists and PCPs in North Carolina. METHODS: We conducted a qualitative study with eight focus groups (4 hospitalist groups and four primary care groups, 59 total participants). The team recruited practices from diverse settings; six of the groups shared patients between inpatient and outpatient settings. Interview guides were based on the AHRQ Care Coordination Measurement Framework and included questions about care coordination, information exchange, follow up care, accountability, and medication management. Focus groups were digitally recorded, transcribed verbatim, and analyzed in ATLAS.ti. Two team members independently coded transcripts and resolved differences through discussion. The constant comparison method was used to evaluate differences between hospitalists and PCPs. RESULTS: Overall, we found that hospitalists and PCPs encounter many similar challenges to care coordination, including: 1) lack of time, 2) difficulty reaching the other provider by telephone, 3) lack of personal relationships with the other provider, 4) lack of clarity about accountability for pending tests and home health orders, and 5) discrepancies between what medications patients are taking and should be taking at home. Specific concerns for hospitalists include: 1) uncertainty about whether PCPs receive information about patient hospitalizations, and 2) difficulty obtaining timely follow up appointments for patients, particularly for after-hours or weekend discharges. Specific concerns for PCPs include: 1) not knowing when their patients are hospitalized, 2) not having records available for post-hospitalization appointments, 3) feeling disrespected by hospitalists when medication changes are made for unclear reasons without their involvement, and 4) difficulty locating important information (e.g., pending tests) in discharge summaries. Hospitalists and PCPs identified common themes to achieving successful care coordination: 1) making extra efforts to coordinate care for high risk patient populations (e.g., patients on Warfarin needing INR checks), 2) improving direct telephone access to each other, 3) improved information exchange through sharing EMRs with improved integration of care transitions workflow into EMRs (e.g., enhanced messaging capabilities, scheduling capabilities for follow up), and 4) enhancing personal relationships with clinicians in the other setting. CONCLUSIONS: Hospitalists and PCPs encounter many similar challenges to care coordination, yet have important experiential differences related to sending and receiving roles for hospital discharges. Challenges that were cited in one setting were often not experienced by clinicians in the other setting. Efforts to improve care coordination between hospitalists and PCPs should aim to understand perspectives of clinicians in each setting. Such solutions include: 1) enhancing direct communication between providers in different settings, 2) clearly defining accountability for tests and home health orders, and 3) building personal relationships between clinicians in each setting. A HEALTH LITERACY INTERVENTION TO IMPROVE MEDICATION ADHERENCE USING MEDUCATION TECHNOLOGY Leah L. Zullig; Felicia McCant; S. Dee Melnyk; Susanne Danus; Hayden bOSWORTH. Durham Veterans Affairs Medical Center, Durham, NC. (Tracking ID #1926044) BACKGROUND: Many Americans are reliant upon medication management for chronic conditions. Medication non-adherence is rampant, leading to increased healthcare cost and worsened health outcomes. Low health literacy is a contributing factor to medication non-adherence. Our objective was to determine whether antihypertensive medication adherence could improve using a Meducation technology health literacy intervention. METHODS: We conducted a six-month feasibility study among patients with cardiovascular disease risk factors who received care from hospital-based primary care clinics associated with Durham Veterans Affairs Medical Center (VAMC). To enhance potential benefit and future generalizability, the intervention was integrated into clinics practice tailored to address needs of high risk, potentially non-adherent patients. All patients received the intervention, which included a personalized Meducation calendar. We evaluated changes in medication adherence and clinical outcomes at 6 months. Potential participants were initially identified through an electronic health record data pull. To be eligible for study inclusion, patients must have met all of the following criteria: enrolled in one of three primary care clinics affiliated with the Durham VAMC for at least 1 year; had at least one visit to an affiliated primary care provider in the previous 12 months; an outpatient diagnostic code for hypertension (ICD9 401.1, 401.9) diabetes (ICD9 250.x), and/or hypercholesterolemia (ICD9 272.x); prescribed metoprolol tartrate (beta blocker) and have poor medication adherence for the 12 months prior to baseline (medication possession ratio <80 %); and be prescribed more than nine active medications. RESULTS: There was a 42 % enrollment rate (n = 23). The sample was predominantly married (57 %), African American (61 %), and male (91 %). Few participants (9 %) completed less than 12 years of school. Forty percent had low health literacy, defined as less than 9th grade reading level. The majority of participants were diagnosed with hypertension (91 %), hyperlipidemia (87 %), or diabetes (61 %). At 3 months self-reported medication adherence improved. There was a 45 % decrease in participants who agreed or strongly agreed that they sometimes forget their heart medications and 58 % fewer participants reported that they agreed or strongly agreed that they were sometimes careless with how they take their medications. At 6 months, medication possession ratio improved nearly 13 % and there were decreases in patients average systolic blood pressure (0.5 mmHg), diastolic blood pressure (1.5 mmHg), and body weight (3.6 lb) (p > 0.05). CONCLUSIONS: A health literacy intervention is a feasible mechanism to improve cardiovascular-related medication adherence and outcomes. In this sample, 9 % of participants attained less than a high school level of education, yet 40 % had health literacy levels below the 9th grade reading level. This is problematic as much health promotion and medication adherence information is targeted to a tenth grade reading level, which may be inaccessible to many patients. Interventions to improve adherence through health literacy have the potential to require relatively few resources to implement, making them particularly important tools to combat medication non-adherence. Future interventions may target and/or oversample patients at greatest risk for health literacy-related medication non-adherence. A LONGITUDINAL ANALYSIS OF THE EFFECT OF TOTAL OPIOID ANALGESIC DOSE ON SUBSEQUENT USE AND DURATION OF HOSPI TALIZATION Barbara J. Turner1,3; Benjamin R. Ehler2; Yuanyuan Liang2,3. 1University of Texas Health Science Center at San Antonio, San Antonio, TX; 2University of Texas Health Science Center at San Antonio, San Antonio, TX; 3University of Texas Health Science Center at San Antonio, San Antonio, TX. (Tracking ID #1937908) BACKGROUND: High opioid analgesic (OA) doses for chronic non-cancer pain (CNCP) increase the risk of adverse events such as drug overdose and car accidents but it is not known whether these drugs are associated with an increased risk of future all cause hospitalization. Objective: To conduct a population-based study of the association of total dose of filled OA prescriptions (Rxes) in a baseline 6 month interval and patient age with any hospitalization and duration of hospitalization in the next 6 months. METHODS: For persons enrolled >1 year from 1/2009 to 7/2012 in Aetnas national HMO (2.1 million), 278,975 persons aged 18 to 64 years were selected if they filled 2 Rxes for non-injectable Schedule II or III OAs. From claims and enrollment files for these subjects, a database was developed including demographics, encounters, ICD-9-CM codes, and filled Rxes. For this analysis, we selected persons aged 4564, no cancer diagnosis (except basal cell) within 6 months of an OA Rx, and no OAs for opioid dependence. Follow-up after the first OA Rx was divided into 6-month intervals until 1/2012 (excluding incomplete intervals). The total morphine equivalent dose (TMED) was calculated from all filled OAs Rxes for each 6month interval and categorized for analysis as none and approximate quartiles: NO risk (0 days); LOW risk (1190 TMED); MEDIUM risk (191450 TMED); HIGH risk (4511830 TMED); and VERY HIGH risk (>1830 TMED). Days covered by filled Rxes for other drugs used for CNCPbenzodiazepines, zolpidem, and antidepressantswere analyzed per 6-month interval. From ICD9 codes, we created indicators for five categories of CNCP, mental health conditions, and substance abuse. Repeated measures logistic regression was used for hospitalization (Y/N) and repeated measures Poisson regression was used for inpatient days. All analyses adjust for patient demographic, clinical, and treatment factors. No significant interaction was observed for OA risk and age; therefore, we only report results for OA risk. RESULTS: The cohort totaled 109,292 persons with a mean age of 53.7 years (SD = 5.5), 54.7 % female, and residence in South (46.9 %), West (16.8 %), Midwest (6 %) and Northeast (30.3 %). In the cohorts 253,531 six-month follow-up intervals, the proportion hospitalized was 6 % and median hospital days was 4 (IQR = 36). For hospitalization, the adjusted odds ratio [95 % CI] for OA risk categories vs. NO risk were: VERY HIGH: 1.45 [1.38, 1.54]; HIGH: 1.15 [1.08, 1.21]; MEDIUM: 0.98 [0.93, 1.04]; LOW: 1.04 [0.98, 1.10]. For inpatient days, the incident rate ratio [95 % CI] for OA risk vs. NO risk were: VERY HIGH: 1.47 [1.34, 1.62]; HIGH: 1.10 [1.01, 1.20]; MEDIUM: 0.87 [0.80, 0.95]; and LOW: 0.95 [0.87, 1.03]. CONCLUSIONS: High or very high total dose of filled OA Rxes in a baseline 6month interval is associated with an increased risk of subsequent hospitalization and more inpatient days independently of age, pain-related comorbidities, and other prescribed medications commonly used for pain. Reasons for these hospitalizations need to be examined to inform interventions to address this high use of costly inpatient care. A MIXED METHOD STUDY OF THE QUALITY OF WRITTEN FEEDBACK BY ATTENDINGS OF INTERNAL MEDICINE RESIDENTS Cynthia Kay1; Jeffrey L. Jackson2,1; Wilkins Jackson3. 1Medical College of Wisconsin, Milwaukee, WI; 2Zablocki VAMC, Milwaukee, WI; 3University of Wisconsin-Milwaukee, Milwaukee, WI. (Tracking ID #1920218) BACKGROUND: Residents uniformly criticize the quality of feedback they receive. The purpose of this study was to evaluate the quality of feedback in an internal medicine residency program. METHODS: Subjects for this retrospective analysis were Medical College of Wisconsin (MCW) internal medicine residents who completed residency training between 2004 and 2012. Over the 8 years, there were 6603 evaluations by 334 attendings. These evaluations followed ACGME guidelines and assessed residents in six domains and included a comment box. From these 6603 written comments, we randomly selected 500 and coded them using grounded theory methods. An initial coding tree was developed by two raters (CK, JLJ). Once the tree was saturated, 25 % of utterances were double coded with good inter-rater reliability (Spearman: 0.94). Utterances were further classified as high, moderate or low quality, also with good inter-rater reliability (Quadratic Kappa: 0.83). Focus groups were also conducted with medicine faculty and housestaff. First written comments were placed on notecards and the groups sorted the cards into high, moderate and low quality feedback. As they did this, they discussed the characteristics that merited this rating. The last few minutes of each focus group was spent reflecting and developing criteria for high, moderate and low quality feedback. RESULTS: The majority of the attendings 500 written comments were rated as moderate quality (65 %); 22 % were rated as high and 13 % low. The most common type of feedback (29 %) was nonspecific comments, often about the residents personality, 16 % about patient care, 14 % interpersonal communication, 7 % medical knowledge, 6 % professionalism, 4 % each commented on practice based learning or systems based practice. The majority of written feedback comments (88 %) were positive, 8 % were negative and 4 % were neutral; three ABIM domains were more likely to include negative comments (medical knowledge: OR: 3.5, 95 % CI: 2.25.6; Practice Based Learning: OR: 2.5, 95 % ci: 1.34.8, Systems Based Practice: OR: 4.6, 95 % CI: 2.58.3). Nonspecific comments and comments on residents attitude or personality were much less likely to include negative comments than the other domains (nonspecific: OR: 0.22, 95 % CI: 0.130.39, attitude/personality OR: 0.53, 95 % CI: 0.340.82). Focus groups uniformly sorted quality into low, moderate and high quality feedback based on three criteria: specificity, quantitative ratings, actionable, Low quality feedback was nonspecific, moderate feedback had at least some specificity and high quality feedback included specificity plus either some aspect of quantification or was actionable. Despite the uniformity of the criteria espoused by focus groups, the inter-rater reliability between the focus groups was low, (quad kappa: 0.24). CONCLUSIONS: Written feedback suffers from lack of balance, lack of specificity and lack of actionability. Much of it is focused on personality characteristics of the resident. While the focus groups developed nearly identical criteria for quality, agreement between the groups was low because the groups disagreed as to what constituted specific and actionable feedback. Residents are justified in expressing chagrin at the quality of feedback they are receiving. A MIXED-METHODS EVALUATION OF A VA PATIENT-CENTERED MEDICAL HOME FOR SPECIALTY CARE Michael Ho2; Catherine Battaglia1; Leah M. Haverhals2; Katherine M. Fagan2; Jackie G. Szarka3; Rebecca Alexander3; George Sayre3; Christian Helfrich3; Susan Kirsh5; David Au3; David Aron6; Julie Lowery4. 1Eastern Colorado Health Care System, Denver, CO; 2Eastern Colorado Health Care System, Denver, CO; 3Puget Sound Health Care System, Seattle, WA; 4Ann Arbor Health Care System, Ann Arbor, MI; 5Case Western Reserve University, Cleveland, OH; 6Louis Stokes VA Medical Center, Cleveland, OH. (Tracking ID #1938016) BACKGROUND: The Veterans Health Administration (VA) implemented a pilot program called Specialty Care Neighborhood (SCN) to align with the VAs Patient Centered Medical Home to deliver comprehensive patient-centered care. SCN is a team-based specialty care model, in which teams of specialists engage in proactive and long term collaboration with primary care in order to provide coordinated, optimal care for Veterans with the greatest specialty care needs. SCNs were implemented at 25 VAs in 20 specialties including Heart Failure, Diabetes, Cancer, Pain, Orthopedic Surgery, and HIV. We completed an evaluation of the SCN initiative for the VA Office of Specialty Care (OSC). METHODS: Findings are presented here from semi-structured interviews and staff surveys. A goal of the evaluation was to identify factors that differentiated four high from two low performing sites as determined by OSC from regular site reports on implementation progress. The evaluation consisted of 27 semi-structured interviews with Primary Care Providers, specialists, and key stakeholders (administrators, chiefs of staff, and facility directors) at the participating sites. Additionally, SCN staff from the six sites returned 39 surveys about their experience with SCN. Interview and survey data were coded using the Consolidated Framework for Implementation Research. RESULTS: SCNs were viewed positively. Participants from all sites described improved professional relationships and communication as well as improved quality and quantity of specialty services. Interviewees felt improved relationships led to increased job satisfaction and engagement. Participants expressed frustration as to lack of clarity regarding how to implement SCN, and despite this barrier, pulled together to creatively implement SCNs best adapted to meet local needs. In structured ratings of implementation facilitators and barriers, High performing sites, in contrast to low performing sites, engaged leadership, utilized a thoughtful planning process that allowed for ample implementation time, and used systems redesign techniques to streamline processes. Additionally, there were factors from interview data that facilitated implementation at both high and low sites: 1) positive knowledge and beliefs about the program; 2) obtaining patient and provider feedback; 3) recognizing patient needs; 4) active champions; and 5) ability to adapt the program easily. Common barriers included 1) problems with hiring staff, 2) lack of commitment to long-term staff funding, 3) lack of administrative support, 4) insufficient materials for facilitating implementation, and 5) coordination and communication challenges among different providers and services. Interview data and staff survey data were largely congruent. However, survey participants reported high leadership engagement despite interview data suggesting this was less positive among low performing sites. CONCLUSIONS: The findings were used to identify areas on which to focus attention for improving dissemination of the program. Recommendations were based on factors that distinguished high from low performing sites and included recommendations to overcome barriers: 1) Provide opportunities for sharing best practices across sites; 2) Encourage innovation through site-based flexibility and individualization while limiting micromanaging; 3) Facilitate access to systems redesign expertise; 4) Minimize complexity by utilizing programs in place; and 5) Begin planning process early and involve all key stakeholders. A MODEL TO PREDICT RISK OF CERVICAL ABNORMALITIES IN PRIMARY CARE Michael B. Rothberg; Bo Hu; Sarah Schramm; Kelly Nottingham; Glen Taksler; Xian W. Jin; Laura Lipold; Andrea L. Sikon. Cleveland Clinic, Cleveland, OH. (Tracking ID #1937479) BACKGROUND: Cervical cancer screening tests are highly sensitive but have low specificity, resulting in high false positive rates among low-risk women. Although there are known risk factors for cervical cancer, there are no validated tools for assessing risk in primary care. Using data from electronic health records, we derived and validated a risk prediction model for cervical intraepithelial neoplasia grade 2 or higher (CIN2+) lesions. METHODS: Our dataset included information on all women aged 30 years who underwent cervical cancer screening with cytology and human papilloma virus (HPV) testing at 33 physician practices associated with the Cleveland Clinic between 2004 and 2010. Patient risk factors at the time of screening (age, race, marital status, insurance type, smoking history, median income based on zip code and physician practice) were determined by reviewing the electronic health record. Pathology results were obtained from the Cleveland Clinics pathology database (for patients who underwent colposcopy with biopsy) or by imputation (for patients with abnormal cytology who did not undergo colposcopy). Imputation was based on observed associations between patient risk factors, HPV and cytology results and CIN2+ status among women who underwent colposcopy. Using CIN2+ as our outcome, we then derived a hierarchical generalized linear model with a logit link that included all patient risk factors available before screening (i.e. not HPV or cytology results) as independent variables. We then stratified patients into tertiles of low, medium and high predicted risk of CIN2+. We then compared the actual rate of CIN2+ (either observed or imputed) in the low-risk tertile to an implied threshold value, which was based on combinations of HPV and cytology results for which colposcopy is typically not recommended. The threshold was 0.20 %, which is equivalent to the rate of CIN2+ among patients with low risk ASCUS and negative HPV. The final model was validated using 100 bootstrap iterations for the entire population and noting the proportion of low risk patients who had CIN2+. RESULTS: The dataset included 99,388 women. Of these, 574 had pathology findings of CIN2+ and 175 had CIN2+ assigned by imputation (overall rate of CIN2+=0.75 %). The final multivariable model had a c-statistic of 0.73. All factors but race were associated with CIN2+: older age (OR 0.69 per 5 years, 95 % CI 0.66, 0.73), Medicare vs. private insurance (OR 0.51, 95 % CI 0.34, 0.75) and higher income (OR 0.97 per $5 K, CI 0.94, 0.99) were negatively associated, while being single vs. married (OR 1.50, 95 % CI 1.24, 1.82) and smoking (OR 1.66, 95 % CI 1.35, 2.04) were positively associated. Tertiles of predicted risk were 0.20 %, 0.53 %, and 1.51 %, with observed rates of CIN2+ over the same tertiles of 0.19 %, 0.52 %, and 1.53 %. Therefore, before cervical cancer screening, women in the lowest risk tertile had observed CIN2+ risk slightly below that of women with ASCUS and HPV negative (0.20 %). In 100 bootstrap iterations, the observed risk in the lowest tertile ranged from 0.13 to 0.25 % (IQR 0.18 % to 0.22 %). CONCLUSIONS: A multivariable model based on data derived from the electronic health record was able to risk stratify women across an 8-fold gradient of risk for cervical abnormalities and to identify a large group of women who are at very low risk for cervical cancer and may not require screening. After further validation, use of a similar model could enable more targeted cervical cancer screening. A NATIONAL SURVEY OF PATIENT CENTERED MEDICAL HOME INITIATIVES Samuel T. Edwards1,2; Asaf Bitton3,4; Johan S. Hong4; Bruce E. Landon4,5. 1Boston VA Healthcare System, Boston, MA; 2Harvard Medical School, Boston, MA; 3Brigham and Womens Hospital, Boston, MA; 4Harvard Medical School, Boston, MA; 5Beth Israel Deaconess Medical Center, Boston, MA. (Tracking ID #1932105) BACKGROUND: The reorganization and improvement of primary care through the implementation of the patient-centered medical home (PCMH) is central to healthcare delivery system reform efforts. PCMH initiatives, often organized by health plans, states or multi-stakeholder groups, seek to create the appropriate context to promote transformation of participating primary care practices into medical homes by building learning collaboratives, providing practice coaching, and reforming physician payment to support medical home services. Although there has been expansion of PCMH activity across the country, the current national scope of PCMH initiatives remains undefined. To address this gap, we conducted a nationwide survey of PCMH initiatives. We describe their size, scope, organization and payment methodology and assess their approach to practice transformation and program evaluation. METHODS: We performed a cross-sectional survey of all PCMH initiatives nationally that included payment reform as an operational feature. We identified initiatives for the survey from existing databases supplemented by formal literature review, internet searches, and by contacting known experts, state Medicaid directors, and representatives from all major national health plans. We included initiatives that were active as of February 1st, 2013 but excluded programs directed solely toward specific medical conditions or serious mental illness. The survey included descriptive questions including initiative start and end dates, numbers of participating practices, physicians and patients, and identity of key stakeholders. We asked about the selection process for practices to enter the initiative, the use of formal PCMH recognition, and payment methods. Finally, we asked about how initiatives facilitated practice transformation and plans for program evaluation. The survey was fielded from February 1st 2013October 31st 2013 via web or telephone. Respondents were physicians, senior administrators, or senior health policy experts from each initiative. We present descriptive statistics to compare initiatives sponsored by commercial health plans, state Medicaid programs, and multiple payers. RESULTS: Of 119 identified initiatives, we received 114 responses (96 % response rate). Forty-one programs were small commercial single payer initiatives, 29 were large single payer initiatives, 23 were Medicaid only and 21 were multi-payer. In total, the programs included 63,011 providers caring for 20,389,327 patients nationally. Small commercial programs had a median of 3,896 patients and typically included a small number of practices (median = 4). Large commercial single payer initiatives included a median of 160,000 patients, often covering most or all practices in a specific region or an entire state (median 105 practices). Medicaid initiatives were typically large statewide programs (median 224,040 patients) run directly by state Medicaid agencies. Multi-payer initiatives included a median of six health plans (IQR 47) with a median of 187,343 patients, and 68 practices. These initiatives included commercial health plans, Medicare and Medicaid. Formal PCMH recognition was required by 69 % of initiatives. Nearly half of initiatives use practice consultants (46 %) or learning collaboratives (48.7 %), but this was more common in multi-payer initiatives (90.5 % and 95.2 %, respectively). Nearly all initiatives had a formal evaluation plan (92 %). A combination of fee for service payments (FFS) for office visits, a per-member-per month care management fee (PMPM), and pay for performance bonuses was the most common payment structure (55 % of programs). Median PMPM was $4.90 (range $0.25$60.00), but was higher in multi-payer initiatives (median $7.00). Since 2009, fewer initiatives are time-limited pilots, shared savings programs are more commonly used, and more initiatives are performing formal evaluations (table). CONCLUSIONS: In this national survey of PCMH initiatives, we demonstrate there has been a large expansion of PCMH activity over the last several years, with 114 initiatives representing over 60,000 providers and nearly 21 million patients in 2013. Initiatives are on average larger, feature more robust transformation efforts, and have higher care management fees. This expansion of PCMH initiatives suggests that the PCMH will continue to play an important role in national delivery reform efforts. Comparison of PCMH initiatives in 2009 and 2013 2009 (N = 26) A PROFILE OF HOSPITALIZED SMOKERS AND THE CARE THEY RECEIVE Ellie Grossman1; Alissa R. Link2; Erin Rogers2,3; Binhuan Wang2; Scott Sherman2,3. 1NYU School of Medicine, New York, NY; 2NYU School of Medicine, New York, NY; 3VA New York Harbor Healthcare System, New York, NY. (Tracking ID #1939163) BACKGROUND: Identification and treatment of patients who smoke is a priority treatment area for hospitals. Hospitalization can be a teachable moment for patients contemplating behavior change, and they may be newly motivated by health consequences. Evidence suggests that appropriate treatment should begin during an inpatient stay and continue for at least four weeks post-discharge. To optimize success, treatment should include both counseling and medications. Patients with mental illness tend to smoke at higher rates than those without, and tend to die earlier of cardiovascular disease, but there is a paucity of evidence about the state of treatment offered to these patients. We aimed to describe the care offered to hospitalized smokers at two urban safety-net hospitals. METHODS: As part of a randomized controlled trial of smoking cessation interventions focused on the post-discharge time period, we assessed the care for smokers hospitalized at two large public hospitals in New York City. Adult patients were eligible for study inclusion if they: smoked at least one puff of a cigarette in the past 30 days; spoke English, Spanish, or Mandarin; were not incarcerated or in police custody; were not pregnant or breastfeeding, and had a U.S. phone number. At the time of discharge, study staff reviewed participants electronic medical records (EMR) for information about type of inpatient unit, history of medical or psychiatric illnesses, and tobacco treatment received during hospitalization and on discharge. RESULTS: From July 2011 to December 2013, we enrolled 1486 study participants. 511 (35 %) were on psychiatric inpatient units and 934 (65 %) were on medical/ surgical units (n = 41 missing). Mean age was 47.9 years, 79 % were male. Of these patients, 17 % had EMR documentation of coronary artery disease, 12 % chronic obstructive pulmonary disease, 14 % asthma, 5 % cerebrovascular disease, 4 % congestive heart failure, 55 % significant alcohol or drug use history, and 10 % schizophrenia or schizoaffective disorder. We found EMR documentation that smoking status was assessed at admission in 1444 (97 %) and bedside counseling was provided in 1294 (87 %). Of those that received counseling, the EMR indicated that 1015 received it from nurses and 145 received it from physicians (for the rest the personnel was not specified). For 647 patients (44 %), physicians ordered smoking cessation pharmacotherapy during hospitalization (36 % NRT patch, 3 % NRT gum, 2 % NRT lozenge, 1 % other). On discharge, 274 patients (18 %) were prescribed cessation medication. Patients on inpatient psychiatric units were more likely to be counseled by their physician regarding smoking cessation (odds ratio (OR) 1.93, 95 % CI 1.203.09) and were more likely to be prescribed cessation medication in hospital (OR 3.21, 95 % CI 2.514.11) than were patients on non-psychiatric units. There was no association between type of inpatient service (psychiatric vs. non-psychiatric) and likelihood of prescription of smoking cessation medication at discharge. CONCLUSIONS: At urban safety-net hospitals, patients who smoke have high rates of mental illness and substance use disorders. Although assessment and counseling are provided at relatively high rates, there is much room for improvement in prescribing smoking cessation pharmacotherapy in the hospital and on discharge. A QUALITATIVE ANALYSIS OF HOSPITAL LEADERS OPINIONS ABOUT PUBLICLY REPORTED MEASURES OF HEALTH CARE QUALITY Sarah L. Goff1,2; Tara Lagu2,1; Amy S. Shatz2; Penelope S. Pekow2; Peter K. Lindenauer2,1. 1Baystate Medical Center/Tufts University School of Medicine, Springfield, MA; 2Baystate Medical Center, Springfield, MA. (Tracking ID #1937659) BACKGROUND: Public reports of hospitals performance on quality measures are intended to both encourage hospital leaders to stimulate improvement activities and to enable consumers to make more informed decisions about where to seek care. How effective public reporting is at fostering quality improvement may be influenced by whether and to what extent hospital leaders find the measures valid and meaningful. We sought to qualitatively describe hospital leaders views regarding quality measures reported on the Centers for Medicare and Medicaid Services Hospital Compare website. METHODS: We mailed or e-mailed a 21-item questionnaire to the chief executive officer and the senior executive responsible for quality at a stratified random sample of U.S. hospitals. Stratification was based upon performance on selected quality measures to achieve a balance of high, average and low performers. While the quantitative portion of the study assessed attitudes regarding publicly reported quality measures, opinions regarding quality measures were elicited using an open-ended prompt at the end of the survey, Please share your additional thoughts about publicly reported measures of healthcare quality, including strengths or weaknesses of current measures, ideas for new measures, etc. We qualitatively analyzed responses to this prompt, identifying the presence and frequency of pertinent themes in an iterative process. RESULTS: Leaders (n= 480) from 380 of the 630 hospitals surveyed (60 %) completed the questionnaire. Of those who completed the questionnaire, 146 leaders from 131 hospitals responded to the open-ended prompt. The majority of respondents to the open-ended prompt were chief executive officers (27 %), chief quality officers or vice presidents of quality (17 %) or chief medical officers (19 %). Half of the responses to the prompt came from leaders at urban hospitals and 68 % were from leaders at non-teaching hospitals. Most respondents were from hospitals that had fewer than 200 beds (41 %), with 23 % from hospitals with more than 400 beds. Forty-one percent of the respondents hospitals had performance scores higher than the national average and 27 % scored lower than the national average. Major themes identified included: 1) Positive views of the value of quality measures (12.5 %, e.g., measuring quality can stimulate change and enhance patient engagement); 2) Concerns regarding the validity and utility quality measures (63 %, e.g., measures fail to measure what is important, inadequate risk-adjustment); 3) Challenges to improving quality scores (e.g., inability to impact factors affecting care outside of the hospital, such as patient adherence to prescribed medication); and 4) Recommendations for improving measures (e.g., compare similar hospitals to each other). Many of the positive views expressed were general support for the concept of measuring quality: Oversight and comparison with peers/standards is necessary and beneficial. Criticisms of quality measures included concerns about whether doctors can actually influence change in the entities measured: the scores for which the doctors are held accountable are subject to profound influence by others besides the doctor to whom the score is attributed. Challenges described included lack of resources: Resources are stretched to the limit with meaningful use and data reporting! Recommendations for improvement in quality measures included an emphasis on the continuous learning cycle, suggesting new targets for quality improvement should be identified as older measures reach acceptable levels of compliance: care measures should evolve as compliance becomes ingrained thus continuing the improvement cycle. CONCLUSIONS: Among hospital leaders who responded to an open-ended prompt eliciting their views about quality measures some expressed support for the concept of measuring quality, while the majority offered criticisms concerning the validity and utility of current quality measures or described challenges to improving quality. As current quality measures are reassessed and new measures developed, the concerns of those responsible for implementing system changes to address quality should be taken into account. A QUALITATIVE ANALYSIS OF PATIENT FACTORS ASSOCIATED WITH HOSPITAL READMISSION Colleen M. Rafferty; Tyler Sechrist; Heather Stuckey; Nicole Swallow; Cynthia H. Chuang. Penn State Hershey Medical Center, Hershey, PA. (Tracking ID #1936483) BACKGROUND: Given the focus on hospital readmissions as a quality indicator, health care organizations nation-wide are interested in finding ways to reduce hospital readmissions. Existing readmission prediction models have relied mainly on clinical and administrative data. A weakness in these methods is that they are not informed by the patient perspective. Exploring patient experiences with readmission may broaden our understanding of what leads to readmission, thereby facilitating a more patientfocused approach to identifying high-risk patients, and enabling institutions to allocate resources to the highest risk patients and to the most cost-effective and efficacious interventions. We conducted qualitative interviews with readmitted patients to identify more precisely the patient-centered factors associated with readmission. METHODS: We conducted semi-structured interviews with a convenience sample of hospitalized patients readmitted to the Internal Medicine service within 30 days of discharge to home from a previous hospitalization at our institution, a 500-bed tertiary care academic medical center. Patients were eligible if they received their primary medical care within our medical system and were readmitted from JuneAugust 2011 (n= 19). Open-ended interview questions addressed the patients perspectives on the medical care they received during the previous admission, access to outpatient care, the patients knowledge of his/her medical conditions and treatments, ability to provide self- care, and social support available to them. We also encouraged the patient to discuss from his/her perspective what could have been done differently to help prevent the readmission. The contents of the interviews were transcribed and a thematic analysis was conducted. We also conducted a medical chart review to determine if the patient had a scheduled follow-up appointment that they attended. RESULTS: We found that the majority of patients had a limited understanding of their medical conditions and their medications. Many patients relied heavily on an advocate (family member or friend) to assist them in their disease management. The vast majority of patients had identified advocates, but we found that not all advocates were effective in providing the needed assistance. Also, most patients did not view their health as under their control, but identified an external locus of control (i.e. the doctors, the disease process, God). All patients identified a primary care physician, and all had been scheduled for follow up appointments within 524 day after discharge, but 9 of 19 patients were readmitted prior to their scheduled appointment (8/19 readmitted within 7 day, 10/19 within 10 day). The majority of patients did not feel their readmission was preventable, but attributed it to the disease process, or an issue that was not viewed by the patient as avoidable or predictable. CONCLUSIONS: These findings suggest potential patient-centered factors that may be used to help identify patients at high risk for readmission: insufficient understanding of medical condition or medications, lack of an effective advocate, a belief that ones health is not under the ones own control, and lack of timely access to outpatient follow up care. Early assessment of these risk factors may help target appropriate interventions to high risk patients to help prevent hospital readmission. A QUALITATIVE ASSESSMENT OF CPOE AND THE EFFECTS OF VARIATION IN DRUG NAME DISPLAY Alexandra Robertson1; Arbor J. Quist1; Thu-Trang Thach1; Lynn A. Volk2; Adam Wright1,3; Shobha Phansalkar1,4; Sarah P. Slight1,5; David W. Bates1,3; Gordon D. Schiff1,3. 1Brigham and Womens Hospital, Boston, MA; 2Partners HealthCare Systems, Inc., Wellesley, MA; 3Harvard Medical School, Boston, MA; 4Wolters Kluwer Health, Indianapolis, IN; 5The University of Durham, Durham, United Kingdom. (Tracking ID #1938546) BACKGROUND: Previous work suggests that prescribing errors account for the highest proportion of medication errors (Lisby, 2010) and that one in every four reported in the U.S. can be linked to drug name confusion (Lambert, 1999). Computerized provider order entry (CPOE) has been widely proposed, implemented, and shown to reduce errors associated with traditional, handwritten, medication prescribing. However, CPOE has not fully lived up to its predicted benefits and many attribute these shortfalls to suboptimal design of systems that are neither interoperable nor easy to use (Kellerman, 2013, Health Affairs). Given recent calls for CPOE performance testing post-implementation (Thompson, 2010 & Bates, 2013) and the proportion of medication errors which occur during the prescribing phase, we undertook a study to assess drug name design and display issues in CPOE including: look-alike-sound-alike pairs, similar name adjacency errors, font, and visual display to better understand their potential to contribute to medication errors. METHODS: As part of an FDA Task Order, this qualitative study assessed 10 leading vendor and home-grown CPOE systems in six major U.S. academic medical centers via remote walk-throughs and on-site visits. To standardize assessment, the team developed and iteratively refined a CPOE Assessment Tool (CAT) relating to 18 elements defined by the FDA to systematically test the potential for drug name/ identification errors across various stages of the medication ordering process. This tool incorporated specific medication ordering, review, and deletion scenarios used to guide system assessments with a series of regular and expert users of these systems. Outputs from these evaluations were recorded from each system using screen shots, video and audio transcripts of user and designer comments. Recordings were transcribed, analyzed, and used to inform the content of onsite visits to each medical center to obtain additional system information from medical leadership, IT experts, users, and pharmacists. Inter- and intra-system characteristics were analyzed descriptively for patterns, themes, and variations. RESULTS: We identified a variety of ways drug names are displayed in CPOE systems that could potentially contribute to medication errors. Recurrent issues and themes that emerged included: (a) Brand versus generic name: Presentation of brand and generic names varied both across and within systems depending on task functionality. In nine of the ten CPOEs assessed, for example, one brand name search (Plavix) yielded results displaying the drug in eight different ways. Some systems used all caps, some only initial caps, some used both. Some were inconsistent in the drug name searched and drug name returned. One system displayed two columns which listed both the brand and generic names, others included only one, and others presented one or the other parenthetically, based on the search term. One system allowed ordering by generic name only. If one searched by brand name, the medication did not appear; we observed end user frustrations resulting from inability to search for and find desired drugs searched non-generically. Systems also used varied methods to display brand and generic options to steer prescribing towards preferred products (typically the generic). (b) Combination products: In most systems, searching for a combination drug often yielded the brand name with the ingredients listed parenthetically. We rarely saw systems provide explanations of ingredient strengths. In some systems, searching the brand name yielded only the brand name, without its generic ingredients. (c) Drug-name modifiers: Across systems, drug modifiers seemed to have some consistency in both representation and display. Drug modifiers appeared primarily at the end of the drug name, in all caps, and were abbreviated (e.g. XR, DS, etc.). They were rarely truncated. (d) Extraneous additions to drug name fields: Across systems, we observed several instances of nonsystematically populating (jerry rigging) name fields with additional information ranging from drug indication to which facility the drug could be ordered from. (e) Font, text size, capitalization: These all varied widely with no standardized approaches seen both across and often within systems. CONCLUSIONS: There is little consistency in the ways that leading CPOE systems search for and present drug names. This lack of standardization was seen both across systems and often within a single system or institutions CPOE systems. Brand vs. generic, combination products and drug name modifiers were particularly inconsistent, each presenting the potential for user frustration, confusion, and erroneous orders. A QUALITATIVE STUDY EXPLORING THE VULNERABILITIES OF COMPUTERIZED PHYSICIAN ORDER ENTRY SYSTEMS Sarah P. Slight1,2; Tewodros Eguale1,3; Mary Amato1,4; Andrew C. Seger4; Diana L. Whitney5; David W. Bates1,6; Gordon D. Schiff1,6. 1Brigham and Womens Hospital, Boston, MA; 2Durham University, Stockton on Tees, United Kingdom; 3McGill University, Montreal, QC, Canada; 4MCPHS University, Boston, MA; 5Baylor College of Medicine, Houston, TX; 6Harvard Medical School, Boston, MA. (Tracking ID #1935926) BACKGROUND: Computerized Physician Order Entry (CPOE) systems can prevent medication errors in both inpatient and outpatient settings. Depending on how they are designed, however, they can fail to optimally prevent various prescribing errors or introduce new errors. The Institute of Medicine report Health IT and Patient Safety: Building Safer Systems for Better Care recommended that specific examples of potentially unsafe processes and risk-enhancing interfaces be identified and shared amongst the health IT community. This study aims to test the vulnerabilities of a wide range of CPOE systems to different types of medication errors, and to develop a more comprehensive understanding of how CPOE human factors design could be improved. METHODS: As part of a National Patient Safety Foundation-funded project, we examined a range of leading vendor and home grown CPOE systems (e.g., Cerner; Epic; Medi-tech; LMR; BICS; GE Centricity) in diverse organizations in United States and Canada. Typical users at each of 16 sites were asked to enter 13 different orders on test patients based on scenarios of previously reported CPOE errors. Users were encouraged to use both usual practice and, where-needed, workarounds to enter the erroneous orders, as well as reflect on their overall knowledge and experience of using their system. A research pharmacist and research assistant independently observed test users enter each order and rated the ease or difficulty of these entries using standardized operational definitions. RESULTS: Ease of entry of erroneous orders and the generation of alert warnings in different CPOE systems was highly variable and appeared to depend on a number of factors including how the order information was entered (i.e., in a structured or unstructured way); whether a specific alert functionality (e.g., duplicate-drug checking) was operational in the system; and which drugs or drug combinations were included in the clinical decision support algorithms. Test users found the wording of many of the alert warnings confusing and expressed frustrations with the way irrelevant warnings appeared on the same screen as those more relevant to the current order. The timing of alert warnings also differed across CPOE systems, with dangerous drug-drug interaction warnings displayed, for example, only after both Imdur (isosorbide mononitrate) and Revatio (sildenafil) had been entered and the order signed off in two CPOE systems. Alert warnings also varied in their level of severity in different systems even within the same institution (e.g., test user was presented with a hard stop alert warning in the inpatient system after entering Lovenox (enoxaparin sodium injection) 40 mg subcutaneous daily followed by a second order for Lovenox 100 mg subcutaneous twice daily, whereas the equivalent outpatient system produced an alert for the same order that was easily overridden). Test users commented on how there may be situations where they needed to prescribe the same drug twice, requiring workarounds such as entering the brand name of the drug for the morning dose and the generic name for the evening dose, to avoid getting duplicate drug alert warnings. CONCLUSIONS: We found a high degree of variability of vulnerability in ordering and alerting between different CPOE systems. Detailed qualitative analysis of both the observed CPOE functionality and users think-aloud comments provided rich insights into the ways both systems and users were susceptible to ordering errors. System developers and users need to be aware of these vulnerabilities and build in protection strategies to re-engineer their systems to ensure safer prescribing. A RANDOMIZED TRIAL OF A COMMUNITY HEALTH WORKER LED INTERVENTION TO IMPROVE DIABETES INTERMEDIATE OUTCOMES AMONG LATINOS PATIENTS WITH POORLY CONTROLLED DIABETES Olveen Carrasquillo; Yisel Alonzo; Cynthia Lebron; Natalie Ferras; Ernesto ReyesArrechea; Hua Li; Sonjia Kenya. University of Miami, Miami, FL. (Tracking ID #1936786) BACKGROUND: Diabetes disproportionately impacts Latinos. Community Health Workers (CHWSs) are one approach that may improve health outcomes in this population. However, evidence from rigorous randomized studies is limited. METHODS: The Miami Health Heart Initiative examined the impact of a comprehensive community health worker (CHW) intervention on diabetes intermediate outcomes of systolic blood pressure (SBP), lipids (LDL) and Hemoglobin A1C (HbA1c). We recruited 300 Latino patients with poorly controlled diabetes (HbA1c>=8.0 %) ages 3565 from the primary care clinics of Miami-Dade countys public hospital. Subjects were randomized to usual health care or a comprehensive structured CHW 1 year intervention consisting of home visits, phone calls and group education sessions. The intervention included patient navigation as well as assistance with social and non-medical needs. An RA blinded to group assignment conducted initial and follow-up evaluations at 12 months. We used linear mixed models to statistically test for the impact of our intervention on outcomes using intention to treat analyses. RESULTS: The mean age of our patients was 56 + 7 years, 55 % were female and mean BMI was 31.6 + 7.4 kg/m2. Over half (55 %) were sedentary (IPAQ) and median daily fruit and vegetable was two (BRFSS). Cubans made up 29 % of our sample with no other Hispanic ethnic subgroup representing over 15 % of the sample; 47 % of respondents scored the lowest possible acculturation score (Marin-Marin). Intervention patients received a median of five home CHW visits and 25 phone calls; 84 % of intervention group participants had at least 12 CHW contacts over the course of the year. Participation at group session was more skewed with 52 % of participants not attending any sessions. Of those that attended at least one session, average attendance was four sessions per participant. At 1 year the retention rate was 72 % and similar in both arms. Data on outcomes by group is shown in Table 1. CONCLUSIONS: In a heterogeneous Latino population with poorly controlled diabetes, a rigorous CHW intervention resulted in statistically and clinically meaningful changes in HbAIC but not in blood pressure or lipids. The impact of the intervention on the latter two outcomes may have been limited by the fact nearly half of patients were at the SBP and LDL targets prior to the intervention. *p = 0.01 for HbA1C NS for SBP and LDL A REDUCTION IN TELEMETRY ORDER DURATION DOES NOT CHANGE TELEMETRY ORDERING FREQUENCY AND REDUCES UTILI ZATION Joel Boggan1,2; Ann Marie Navar-Boggan4; Ryan D. Schulteis3,2; David Simel6,5. 1Durham Veterans Affairs Medical Center, Durham, NC; 2Duke University Health System, Durham, NC; 3Durham Veterans Affairs Medical Center, Durham, NC; 4Duke University Health System, Durham, NC; 5Duke University Health System, Durham, NC; 6Durham Veterans Affairs Medical Center, Durham, NC. (Tracking ID #1940176) BACKGROUND: American Heart Association guidelines for appropriate use of telemetry have not decreased inappropriate use of in-hospital cardiac monitoring. We evaluated how changing the default order set for telemetry affected overall telemetry utilization. METHODS: The computerized ordering system at the Durham VA Medical Center was changed so that the maximum number of hours of cardiac monitoring was reduced by 33 %, from 72 to 48 h, before requiring a renewal order. No other intervention or general education about telemetry appropriateness was done. We compared the frequency of orders/week for telemetry and duration of cardiac monitoring for 12 weeks before through 6 weeks after this intervention. To assess the safety of the intervention, we measured the number of rapid response team (RRT) or code events occurring at any time during the hospital stay of general medical or general surgical patients. We compared event rates on telemetry before and after the intervention. RESULTS: Over a 4-month period, there were 1051 telemetry orders placed for general medical or surgical patients before the intervention and 450 after. The average number of orders/week did not differ significantly before vs after the reduction in telemetry order duration (88 +/ 10 vs. 75 +/ 22 orders/week, p = 0.11). The mean duration of telemetry was reduced by 32 %, which was the same percent reduction required by the change in policy (mean reduction in telemetry time decreased from 64 +/ 3.1 h/patient to 44 +/ 1.2 h/patient, p < 0.01). The rate of code events and rapid response team events in hospitalized general medical or surgical patients decreased from 2.5 events/week to 2.2 events/week. Only seven RRT/code events occurred in patients on telemetry prior to the intervention, while one RRT/code event occurred in a patient on telemetry after the intervention (p = 0.28). CONCLUSIONS: A simple change in the default maximum number of hours a provider can prescribe cardiac monitoring before requiring a renewal order led to a dramatic decrease in telemetry utilization. Although the number of orders for patients to receive cardiac monitoring was unchanged, the duration of cardiac monitoring dropped substantially. Even with less monitoring, we found no increase in the rate of adverse events. Similar reductions at other medical centers may reduce telemetry utilization, technologist and nurse staffing requirements, and cost while preserving patient safety. A RESIDENT INITIATIVE TO INCREASE USE OF NALTREXONE FOR ALCOHOL USE DISORDER AT A SAFETY NET HOSPITAL Maria E. Otto1; Jeffrey K. Hom1; Shilpa M. Shah1; Natalie C. Young1; Weston S. Fisher2; Edgar Pierluissi1. 1University of California San Francisco, San Francisco, CA; 2University of California San Francisco, San Francisco, CA. (Tracking ID #1939058) BACKGROUND: Alcohol use disorder (AUD) is extremely common at San Francisco General Hospital (SFGH), with nearly one in four inpatients on the Medicine service affected. These patients frequently present to the Emergency Department (ED) with alcohol-related falls, seizures, and symptoms of withdrawal. Their 30-day readmission rate is 16 %, above the 13 % for the remainder of the hospital. Hospitalization provides a window of opportunity to help this vulnerable population, as patients are often motivated to make changes following an alcohol-related illness but are frequently without primary care doctors. Despite the effectiveness of naltrexone, an inexpensive, FDA-approved medication for treatment of AUD, it has been underutilized. A pilot intervention in 2011 at SFGH increased prescriptions of naltrexone to eligible patients at discharge, but was not sustained. METHODS: A resident-led, quality improvement (QI) project was implemented in February 2013 to 1) sustainably increase the prescribing of naltrexone to eligible inpatients with AUD upon discharge and 2) to measure the effect of the intervention on ED visits and readmissions within 30 days. The intervention 1) identified inpatients with AUD via the electronic handoff system, 2) educated providers about naltrexone, 3) prompted providers to prescribe naltrexone to eligible patients at discharge planning rounds, and 4) utilized principles of QI to continuously evaluate and improve the intervention. Data were collected via retrospective chart review to identify patients with AUD and determine eligibility for naltrexone. Among patients with AUD, number of ED visits and readmissions within 30-days of discharge were also obtained. RESULTS: Among 118 patients with AUD identified between March 1 and April 15, 2013, 43 % (n = 51) were eligible for naltrexone. Following the initial intervention, naltrexone offering increased from 2 to 58 % of eligible patients. This effect was sustained through continuous cycles of improvement, with 67 % of patients offered it most recently. Compared to patients not prescribed naltrexone (n = 31), patients prescribed the medication (n = 20) had 30-day readmission rates and hospital visits that were 74 % and 88 % lower, respectively. CONCLUSIONS: These results are promising and suggest that an intervention to systematically identify and offer treatment to patients with AUD is important in caring for our patient population and decreasing hospital utilization. Additionally, this QI initiative demonstrates that a deep-rooted problem can be addressed with the commitment of an interdisciplinary team, which can include a rotating group of trainees. By assessing the effect of this intervention on hospital utilization, the project shows that measuring true outcomes, rather than just processes, can help build a business case for sustainability. Next steps include 1) expanding the intervention to other hospital services, 2) ensuring the safe transition of patients on naltrexone from the hospital setting to their outpatient providers, and 3) collaborating with hospital leadership to improve systems to identify and offer treatment to all patients with AUD. A ROLLING, INTEGRATED EVIDENCED-BASED MEDICINE CURRICULUMS IMPACT ON EVIDENCE-BASED PRACTICE SKILLS, KNOWLEDGE, BEHAVIOR AND ATTITUDES ON INTERNAL MEDICINE STUDENTS AND RESIDENTS AT THE WASHINGTON, DC VETERANS AFFAIRS MEDICAL CENTER Matthew Tuck; Samantha Mcintosh; David Maron. Department of Veterans Affairs Medical Center, Washington, DC. (Tracking ID #1941446) BACKGROUND: Teaching evidence-based medicine (EBM) skills to students and residents has become an imperative for medical education. The AAMC made recommendations to incorporate problem-based clinical skills in medical schools curricula. Similarly, the Next Accreditation System of the ACGME requires that residency programs train and evaluate residents in practice-based learning and improvement. Developing effective techniques for integrating these practices across the years of medical education remains a challenge for medical schools and residency programs. METHODS: We developed a rolling EBM curriculum for medical students and residents rotating through the Medical Service of the Washington, DC Veterans Affairs Medical Center (VAMC) that provides repeated opportunities for students and residents to learn, practice, and evaluate their skills within the context of real-time patient care. The curriculum is integrated into morning report over the course of a month-long rotation. Using a patient case on the wards, learners are guided through the process of formulating a PICO question, searching the literature to find an article that answers the question, appraising the validity of the article and applying the literature to the case. The aim of the study was to evaluate the curriculum on learners self-reported patient care practices, attitudes and knowledge using validated rubrics. We hypothesized that exposure to the curriculum would improve learners EBM-related patient care practices, attitudes and knowledge. Using a before-and-after design, consented learners were given questionnaires. We used validated rubrics on attitudes and behavior as well as a rubric to assess EBM knowledge (the so-called Berlin Questionnaire) at the beginning and end of their rotation at the VAMC. Incentives were provided for completion of the questionnaires. Means, standard deviations, percentages and univariate paired t-tests were calculated for pre- and post-exposure scores. The research was funded by a Medical Education Research Grant awarded by George Washington University. RESULTS: 231 residents and students were exposed to the curriculum. Sixtythree learners completed the pre-curriculum questionnaire and 21 completed both the pre- and post-curriculum questionnaires. Up to 74 % of respondents had never received formal education on EBM prior to their rotation at the VAMC. Mean scores on attitudinal and belief questions improved after exposure to the curriculum, but to a non-significant level. Similarly, knowledge of EBM, as assessed by the Berlin Questionnaire, improved to a non-significant level (pre-curriculum mean score 4.5, post-curriculum mean score 5.3, p = 0.10). A regression was conducted to determine if the number of days between the pre and post-curriculum assessment predicted improved performance on the Berlin Questionnaire and was non-significant (p = 0.13). CONCLUSIONS: We successfully implemented a rolling, integrated EBM curriculum for internal medicine students and residents. Attitudinal, behavioral and knowledge scores improved to a non-significant level after the month-long exposure to the curriculum. The lack of significant findings may be in part due to small sample size. A SALARY EQUITY STUDY: DIFFERENCES BY GENDER AND RACE, DEPARTMENT OF MEDICINE, UNIVERSITY OF MINNESOTA Anne Joseph; Wesley Miller; Jeremiah Menk; Stewart McMullan. University of Minnesota, Minneapolis, MN. (Tracking ID #1938954) BACKGROUND: Salary inequities for female faculty have been documented in many academic disciplines. METHODS: The University of Minnesota Department of Medicine (DOM) conducted cross-sectional studies in 2011, 2012, and 2013 with the objective of examining gender equity issues related to salary (n = 165, n = 196, and n = 213 respectively). A pre-specified multivariable linear regression was used to adjust the salary comparison between men and women and Whites and nonWhites. The analysis included the following variables: gender, race, academic rank (and time in rank), leadership roles, degree(s), number of ABIM certifications, time from highest degree to current position, percent clinical effort, and subspecialty. RESULTS: In 2011, 8 % of women were full Professors compared to 35 % of men. Leadership roles were held by 7 % of women compared to 19 % of men; similar differences in career characteristics were observed across all 3 years. The unadjusted mean salary for women was $69,717$76,009 lower compared to men depending on the year. In 2011, 2012, and 2013 the adjusted mean comparisons for women vs. men were $-10,513, $9,377, and $16,132 respectively. For non-Whites vs. Whites the comparisons were $9,018, $741, and $13,002. R-squared values for the models were between 0.757 and 0.767. As predicted, the model confirmed that the following were associated with increased salary: a PhD or Masters in addition to MD degree, number of ABIM certifications, rank, time variables, and holding a leadership position. For clinicians, procedural subspecialties (cardiology, gastroenterology, pulmonary, renal) conferred higher salaries than hematology-oncology-transplant, endocrinology, rheumatology, infectious disease, general internal medicine (in that order). CONCLUSIONS: Gender differences persisted over the 3-year period despite preferential salary adjustments for women. Professional accomplishment characteristics do not fully explain faculty salary differences by gender and race. In response, the DOM has formed a Salary Equity Committee to systematically address these differences abiding by principles of transparency, simplicity, value of clinical, educational and research missions, validated benchmarks of productivity, equal access to leadership positions and supplemental income, and sustainability. Adjusted Mean Comparisons A SIMPLE SCREENING METHOD TO DETECT MALIGNANCY USING RED CELL DISTRIBUTION WIDTH (RDW) Hiroaki Wakakuri; Toshihiko Ohara; Makoto Suzaki; Sonoko Kirinoki; Naoko Onodera; Takashi Araki; Hideya Hyodo; Makoto Kawai; Masahiro Yasutake. Nippon medical school, Bunkyokusendagi, Japan. (Tracking ID #1928358) BACKGROUND: RDW is rapidly calculated parameter using an automatic cell counter. Recent studies showed that RDW is strongly associated with prognosis in disorders such as chronic heart failure, idiopathic pulmonary fibrosis, sepsis, various cancer, and chronic kidney disease. However, it is still unclear that the relationship of RDW and diagnosis of malignancy. The aim of this study is to evaluate the usefulness of RDW to screening malignancy at the first visit outpatient clinic. METHODS: Fifty one consecutive patients diagnosed as malignancy at ER and general internal medicine of our hospital were reviewed in the period between September 2012 and September 2013. Fifty one patients, matching age and sex, with common diseases, including hypertension and dyslipidemia were assigned to control group. The patients complicated severe anemia (Hb < 9.0), chronic heart failure, chronic inflammation disease, with a dosing history of medication that have effect on RDW were excluded from control group. RDW was calculated using an automatic cell counter (Sysmex XE-5000). We also measured Hb, BUN, Cre, TC, albumin and CRP at the same time. The receiver operating characteristic (ROC) curve showed the best cut-off value of RDW for predicting malignancy. RESULTS: Malignancy group (age 67.8 [2488], man 27, woman 24) was classified into the digestive organs n 26 (gastric ca, colon ca, cecal ca, vater ca, HCC, GIST), hemopathy n 10 (malignant lymphoma, multiple myeloma, AML), respiratory organs n 6 (lung ca), gynecologic disease n 3 (ovarian ca, uterine ca), urinary organs n 3 (prostate ca, renal pelvic ca), breast cancer n 2, and liposarcoma n 1. Control group (age67.4 [2886]) included 27 men and 24 women. Malignancy group had lower Hb and alb values than control group, and higher CRP values than control group (Hb: 12.0 [5.017.3] vs 13.6 [10.516.9], albumin: 3.64 [2.64.9] vs 4.21 [3.24.8], CRP: 3.33 [0.0321.3] vs 0.37 [0.012.88], P < 0.05). There was no strong correlation between the RDW and other parameters (vs BUN r 0.096, vs Cre r 0.071, vs TC r 0.284, vs CRP r 0.182, vs albumin r 0.437, vs Hb r= 0.537). RDW of malignancy group (14.6 [12.625]) was significantly higher than RDW of control group (13.3 [12.215.6]), (P< 0.05). Using ROC analysis (AUC = 0.744), RDW of 13.9 % was the best cut-off value for predicting malignancy with a sensitivity of 58.9 % and a specificity of 78.5 %. CONCLUSIONS: Our result shows that RDW of malignancy group is significantly higher than control group. It is not enough to discriminate patients with malignancy diseases between healthy persons by just RDW value, but an elevated RDW > 13.9 % value is a reliable additional predictor for screening malignancy at the first visit outpatient clinic. A STRUCTURED REVIEW OF BENEFIT AND HARM DATA IN CANCER SCREENING GUIDELINES AND POINT-OF-CARE CLINICAL RE SOURCES: DO CLINICIANS HAVE THE INFORMATION THEY NEED? Tanner Caverly; Elyse Reamer; Rodney A. Hayward; Michele Heisler; Angela Fagerlin. University of Michigan Medical School, Ann Arbor, MI. (Tracking ID #1939040) BACKGROUND: Data on the absolute benefits and harms of preventive interventions represent essential information for clinicians. Clinicians cannot practice informed or shared decision making if they lack ready access to this information, which is needed to help weigh trade-offs between potential benefits and harms. We might expect that transparent quantification of potential benefits and harms would be provided in prevention guidelines and point-of-care resources (such as UpToDate), but to date, a systematic assessment of whether these clinical resources provide this important clinical information has not been conducted. We sought to address this gap by conducting a review of current cancer prevention guidelines and point-of-service resources to examine what information they provide on the chance of benefit and the potential for harms. METHODS: We reviewed all recommendation statements related to cancer prevention and screening from the United States Preventative Services Task Force (USPSTF), excluding statements rated I for insufficient evidence, pediatric and adolescent guidelines, and those only discussing counseling interventions. To capture other non-USPSTF guidelines in the same topic areas, we examined within the National Guidelines Clearing House (NGC) all discrete guidelines developed by US organizations. We also searched two popular point-of-care clinical resources (UpToDate and Dynamed). A standardized abstraction form was iteratively created. We abstracted basic guideline demographic data and information about how benefits and harms are presented. RESULTS: We reviewed 11 cancer prevention guidelines from the USPSTF, 19 guidelines from other organizations within the same cancer-prevention areas, 15 related articles from Dynamed and 14 from Up-to-Date. These resources contained 114 separate recommendations, 77 of which described recommendations for the use of the preventive service in at least some sub-populations (positive recommendations). Among the positive recommendations, 31 % (24 out of 77) simply made a recommendation without mentioning any estimate of the chance of benefit. Twentyseven percent (21 out of 77) presented the benefit as an absolute risk reduction. Regarding the potential for harm inherent in any intervention, 48 % (37 out of 77) of the recommendation statements did not mention any harms and an additional 18 % (14 out of 77) mentioned but did not quantify those harms. Of the 27 % (21 out of 77) that quantified both potential benefits and potential harms, 19 % (4 out of 21) presented benefit as a relative risk reduction (larger number) while presenting harm as an absolute risk increase (smaller number)a format known as mismatched framing that exaggerates the relative magnitude of the benefit compared with the harm. Only 21 % (n = 16; 10 from guidelines and 6 from point-of-care resources) of the 77 positive recommendations quantified both the potential benefits and harms as absolute risks so that they could be directly compared. CONCLUSIONS: Guidelines and point-of-care clinical resources for cancer prevention interventions frequently do not provide needed information on the chance of the intervention resulting in patient benefit or the potential of being harmed. The resources we reviewed may not represent all of the resources that clinicians use in daily practice. However, other resources are unlikely to present this information if the resources we reviewed did not. After undertaking the effort to summarize and rate the quality of evidence supporting use of a preventive service, evidence review bodies should also seek to quantify and present the potential benefits and harms of the proposed intervention. A satisfactory way to present this data would be to depict estimates of absolute benefits and harms together in a single table within the guidelineas is promoted by the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach. BACKGROUND: Cervical cancer disproportionately impacts minority women. While screening substantially reduces cervical cancer mortality, many women in safety-net settings remain unscreened. System level barriers include limited resources and staffing. Patient barriers include financial (lack of insurance and co-payments), health literacy, and cultural norms. Provider barriers such as visit time constraints also exist. Human Papilloma Virus (HPV) self-sampling, in which women use a cotton swab device to self-collect material for an HPV assay, may be an attractive option in such settings. METHODS: 1) To measure the acceptability and feasibility of HPV self-sampling among patients, we had two Community Health Workers (CHWs) recruit women from two clinics in Miami-Dade County. One site was a public hospital based clinic primarily serving Hispanics; the other a community based facility serving Haitians. CHWs identified women aged 3065 years with no Pap smear in the past 3 years and gave a brief description of the study. We offered all eligible women the option of HPV self-sampling or discussing screening through Pap smears with their provider. If HPV self-sampling was selected, instructions on sampler use and assistance with collection was provided by the CHW. Patient attitudes towards HPV self-sampling were assessed using a 12-item survey modeled after existing surveys. Among women in whom the HPV self-sampler option was not chosen, the electronic medical record or paper chart were reviewed to assess if a Pap smear was performed. 2) To measure acceptability and feasibility among clinic staff (medical assistant, nurse, nurse practitioner, and physician), we developed a 9-item survey based on prior studies of provider screening acceptability. Through emails and personal invitations, staff were encouraged to participate in an anonymous online questionnaire. RESULTS: Of 1725 women assessed, 62 % were not eligible due to having a Pap smear in the last 3 years. Of the 184 study eligible women, 91 % agreed to participate (n = 167). Latinas represented 73 % and Haitians 27 % of the sample; 99 % were uninsured. The HPV self-sampling approach was selected by 67 % of participants. The majority of participants agreed that HPV self-sampling was faster than the Pap smear (86 %), more private (93 %), and easy to use (97 %). Nearly all participants agreed they would use the self-sampler again (97 %) and recommend it to a friend or family member (96 %). Overall, 76 % of participants preferred the self-sampler to the Pap smear. High-risk HPV was detected in 11 % of samples. The CHW provided navigation to Pap smear for all women having high-risk HPV strains. Of those not selecting the selfsampler, 34 % had not yet had a Pap smear at 1 month post-recruitment. Of 178 clinic staff invited to participate in our survey, 16 % (n = 29) completed the survey. Of those completing the survey, 69 % agreed it was difficult to perform a Pap smear because of time constraints and 79 % agreed they would be willing to incorporate HPV self-sampling into their practice. CONCLUSIONS: HPV self-sampling was both feasible and highly acceptable in the safety-net clinics studied. A low response rate by clinic staff may limit the generalizability of the provider survey results. HPV self-sampling may be an important tool to address barriers to cervical cancer screening in resource limited settings. ACCESS TO AND PREFERENCES FOR TEXT MESSAGING FOR MEDICAL AND INSURANCE REMINDERS IN A SAFETY NET POPULATION Leah Zallman1,2; Alex Harsha1,3; Danny McCormick1,3. 1Cambridge Health Alliance, Cambridge, MA; 2Institute for Community Health, Cambridge, MA; 3Harvard Medical School, Boston, MA. (Tracking ID #1934972) BACKGROUND: Text messaging has been used successfully to deliver health information interventions to underserved communities. Evidence suggests that text message reminders can increase patient appointment and treatment completion, and is a reliable method of communication even in communities with low literacy. Text messaging may offer similar opportunities to reach patients in health care safety net settings who are disproportionately impacted by homelessness or unstable housing, low health literacy and may be lost to follow up. The immediate, direct-to-patient nature of text messaging and the ability to send automated reminders at relatively low cost may help overcome communication barriers between healthcare providers and underserved communities in the US. While the opportunity to improve this communication may exist through text messaging, preferences for text messaging and patient characteristics associated with preferring text messages have not been elucidated METHODS: Between August and December 2013, we conducted a face-to-face verbally delivered survey regarding preferences for text messaging with 415 patients presenting to three emergency departments at a large integrated safety net health care system in the greater Boston area. We included patients aged 1864, who had an Emergency Severity Index (ESI) of 25 (excluded most severely ill), had only one insurance type and spoke English, Portuguese, Spanish or Haitian Creole. We asked participants about their access to mobile phones with text messaging capability, preferences for receiving specific types of reminders by text (appointment, medication, and expiring insurance), and preferences for text messaging over other forms of communication. We used logistic regression models that controlled for age, gender, race/ethnicity, health insurance type and medical comorbidities to identify predictors of preferring text messages for medical reminders over other forms of communication with medical providers. RESULTS: Overall, 96.5 % of participants reported having access to mobile phones with text messaging capability. Half reported a preference for text messaging over other forms of communication such as phone, mail or email. Respondents expressed preferences for appointment reminders (77 %), reminders regarding expiring insurance (55 %), and medication reminders (32 %). A minority (17 %) indicated that they preferred not to be text messaged by their healthcare provider. Reasons for not preferring text messaging included: cost of receiving text messages (11 %), privacy (17 %), not liking or understanding text messaging (40 %) and other (31 %). In logistic regression analyses, Hispanic ethnicity (1.68 [1.032.75]) and younger age (0.98 [0.961.00]) were associated with preferences for medication reminders. Similarly, Hispanic ethnicity (1.93 [1.093.43]) and younger age (0.98 [0.951.00)) were associated with preferences for expiring insurance reminders. CONCLUSIONS: In a safety net population, the vast majority of patients have access to text messaging and a high proportion express preferences for text message reminders over other forms of communication. Reminders regarding appointments and expiring insurance were most highly desired. Text messaging may be a promising method to improve healthcare provider communication with underserved communities in the US. ACCESS TO DEPRESSION CARE SERVICES: A NEEDS ASSESSMENT IN AN URBAN ACADEMIC PRIMARY CARE CLINIC Lauren Peccoralo; Susan Truong. Mount Sinai Medical Center, New York, NY. (Tracking ID #1937528) BACKGROUND: Major depressive and anxiety disorders are common and these and other psychiatric conditions are often encountered by primary care practitioners. Urban academic training centers serve patients at high risk for mental health disorders who often have little access to care, yet little is known about the perceived need for mental health services amongst primary care providers in these environments. The US Preventive Services Task Force endorses routine screening of all adults for depression in settings with adequate staff-assisted depression care support to provide appropriate care and follow up. The purpose of this survey was to assess the needs and barriers to depression care services amongst physician primary care providers in an urban academic training practice in the United States prior to the implementation of universal screening and a depression care program. METHODS: From April through June 2013, we conducted an online survey about the perceptions of barriers to depression care and unmet psychiatric needs in the primary care practice in our clinic. Survey participants included primary care providers (all internal medicine residents and faculty providers) at the resident training clinic practice at the Mount Sinai School of Medicine in New York, New York. The survey was analyzed as descriptive data. RESULTS: Ninety-two participants responded to the survey (74 resident trainees and 18 attending physicians, 62 % and 78 % response rate, respectively) during the study period. Eighty-three percent of residents and 88 % of attendings are currently caring for a patient(s) with depression. Eighty-six percent of residents and 100 % of attendings believe there are barriers to providing optimal care for patients with depression. The most commonly identified barriers to care are lack of psychiatrists (85 % residents, 94 % attendings), lack of coordination between primary care and psychiatry (83 % residents, 100 % attendings). and lack of patient education (54 % residents, 78 % attendings). Ninety-six percent of residents and 94 % of attendings have at least one patient with an unmet psychiatric need and 18 % of residents and 35 % of attendings have >10 patients with such a need. The most commonly cited psychiatric needs are depression (93 % residents, 100 % attendings), anxiety (74 % residents, 94 % attendings) and substance abuse (45 % residents, 47 % attendings). CONCLUSIONS: Primary care providers at all levels perceive high need for and multiple barriers to depression and other mental health care. These data support the recommendation that appropriate mental health services should be available in tandem with the initiation of universal depression screening. Further research is needed to determine the impact of implementation of universal screening and depression care program in a primary care training environment. ACCULTURATION AND MEASURES OF CARDIOVASCULAR RISK AMONG FILIPINA IMMIGRANTS Jane Jih1; Alka M. Kanaya1; Maria Rosario G. Araneta2. 1UCSF, San Francisco, CA; 2UCSD, San Diego, CA. (Tracking ID #1938915) BACKGROUND: Acculturation is a dynamic process that encompasses a broad spectrum of cultural changes, interactions and adaptations that affect health. In immigrant populations, acculturation has been associated with increased prevalence of obesity, diabetes and coronary heart disease (CHD). CHD and its risk factors are highly prevalent among Filipino Americans. However, limited studies have examined acculturation and CHD specifically among Filipino Americans. We used the University of California San Diego (UCSD) Filipino Womens Health Study to examine the relationship of acculturation and CHD and related risk factors among Filipina immigrants. METHODS: The UCSD Filipino Womens Health Study is a longitudinal study of 453 Filipino American women ages 3986 years recruited from 1995 to 1999 in which the prevalence of and risk factors for CHD, hypertension (HTN), diabetes and obesity was measured. In addition to collecting immigration data, an adapted acculturation scale with three domains [a) functional integration via language use; b) values on preserving culture; c) attitudes on family structure and sex-gender roles] was administered. Outcome variables included prevalence of CHD (defined by ECG, Rose questionnaire and history of myocardial infarction), HTN, diabetes and obesity. We examined several measures of acculturation including length of residence in the United States (US), proportion of life years in the US (years residing in US/age at enrollment) and each domain of the acculturation scale. We used logistic regression, adjusting for age, education and exercise, smoking, alcohol use and body mass index (BMI) for the CHD outcome, to estimate the association of each measure of acculturation to each outcome. RESULTS: The mean age of Filipina immigrants (n = 446) was 58 9 years old. The mean length of US residence was 16 years (range 048). The prevalence of CHD was 21 %, HTN was 51 %, and type two diabetes was 33 %. Mean BMI was 25 3 kg/m2 and 23 % were obese as defined by the World Health Organization (WHO) Asian BMI scale (BMI 27.5 kg/m2). Measures of acculturation were not associated with CHD, HTN and diabetes outcomes. Length of residence in the US but not proportion of life years or the acculturation scale was associated with obesity defined by the WHO Asian BMI scale. For each additional year of living in the US, Filipina immigrants had a 21 % increase in the odds of obesity (adjusted odds ratio 1.21 (95 % CI 1.011.47)). CONCLUSIONS: Filipina immigrants have a substantial burden of CHD and risk factors. Unlike other immigrant populations, acculturation to life in the US was not associated with CHD, HTN and diabetes prevalence in Filipino women. However, longer duration of residence in the US was associated with increased odds of obesity. Exposure to a westernized lifestyle during the Spanish and US colonization in the Philippines may mitigate the impact of migration on cardiovascular health among Filipina immigrants. ACCURACY AND GAMING OF THE ACUTE BRONCHITIS ANTIBIOTIC PRESCRIBING PERFORMANCE MEASURE Patrick P. Dempsey1; Harry Reyes Nieva1; Alexandra C. Businger1; Jeffrey A. Linder1,2. 1Brigham and Womens Hospital, Boston, MA; 2Harvard Medical School, Boston, MA. (Tracking ID #1938252) BACKGROUND: Acute bronchitis is a cough-predominant illness of less than 3 weeks duration in a patient without chronic lung disease. A national performance measure states that antibiotics are not indicated for acute bronchitis. However, concerns have been raised about the measures accuracy and susceptibility to gaming. We compared the performance measure to chart review to assess the accuracy of the performance measure and detect gaming. METHODS: We assessed 57,766 visits to an academic primary care practice between May 2011 and September 2012. For the chart review, we included patients aged 1864 and excluded patients with chronic lung disease on the problem list or who made a study clinic visit in the prior 30 days (n = 26,539). Two trained chart abstractors reviewed these notes and coded whether the treating clinician diagnosed the patient with acute bronchitis. We calculated a modified Healthcare Effectiveness Data and Information Set (HEDIS) performance measure by including patients aged 1864 with ICD-9 diagnosis codes of 466 or 490, excluding patients with chronic lung disease, immunosuppression, and cancer. Using chart review as the gold standard, we calculated the sensitivity, specificity, and positive predictive value of the acute bronchitis performance measure. RESULTS: There were 31,174 visits which neither the performance measure nor chart review identified as an acute bronchitis visit. According to the performance measure, there were 90 acute bronchitis visits (88 with a diagnosis code of 466 and 2 with a diagnosis code of 490). There were 143 visits identified as acute bronchitis visits by both the performance measure and chart review (n = 41; true positives), the performance measure only (n = 49; false positives), or chart review only (n = 53; false negatives). These patients had a mean age of 48 years old; 77 % women; and 44 % white, 28 % Black, and 17 % Latino. Compared to chart review, the performance measure had a positive predictive value of 46 %, a sensitivity of 44 % and a specificity of >99 %. Of the 49 false positives, according to the chart, 10 visits were for other ARIs, seven visits were to other practices, seven the patient had been coughing for more than 3 weeks, five the patient did not have cough, five the patient had chronic lung disease, and 15 the patient had a visit in the prior 30 days. Of the 53 false negativesfor which the clinician diagnosed the patient with acute bronchitis in their note52 did not have an acute bronchitis diagnosis code (most common diagnosis codes were cough, hypertension, upper respiratory infection, and viral infection) and one had a diagnosis code of asthma that was not mentioned in the chart. The antibiotic prescribing rate for true positives was 22 %, for false positives was 4 %, and false negatives was 30 %. CONCLUSIONS: Compared to chart review, the acute bronchitis performance measure has a sensitivity and positive predictive value of less than 50 %. The performance measure omitted visits with antibiotic prescribing and falsely included visits without antibiotic prescribing, suggesting gaming. ADAPTIVE RESERVE AT COMMUNITY HEALTH CENTERS: THE CANCER PREVENTION AND CONTROL RESEARCH NETWORK (CPCRN) MULTI-STATE SURVEY Shin-Ping Tu1,2; Alan Kuniyuki2; Vicki M. Young3; Maria Fernandez4; Rebecca S. Williams5; Amanda Kimura2. 1Virginia Commonwealth University, Richmond, VA; 2University of Washington, Seattle, WA; 3South Carolina Primary Healthcare Association, Columbia, SC; 4University of Texas at Houston, Houston, TX; 5UNC Chapel Hill, Chapel Hill, NC. (Tracking ID #1936531) BACKGROUND: Adaptive Reserve comprises intangible elements of a practices capacity (i.e., human relationships, teamwork, a learning culture) that provides flexibility, resilience, and enhance positive change. Our objective is to determine the Adaptive Reserve at Community Health Centers (CHCs) in seven states (California, Colorado, Georgia, Missouri, South Carolina, Texas, and Washington). METHODS: We selected the Patient Centered Medical Home (PCMH) and the Consolidated Framework for Implementation Research to guide this research. Our survey included questions on colorectal cancer screening PCMH best practices and items from the Clinician Staff Questionnaire to examine the practice Adaptive Reserve (PAR) of participating CHC clinics. PAR scores were scaled from 0.00 to 1.00, with higher scores representing greater Adaptive Reserve. In collaboration with Primary Care Associations and CHCs, we recruited a convenience sample of 76 CHC clinics and invited their clinical staff (providers, nurses, and medical assistants) to participate in our CPCRN Community Health Center Survey. RESULTS: Survey respondents (N = 246) are mostly female (82 %) and non-Hispanic (61 %). Forty-two percent reported their clinic had 78 colorectal cancer (CRC) screening PCMH best practices (maximum 8). Preliminary PAR scores in the states ranged from 0.61 to 0.7. In a study on PCMH implementation, 36 practices expected to have significant capability for change had a mean baseline PAR score of 0.69. Our survey indicates higher levels of PAR to be positively associated with CRC screening best practices in the clinics (adjusted OR = 3.62; 95 % CI 1.31, 10.02) and how frequently staff performed the best practices (p = 0.005). CONCLUSIONS: Evidence from primary care transformation indicates that practices with strong Adaptive Reserve are more successful at incorporating change. With the Affordable Care Act, understanding the Adaptive Reserve of CHC clinics will be timely for the many changes needed to successfully expand primary care at CHCs. ADVANCE DIRECTIVE INTERVENTION FOR CHRONICALLY HOMELESS PEOPLE: ACCEPTABILITY, FACTORS ASSOCIATED WITH COMPLETION, AND TREATMENT PREFERENCES Alexander Leung1; Dhruv Nayyar1; Manisha Sachdeva1; John Song2; Stephen Hwang1. 1St. Michael, Toronto, ON, Canada; 2University of Minnesota, Minneapolis, MN. (Tracking ID #1933803) BACKGROUND: End-of-life care planning is relevant for homeless individuals because they experience high morbidity and mortality. A previous study has shown that homeless persons are more likely to complete advance directives with a one-onone counselor-guided intervention than if the advance directive is self-completed. However, the acceptability of a counselor-guided intervention for the completion of advance directives among chronically homeless individuals has not been investigated. The aim of this study was to determine the acceptability of a counselor-guided intervention for the completion of advance directives among chronically homeless individuals. Secondary objectives were to identify characteristics associated with willingness to complete an advance directive and to characterize treatment preferences in this population. METHODS: In a single-group intervention study, homeless men residing at a shelter in Toronto, Ontario were recruited from April to June 2013. Recruitment took place among programs that served chronically homeless individuals who were residents in a managed alcohol harm reduction program, a medical respite program that treats individuals with acute or uncontrolled chronic illnesses, and a long-term program for individuals who have been homeless for extended periods of time. Two hundred five homeless men (89.1 % of those approached) participated in the study. Participants completed a survey which obtained information on potential predictors of advance directive completion, including socio-demographic characteristics, health status, health care use, and attitudes towards end-of-life care and death. After completing this survey, all participants were offered an opportunity to complete an advance directive with a trained counselor. The primary outcome was the proportion of participants who completed an advance directive. The characteristics associated with completion of an advance directive were examined. RESULTS: The duration of homelessness was 12 months or longer in 72.8 % of participants. A total of 103 participants (50.2 %) completed an advance directive. Socio-demographic characteristics, health status, and health care use were not associated with completion of an advance directive. Participants were significantly more likely to complete an advance directive if they reported thinking about death on a daily basis, believed that thinking about their friends and family was important, or reported knowing their wishes for end-of-life care but not having told anyone about these wishes. Of 103 individuals who completed an advance directive, the proportion expressing a preference to receive cardiopulmonary resuscitation in the event of a cardiorespiratory arrest was 94.1 %, 81.3 %, 64.7 %, and 36.3 % if there was a chance of returning to their current state of health or hypothetical health states of moderate dementia, severe dementia, and permanent coma, respectively. CONCLUSIONS: As the homeless population continues to age, health care professionals will encounter an increase in chronically homeless persons facing complex end-of-life care needs. A single-encounter, one-on-one counselor-guided intervention can achieve a high rate of advance directive completion among chronically homeless individuals. Health care providers should initiate advance care planning discussions with their patients who are homeless. Future research aims to ascertain whether the completion of an advance directive has an effect on subsequent care in homeless individuals. ADVANCED DIRECTIVE USE IN INTERNAL MEDICINE RESIDENCY CLINICS: A RETROSPECTIVE REVIEW Briana N. Ketterer; Corinne Self; Guadalupe Martinez; Jennifer Gabbard; Macy Whitley; Ellyn Lee. University of Arizona, Tucson, AZ. (Tracking ID #1923316) BACKGROUND: Advanced directives (AD) allow patients to voice preferences about life sustaining therapies and end-of-life (EOL) care. ADs improve satisfaction, reduce stress, anxiety and depression at EOL. However, this information needs to be discussed and documented to be effective. Resident physicians play a key role as one of the first providers to encounter patients thus, making them an integral hub in the health delivery system. We sought to evaluate Internal Medicine (IM) Residents practice habits and assess participation in EOL/AD planning, and documentation of those preferences. This information will serve to address potential weaknesses in the Residency curricula and areas for improvement. METHODS: Retrospective chart review of patients seen in IM Residency clinics at The University of Arizona Medical Center (UAMC) from July 20112013. Inclusion criteria: 65 years of age with at least one chronic, debilitating diagnosis associated with morbidity or mortality (dementia, AIDS, cancer, neurodegenerative disease, chronic obstructive pulmonary disease, chronic kidney disease, cerebral vascular accident or transient ischemic attack, liver cirrhosis, coronary artery disease or myocardial infarction, post-transplant) documented in the electronic medical record, AllScripts. Documentation was reviewed and ADs were recorded along with demographics (age, gender), number and type of diagnoses. RESULTS: Reviewed 1075 charts, 165 met inclusion criteria (M 47 %, F 53 %; average age, 75; average number of problems, 7) with most frequent diagnoses of: cardiovascular disease (52 %), cancer (34 %), CKD (17 %). AD documented in 17 of 165 charts (10 %) which included: AD discussion (2), no specific qualifiers (1), Medical Power Of Attorney (8), DNR/DNI (9), Living Will (4), Hospice (1), and documentation for first responders (4). CONCLUSIONS: IM Residents at UAMC clinics documented ADs in a mere 10 % of seemingly tenuous patients 65 years of age. This represents a cohort who would benefit from AD conversations to preserve their wishes at EOL. Although residents have some training in EOL care and AD, the curricula appears insufficient as the rate of AD documentation is very low and residents feel uncomfortable with those discussions as unpublished data from our group similarly suggests. One way to improve upon this is to implement an educational intervention to enhance training, improve AD discussions, and address all AD components with patients. Improvement in this area is crucial as it would not only fulfill residency competency requirements but also lead to more complete ADs allowing patients greater autonomy in their EOL care. AGING WITH HIV: SYMPTOM BURDEN AND IMPACT ON QUALITY OF LIFE Meredith Greene1,2; Christine Ritchie1; Jeff Martin4; Joy A. Madamba3; Monica Mattes5; Yinghui Miao2; Steven Deeks3; Victor Valcour1,6. 1University of California San Francisco, San Francisco, CA; 2San Francisco VA Medical Center, San Francisco, CA; 3University of California San Francisco, San Francisco, CA; 4University of California San Francisco, San Francisco, CA; 5University of Central Florida College of Medicine, Orlando, FL; 6University of California San Franscico, San Francisco, CA. (Tracking ID #1932204) BACKGROUND: As HIV infection has shifted from a life-threatening disease to a chronic illness, the HIV-infected population is aging and often facing multiple comorbid conditions. Symptom assessment is an important clinical tool in the management of patients with complex chronic disease, yet limited data exists about symptom burden in older HIV-infected adults especially in the modern treatment era. The primary aim of this study was to describe the symptom burden (frequency and severity of symptoms) among HIV-infected adults over age 50 on combination antiretroviral therapy. We also examined factors associated with symptom burden in this population and the association of symptom burden on health related quality of life. METHODS: We conducted a cross-sectional survey in 20122013 of HIV-infected adults over age 50 that had an undetectable viral load (HIV RNA) on antiretroviral therapy and were enrolled in a clinic based cohort at a county medical center in San Francisco. Sixteen symptoms were measured by self-report in a two part question based on the 12 item HIV symptom index:1) yes/no for presence of each symptom in the past week and 2) degree of severity of each symptom (not a problem, mild, moderate, severe, very severe) as judged by symptom frequency, intensity, and duration. Symptom severity was analyzed by examining the number of symptoms rated as moderate, severe, or very severe. Health related quality of life was assessed by a single-item self-report of health as excellent, very good, good, fair or poor. Demographics, HIV related factors, co-morbidities and medications were measured by self-report and verified by chart review. Descriptive statistics were used for data analysis. RESULTS: 142 participants were enrolled, of which 94 % were male, 63 % were Caucasian, with a median age of 57 (range 5074). The median CD4 count was 576 (IQR 393715) and the median length of HIV infection was 22 (IQR 1825) years. Participants had a median of 4 (IQR 36) co-morbid conditions, most commonly hyperlipidemia (61.3 %) and hypertension (52.8 %) and were taking a median of 9 (IQR 512) non-antiretroviral medications. Sixty-nine percent of participants reported two or more symptoms, with a median of 3 (IQR 15) symptoms reported. Fatigue (n = 72, 51 %), feeling sad or depressed (n = 60, 42 %), pain, tingling, loss of feeling in feet or hands (n = 54, 38 %) and insomnia (n = 53, 37 %) were the most commonly reported symptoms. Fifty-five percent of participants reported at least one moderate to very severe symptom with a median of 1 (IQR 02) moderate to very severe symptom reported. Fatigue (n = 31, 22 %), feeling sad (n = 31, 22 %) and insomnia (n = 25, 18 %) were the most frequent severe symptoms. The total number of co-morbid conditions was associated with both the total number of symptoms reported (rs = 0.21, p = 0.01) and symptom severity (rs = 0.27, p < 0.001) and the total number of non-antiretroviral medications was associated with symptom severity (rs = 0.21, p= 0.01). HIV related factors such as CD4 count and length of HIV infection did not have statistically significant correlations with symptom frequency or severity. Both the total number of symptoms and symptom severity were inversely correlated with overall health related quality of life (rs 0.42, p < 0.001; rs 0.47, p < 0.001). CONCLUSIONS: Among adults age 50 and older with well-controlled HIVinfection, symptom burden was substantial and associated at least in part with comorbid conditions. Addressing symptom burden during routine clinical encounters merits attention and may improve quality of life as the HIV-infected population ages. ALLOWING PATIENT CHOICE BETWEEN TWO DIVERSE DIETS DID NOT IMPROVE WEIGHT LOSS William S. Yancy1,2; Cynthia Coffman1,3; Valerie Smith1; Stephanie B. Mayer4; Ronette L. Kolotkin6; Paula J. Geiselman5; Megan McVay1; Eugene Z. Oddone1,2; Corrine Voils1,2. 1VAMC, Durham, NC; 2Duke University, Durham, NC; 3Duke University, Durham, NC; 4Virginia Commonwealth University, Richmond, VA; 5Louisiana State University, Baton Rouge, LA; 6Quality of Life Consulting, Durham, NC. (Tracking ID #1941417) BACKGROUND: A variety of diet approaches lead to modest weight loss, and dietary adherence is a strong predictor of success. Allowing individuals to choose diet composition after considering their food preferences may improve adherence by increasing autonomy and self-efficacy. In a randomized trial, we examined whether participants who chose a diet experienced greater weight loss than those randomly assigned a diet. METHODS: Participants were Durham Veterans Affairs Medical Center outpatients aged <75 years with body mass index (BMI) 30 kg/m2 and no unstable health problems. They were randomized to the Control or Choice arm, stratified by sex, BMI < or 40 kg/m2, and presence of type two diabetes. Control participants were randomly assigned to follow a low-carbohydrate (LCD) or a low-fat/reduced-calorie (LFD) diet. Choice participants were presented individual results from the Geiselman Food Preference Questionnaire, indicating which diet fit their food preferences best, as well as information about the two diets; 1 week later, they were called to elicit their diet choice. All participants then received diet counseling led by a dietitian in groups of 819 participants separated by arm and diet approach. In-person group sessions occurred every 2 weeks for 24 weeks, then every 4 weeks for 24 weeks interspersed with one-on-one telephone sessions mid-month from the dietitian focusing on behavioral goal-setting. All participants received literature specific to their diet; a pocket calorie/carbohydrate/fat counting guide; and a recommendation to exercise 150 min per week. At week 12, Choice participants could switch to the other diet if dissatisfied with their initial diet choice. The primary outcome, weight, was measured on a digital scale at the 19 visits in light clothing and no shoes. Secondary outcomes measured every 12 weeks were dietary intake by Block food frequency questionnaire and the Impact of Weight on Quality of Life-Lite (IWQOL-Lite) questionnaire (scale 0100; higher score indicates higher quality of life). Linear mixed models were fit for these continuous outcomes, and included a common intercept, time effect, time*treatment interaction, randomization stratification variables and random effect to account for group clustering. RESULTS: A total of 207 participants were enrolled. Mean age was 54 years, 27 % were women, 51 % were White, 43 % were Black, 22 % had diabetes. Mean weight at baseline was 109 kg and mean BMI 36 kg/m2. In the Choice arm, 61 (58 %) chose the LCD and 44 (42 %) chose the LFD; only three LCD and two LFD participants switched diets at 12 weeks. Final measurements were collected in 88 of 102 (86 %) in the Control arm and 87 of 105 (83 %) in the Choice arm. Mean weight loss at 48 weeks was 5.3 % in Control participants and 4.2 % in Choice participants (difference 1.1 %, 95 % CI 0.0 to 2.2, p = 0.06). Dietary adherence was similar in the Control and Choice arms, with mean percent deviation from macronutrient goal differing by 0.6 % (10.1 % vs 10.7 %, respectively) at week 48. Mean IWQOL-Lite total scores were similar at week 48 (Control 82.9, Choice 81.7, difference 1.3, 95 % CI 4.5 to 2.0). CONCLUSIONS: Educating patients on diet approaches and their food preferences, and then allowing them to choose their diet composition, did not result in greater weight loss compared to random diet assignment. Future research could consider targeting diet to individuals based on metabolic, genetic or other characteristics that might predict improved outcomes. AN ACHILLES HEEL OF DISPARITIES RESEARCH: MISSING RACIAL DATA IN A STUDY OF MORTALITY AFTER CRITICAL ILLNESS Mallika Mundkur1,2; Swapna Abhyankar1; Fiona M. Callaghan1; Clement J. McDonald1. 1National Institutes of Health, Bethesda, MD; 2George Washington University, Washington, DC. (Tracking ID #1936298) BACKGROUND: Research has consistently demonstrated the presence of racial disparities in care and outcomes across inpatient and outpatient settings. However, studies focused on disparities in the intensive care unit (ICU) have yielded conflicting results, albeit with considerable variation in design and casemix. Some analyses of ICU outcomes have detected greater mortality for minorities as compared to non-minority populations for conditions such as severe sepsis or trauma (Mayr et al., 2010; Haider et al., 2007), while others have found lower mortality rates for minorities with conditions such as community-acquired pneumonia (Frei et al., 2010). Recently, a retrospective study of 9518 patients in California concluded that in-hospital all-cause mortality for ICU patients was equivalent across race categories after adjusting for socioeconomic status, resuscitation preferences and severity of illness (Erickson et al., 2011). In the study by Erickson et al., patients without known race, equivalent to 10 % of overall study population, were excluded from the analysispotentially biasing results, but noted to be common practice among other studies of racial disparities (Long, Bamba, Ling, Shea, 2006). We designed our study not only to contribute to the unresolved discourse on disparities in the ICU but also to ensure that patients with missing race would be examined. METHODS: We performed a retrospective cohort study utilizing the Multiparameter Intelligence Monitoring in Intensive Care (MIMIC-II) research database (Saeed et al., 2011; Lee et al., 2011). MIMIC-II contains de-identified clinical information on 32,425 adult and neonatal ICU patients admitted to Beth Israel Deaconess Hospital in Boston, MA between 2001 and 2008. Our cohort was comprised of adult patients (age >18) admitted to a medical or surgical ICU for at least 4 h for a non-elective admission. For patients with multiple ICU admissions, only the first stay was included. Our primary outcome was 30-day all-cause mortality. Covariates were: race, age, gender, marital status, type of insurance coverage (indicator of socioeconomic status), code status within 2 h of ICU admission, Simplified Acute Physiology Score (SAPS) (severity of illness indicator), admission source (e.g., emergency room vs. transfer) and comorbidities including congestive heart failure, chronic renal failure/ end-stage renal disease, chronic liver disease, and cancer. We used demographic data to group patients into the following race categories: White, Black, Hispanic, Asian/ Pacific Islander, Other and included an Unknown category for patients without documented race. We used the software package R for all statistical analysis. We used Kruskal-Wallis, Pearsons chi-squared, and Fishers exact tests to assess the AN EVALUATION OF CLINICALLY PROVEN DIETARY SUPPLEMENTS Alysha Melnyk; Bimal Ashar; April S. Fitzgerald. Johns Hopkins University, Baltimore, MD. (Tracking ID #1939374) BACKGROUND: The Dietary Supplement Health and Education Act (DSHEA) permits manufacturers to sell products without providing pre-market evidence of safety or efficacy. One fundamental reason for the passage of the DSHEA was to empower consumers to make their own choices. Yet, data suggests that the public is unaware of the limited regulation on product marketing and advertising. Dietary supplements are frequently advertised as being clinically proven. Yet, there are not standards in place to determine acceptable use of this term. This study was undertaken to evaluate products that have been advertised as being clinically proven in order to determine the availability of data used to make such an assertion. METHODS: An internet survey for products was performed by using the search the search tems clinically proven dietary supplements, dietary supplements, best dietary supplements in common search engines. Advertisements were then reviewed to see if they contained terms that suggested definitive efficacy. Products whose advertisements contained terms such as clinically proven, scientifically proven, clinically supported were included in the study. Products returned via the search were then categorized for what they were purported to be proven to do. Additionally, the supplement advertisements were analyzed to determine if evidence was readily available to consumers that would support the clinically proven label. If evidence supporting the statement was not readily available on their website, companies were contacted to attempt to obtain any information supporting their claim. Descriptive statistics were used to categorize products and determine what percentage of such products had available references to studies supporting their claims. RESULTS: A total of 99 dietary supplements were found that referenced definitive scientific efficacy. Most of these products (63 %) were advertised as being useful for helping consumers lose weight and/or build muscle. Despite the assertion for efficacy, only 28 made direct reference on their website to any scientific studies. After being contacted, an additional eight companies provided studies and/or links to evidence. The remaining companies (64 %) either had no testing to report (21 %), claimed that they had testing but would not provide it (4 %), claimed that their evidence was all on their website (13 %), did not respond to inquiries (14 %), or had no listed contact information on their internet advertisement (11 %) CONCLUSIONS: A number of dietary supplements are advertised as being clinically or scientifically proven, yet provide little to no evidence to consumers to substantiate their claims. Truth-in-advertising standards for dietary supplements should define parameters for using such terminology in order to protect the public from potentially misleading claims. AN EVALUATION OF COMPUTERIZED DRUG-DRUG INTERACTION ALERT OVERRIDES IN AMBULATORY CARE Sarah P. Slight1,2; Diane L. Seger3; Karen C. Nanji4,5; Insook Cho1,6; Nivethietha Maniam3; Patricia C. Dkyes1,4; David W. Bates1,4. 1Brigham and Womens Hospital, Boston, MA; 2Durham University, Stockton on Tees, United Kingdom; 3Partners Healthcare Systems, Inc.,, Boston, MA; 4Harvard Medical School, Boston, MA; 5Massachusetts General Hospital, Boston, MA; 6Inha University, Incheon, MA. (Tracking ID #1935807) BACKGROUND: Computerized physician order entry with clinical decision support (CDS) represents a valuable tool for improving medication safety, quality, and efficiency. CDS systems can provide health care providers with real-time, relevant, patient-specific information and alert them to potential prescribing errors. Providers may choose to override these alerts however, selecting a coded reason in order to proceed with the medication order. It is often unclear whether these particular actions, e.g., will monitor drug levels as recommended, are carried out by the provider or just simply ignored. We evaluated why providers overrode drug-drug interaction (DDI) alerts and what actions they took as a consequence of the alert. METHODS: This study was a cross-sectional, observational study of DDI alerts generated over a three-year period between January 2009 and December 2011. All Level 2 DDI alert overrides were downloaded with IRB approval from 36 primary care practices affiliated with two Harvard teaching hospitals (total 14,966 overrides, 60.2 % of alerts generated). Any duplicates were removed and replaced. A physician and pharmacist expert panel assessed the appropriateness of these DDI alert overrides. A detailed chart review was conducted for each of the appropriate overrides to ascertain whether the provider had carried out their intended action(s). RESULTS: Providers appropriately overrode just over two-thirds (338/496) of the DDI alerts and carried out the intended action in less than two-thirds (214/ 338) of these cases. Eight drugs were responsible for generating approximately three quarters of these alerts. The most common coded reasons for overriding DDI alerts were will monitor as recommended (43.9 %, n = 218), will adjust dose as recommended (16.9 %, n = 84), and patient has already tolerated combination (15.7 %, n = 78). Of the 121 appropriate alert overrides where the provider said they would monitor as recommended, only 35.5 % (n = 43) actually completed the monitoring. Our study also revealed how only 60 % (n = 21) of providers who indicated that they would adjust the dose as recommended actually did. Furthermore, some providers selected the coded reason patient has already tolerated combination, yet no information was found to suggest that the patient had been taking both drugs together previously. CONCLUSIONS: We found that providers continue to override important and useful alerts, some of which likely cause serious patient injuries, although we did not measure injuries in this study. A key difference between our study and previous studies1 is that we also reviewed the charts to see whether the provider actually took action as a consequence of the alert. We found that the intended action was carried out in only two-thirds of cases. More research is needed to explore the human factors elements that influence provider behaviour, such as patient demands, workload, and time constraints. References: 1. Shah NR, Seger AC, Seger DL et al. Improving Acceptance of Computerized Prescribing Alerts in Ambulatory Care. JAMIA 2006;13:511. AN EVALUATION OF COMPUTERIZED MEDICATION ALERT OVERRIDE BEHAVIOR IN AMBULATORY CARE Nivethietha Maniam1; Sarah P. Slight1,3; Diane L. Seger1; Mary Amato1,7; Dustin McEvoy1; Karen C. Nanji4,5; Insook Cho1,6; Patricia C. Dkyes2,4; David W. Bates1,2. 1Partners HealthCare, Cambridge, MA; 2Brigham and Womens Hospital, Boston, MA; 3The University of Durham School of Medicine, Stockton on Tees, United Kingdom; 4Harvard Medical School, Boston, MA; 5Massachusetts General Hospital, Boston, MA; 6Inha University, Incheon, Republic of Korea; 7Massachusetts College of Pharmacy and Health Sciences University, Boston, MA. (Tracking ID #1939988) BACKGROUND: While evidence suggests that computerized decision support (CDS) increases safety and quality of care, understanding how physicians respond to CDS alerts is a critical factor in achieving meaningful use of electronic health records. We continue to observe a high level of medication alert overrides for many prescription domains. While some overrides are justified clinically, others are not, and it is important to be able to reach out to those providers who are not prescribing optimally and understand their reasons for overriding alerts. METHODS: All Level 2 alert overrides that required a coded reason to be provided at the time of prescribing were downloaded between January 2009 and December 2011 in the outpatient primary care setting. We limited our sample to providers who had received 20 or more alerts in each of the prescribing domains (drug-drug interaction, drug-allergy interaction, renal suggestion, age-based, duplicate-drug, and formulary substitution alerts) and the number of times each provider overrode these alerts was calculated. Of the 725 providers eligible, those with a high inappropriate override rate (above 75 % within a specific domain or overall) were targeted for interviews. The sessions were tailored to each providers overrides and conducted by a research pharmacist or physician trained in counter-detailing techniques. Graphical material, including performance level data, a list of their inappropriate overrides and supporting evidence-based summaries, was presented to each provider during the discussion. An analysis of the data was carried out and prevalent concepts related to general views on alert functionality and specific prescribing behavior identified. RESULTS: We conducted 23 interview sessions across primary care clinics affiliated with Brigham and Womens Hospital and Massachusetts General Hospital. We identified seven high level content categories: clinical satisfaction, clinical utility and relevance, variant user knowledge, impact on clinician reviewing process, patient preferences, current alerting tool challenges, and considerations for the future. Overall, providers were favorable towards the alerts and found them helpful in identifying possible adverse interactions. Many providers, however, found that the clinical relevance of the alerts could be improved by including magnitude of risk, suggested alternate treatments, recent laboratory values, and more up-to-date reference material. Additionally, many participants said they failed to provide a valid override reason because they found the alerts time-consuming and disruptive. Limitations of the existing EHR infrastructure such as inaccurate medication and allergy lists were identified as contributing to unnecessary alerts. Creating a more patient specific alerting tool was recognized as a future area of focus. CONCLUSIONS: Many providers were unaware of their relatively high rate of overriding and this study allowed providers the unique opportunity of objectively assessing their prescribing behavior. Key issues that emerged from the sessions included the perceived risk to physician autonomy in decisionmaking, the increase of clinically irrelevant alerts leading to alert fatigue, and the lack of supplementary clinical information. By incorporating provider preferences, customizing alerts to the context of the visit, and offering additional clinical data, providers felt that CDS alerts would be less likely to be overridden providing more effective, efficient, and equitable care. AN INTERNATIONAL EVALUATION OF DRUG-DRUG INTERACTION ALERTS THAT SHOULD BE NON-INTERRUPTIVE IN U.K. AND U.S. SETTINGS Alexandra Robertson1; Sarah P. Slight1,2; Diane L. Seger3; Sarah K. Thomas4; Jamie Coleman4; David W. Bates1,5; Shobha Phansalkar1,6. 1Brigham and Womens Hospital, Boston, MA; 2The University of Durham, Durham, United Kingdom; 3Partners HealthCare Systems, Inc., Wellesley, MA; 4The University of Birmingham, Birmingham, United Kingdom; 5Harvard Medical School, Boston, MA; 6Wolters Kluwer Health, Indianapolis, IN. (Tracking ID #1937914) BACKGROUND: Clinical Decision Support (CDS) has the potential to improve patient safety by providing knowledge and support at the point of prescribing. CDS systems can, however, deliver an overdose of alerts that may result in alert fatigue, with providers overlooking clinically significant alerts as well as those considered less critical. In a previous study, Phansalkar et. al. (2012), identified 33 non-critical drug-class and class-class interaction alerts that could drug effects including medication errors, all of which adversely affect patient outcomes. In addition, it leads to increased cost of health care through duplication of services (35). Several studies show that comprehensive interdisciplinary assessments can greatly improve outcomes in transitions of geriatric care (6,7). Utilizing IPE experiences in training health care professionals can foster effective collaboration across these professions and greatly improve transitions of care in the elderly. As part of the University of Virginia Health Systems initiative on IPE, we developed and implemented an interprofessional geriatric education workshop for nursing and medical students with a focus on transitions in geriatric care. SETTING AND PARTICIPANTS: A total of 192 students (107 medical students and 85 nursing students) participated in a 90-minute, interactive, case-based workshop. DESCRIPTION: The monthly workshops were conducted with groups of an average of 12 medical and 9 nursing students over 1 year. These were facilitated by nursing and medical faculty. Self-perceived competencies in necessary IPE skills and attitudes toward interprofessional teamwork were measured through surveys. Data were analyzed using descriptive and nonparametric statistics. EVALUATION: Over 90 % of students were better able to describe the necessary interprofessional communication needed to develop a patient-centered care plan in transitioning patients between clinical sites. Four out of five students reported an enhanced appreciation of interprofessional teamwork and stated that the learning experience was valuable. They were also able to identify legal, financial and social implications in transitions of care (77 %). Interestingly, nursing students consistently rated the workshop more highly than medical students across most domains (P< .05). DISCUSSION/REFLECTION/LESSONS LEARNED: Students improved and demonstrated their knowledge of interprofessional communications skills and strategies required in transitions of geriatric care. Their teamwork skills were also enhanced. Introducing these concepts in medical and nursing training fosters effective interprofessional communication and collaboration. References 1. WHO (World Health Organization) 2010. Framework for Action on Interprofessional Education & Collaborative Practice. Geneva: WHO/HRH/HPN/10.3. 2. Naylor, MD, Hirschman, KB, Bowles, KH, Bixby, MD, Konick-McMahan, J, and Stephens, C. Care coordinator for cognitively impaired older adults and their caregivers. Home Health Care Services Quarterly 26(4), 5778, 2007. 3. Jack BW, Chetty VK, Anthony D, et al. Ann Intern Med 3;150(3):17887, 2009. 4. Boockvar KS, Carlson LaCorte H, Giambanco V et al. Am J Geriatr Pharm 4(3):23643, 2006. 5. American Medical Directors Association. Transitions of Care in the Long -Term Care Continuum. Clinical Practice Guideline. Columbia, MD: AMDA 2010. 6. Berwick, AD, Rees, K, Dieppe, p, et al. Complex interventions to improve physical function and maintain independent living in elderly people: A systematic review and meta-analysis. Lancet 371(9614):725735, 2008. 7. McCusker, J, and Verdon, J. Do geriatric interventions reduce emergency department visits? A systematic review. Journals of Gerontology: Series A: Biological Sciences and Medical Sciences 61A(1):5362, 2006. INTERDISCIPLINARY EVIDENCE-BASED MEDICINE COURSE DESIGN FOR NAS MILESTONE ASSESSMENT THROUGH ONLINE ADAPTIVE LEARNING STRATEGIES Ravi Gupta1; Kell Julliard2; Juan Barrientos2. 1Lutheran Medical Center, Brooklyn, NY; 2Lutheran Medical Center, Brooklyn, NY. (Tracking ID #1940936) NEEDS AND OBJECTIVES: ACGMEs Next Accreditation System (NAS) measures competency in essential tasks of clinical practice, including the ability to find and evaluate clinical research to improve care. Journal Clubs are our primary teaching venue for evidence-based medicine (EBM), but we were unable to fully assess all participants understanding of the concepts presented. Moreover, residents leading journal clubs had variable prior knowledge of EBM, resulting in similarly variable learning experiences for the audience. On the wards, attending physicians evaluated residents integration of EBM with clinical care as opportunities arose, leading to inconsistent assessment. The Association of Professors of Medicine reported that although 59 % of online courses reviewed had some type of assessment, none monitored resident learning or objectively reported on milestones (2012). The study called for development of e-Learning courses perceived by students as providing an advantage over traditional curriculum. Thus, our specific project goals were to: Develop clear EBM learning objectives related to milestones and national guidelines Bring teaching of EBM in line with recent developments in learning theory Take advantage of new learning technologies Enhance resident learning and clinical application of EBM The project team reflected the three essential areas of expertise: a physician provided clinical relevance and accuracy; an EBM expert led content development; and an instructional designer created a pedagogically sound course design informed by learning theory. SETTING AND PARTICIPANTS: Internal Medicine residents and faculty at an inner city community teaching hospital DESCRIPTION: The EBM Curriculum comprises the following modules: Introduction to EBM, Randomized Controlled Trials, Cohort and Case-control Studies, Diagnostic Tests, and Meta-Analyses. The course is SCORM compliant and adaptable to any Learning Management System. Modules are optimized for mobile platforms on Android and iOS. Modules cover the basic principles of each EBM study type and integration of research findings with clinical care through the use of four primary modalities: video, animation, simulation, and text. The curriculum emphasizes userdriven adaptability of course content (adaptive learning): the sequence of the course is dependent on user responses to continuous, embedded assessment. Users are remediated only as necessary. Data document each users proficiency. Two modules are complete: Introduction to EBM and Randomized Controlled Trials. The third will be complete by the date of the conference. Each assessment item is mapped to specific curricular NAS milestones. User responses including applicable milestone, correctness of responses, number of attempts, and need for remediation are routed to departments for reporting to ACGME. EVALUATION: 97 % of residents were either satisfied or very satisfied with the course. Ninety-seven percent believed that the modules would improve their clinical skills and practice. DISCUSSION/REFLECTION/LESSONS LEARNED: Residents participated willingly in taking the modules, and the great majority perceived them as directly benefiting clinical practice. In module testing, residents individual learning preferences became apparent, which will guide development of future modules. Making explicit the evidence grading system for online EBM repositories was perceived by residents and faculty as having great clinical relevance. Significantly more time was needed to develop modules than originally anticipated. ONLINE RESOURCE URL (OPTIONAL): Two modules are available for review at http://lmcmc.certpointsystems.com. Login names are SGIM1, SGIM2, SGIM3, SGIM4, SGIM5; password is password. INTERNAL MEDICINE RESIDENTS ATTITUDES AND BELIEFS OF ACHIEVING THEIR CLINICAL COMPETENCY MILESTONES DURING THEIR NIGHT FLOAT ROTATION Brian E. Foster; Maryann T. Ally. Walter Reed National Military Medical Center, Bethesda, MD. (Tracking ID #1937176) NEEDS AND OBJECTIVES: We created standardized presentations for the internal medicine residents on their two-week night float rotation, but it is not known how this impacts their learning experience. At our medical center, most attending physicians supervise cross-cover issues of night float residents during one 2-week period per academic year, leading to a varied learning experience. Our PowerPoint presentations encompass common, cross-cover issues, including atrial fibrillation with rapid ventricular response, bowel management, chest pain, delirium, hypotension, hypertension, mechanical falls, neutropenic fever, pain management, post-operative complications, and stroke. The presentations were disseminated on our passwordprotected internal medicine website for easy access. The resident on the night float team was instructed to lead a discussion with the interns using these presentations. We developed a survey to assess internal medicine residents attitudes and beliefs of achieving their clinical competency milestones after the implementation of these standardized presentations. SETTING AND PARTICIPANTS: Internal medicine residents in a single-site, academic medical center. DESCRIPTION: A voluntary and anonymous 9-question survey was distributed electronically to our internal medicine residents during a two-week period in December 2013. EVALUATION: Of the total 70 internal medicine residents, 23 residents responded to the survey, including 10 PGY-1, 9 PGY-2, and 4 PGY-3. Of the survey respondents, 12 had completed the night float rotation, and 11 had used the night float standardized presentations as outlined above. 84.6 % had used between 1 and 3 lectures on average per week while 15.4 % used an average of 46 lectures per week. The majority (77 %) spent an average of 515 min per presentation. 45.5 % felt the presentations allowed them to become better teachers, 18.2 % referred to the presentations during the night float rotation when encountering a clinical problem, and 9.1 % referred to the presentations after the night float rotation was completed when encountering a clinical problem. Over half (54.5 %) felt the night float presentations improved their overall clinical knowledge and 1 respondent (9.1 %) stated the presentations improved development of a comprehensive management plan. 27.3 % noted improvement in their effective communication of diagnostic and treatment plans to patients and caregivers after utilizing these presentations. One respondent (9.1 %) felt clinical questions to consultants were better framed as a result of these presentations. DISCUSSION/REFLECTION/LESSONS LEARNED: Using a brief survey, we assessed the attitudes and beliefs of our internal medicine residents regarding their clinical competency milestones during their night float rotation. In developing and implementing standardized presentations on commonly encountered cross-cover issues, we aimed to enhance the learning experience of our residents and to encourage our residents to become better teachers, communicators, and, ultimately, independent practitioners. This survey identified areas of resident improvement, including developing more comprehensive management plans and communicating more clearly with patients and consultants. We will address these needs by strengthening our curriculum and developing more quality improvement initiatives with our residents. LONGITUDINAL PATIENT SATISFACTION DATA AFTER INSTITUTING A 4 + 2 PRACTICE MODEL IN RESIDENT CONTINUITY CLINIC Julie Oyler; Lisa M. Vinci; Wei Wei Lee; Eileen Wang; Vineet Arora. University of Chicago, Chicago, IL. (Tracking ID #1938913) NEEDS AND OBJECTIVES: Many internal medicine programs have recently instituted a block scheduling model such as the 4 + 2 model for ambulatory training. As opposed to traditional training models where residents leave inpatient responsibilities for one half day a week to take care of continuity clinic patients, the 4 + 2 model of training separates inpatient and outpatient responsibilities. Little research has been done on the impact of this training model on patient satisfaction. SETTING AND PARTICIPANTS: On July 1, 2013, an academic internal medicine training program transitioned to a 4 + 2 model in which residents spend 4 weeks on inpatient rotations followed by 2 weeks of outpatient clinic. During the 2 weeks on outpatient, residents participate in 8 half day continuity clinics, 46 half day subspecialty clinics, 4 half day teaching sessions and 24 administrative sessions per 2 week block. During the 4 weeks on inpatient, residents continuity clinic patients are covered by another resident. DESCRIPTION: Since 2010, American Board of Internal Medicine (ABIM) Cancer Screening Practice Improvement Module (PIM) paper surveys have been administered annually from July to November to patients in the resident continuity clinic. PGY2 residents hand anonymous paper surveys to patients, these surveys are then entered into the ABIM PIM website by staff. This annual patient survey data was used to evaluate patient demographics, satisfaction and access to the resident practice before and after the 4+ 2 curricular change. EVALUATION: From 2010 to 2012 (pre data) 273 patient surveys and in 2013 (post data) 90 were completed. The majority of patients were female (pre 188/273, 68 %, post 68/90, 76 %) and over the age of 50 (pre 209/273, 76 %, post 76/90, 84 %). About one third of patients rated their health fair to poor (pre 88/273, 32 %, post 30/90, 33 %) and another third rated their health very good to excellent (pre 64/273, 23 %, post 31/90, 34 %). Patients responded that the practice was excellent at encouraging question and answering them clearly (pre 144/273, 53 %, post 51/90, 57 %) and that they would recommend this practice to others (pre 253/273, 93 %, post 80/90, 89 %). When evaluating access to the resident practice, patients reported no problems with scheduling appointments (pre 167/269, 62 %, post 65/90, 72 %), reaching someone with a question (pre 175/262, 67 %, post 66/88, 75 %), obtaining prescription refills (pre 196/258, 76 %, post 60/87, 69 %), obtaining test results (pre 212/258, 82 %, post 75/84, 89 %), and obtaining referrals (pre 191/241, 79 %, post 65/84, 77 %). DISCUSSION/REFLECTION/LESSONS LEARNED: The 4+2 schedule more closely matches the responsibilities of a general internist. However, there are periods of 4 8 weeks at a time when the resident is not available and covered by a fellow resident. Interestingly, patients were more satisfied with scheduling appointments but less satisfied with obtaining prescription refills. Despite this curricular change there was no significant change in most areas of patient satisfaction or access to the resident practice. LONGITUDINAL RESIDENT-STUDENT PATIENT CARE COLLABORATION IN PRIMARY CARE Amy Pasternack1,2; Katherine Lupton1,2. 1Cambridge Health Alliance, Cambridge, MA; 2Harvard Medical School, Boston, MA. (Tracking ID #1939424) NEEDS AND OBJECTIVES: Although the majority of patient care in internal medicine occurs in the ambulatory domain, the bulk of residency training has traditionally taken place in the inpatient setting. Consistent with this, residents receive little exposure to and training in teaching medical students outpatient medicine. Our near-peer educational exchange between PGY-2 and -3 residents and third year medical students will build residents skills in outpatient teaching, improve continuity of care for complex patients, and increase the number of complex adult primary care patients that medical students follow longitudinally. SETTING AND PARTICIPANTS: Cambridge Health Alliance (CHA) is well known for pioneering a longitudinal integrated clerkship for third year Harvard medical students, the Cambridge Integrated Clerkship (CIC). Most of the student teaching in this program is done by attending physicians in the ambulatory setting; as a result, CHA internal medicine residents have fewer opportunities to teach medical students than do residents at many other medicine training programs. The CHA internal medicine residency follows a 2 + 4 immersion model in which two-week ambulatory blocks with concentrated time in primary care clinic alternate with fourweek inpatient or elective rotations when residents are out of continuity clinic entirely; during inpatient blocks it can be challenging for residents to maintain close follow up with their more complex patients. This project is being piloted at the Windsor Street Health Center, a safety-net community health center that is the ambulatory medicine training site for six internal medicine residents and four CIC students. DESCRIPTION: We are currently piloting a collaborative patient care experience in which CIC students pair with PGY-2 and -3 primary care providers to longitudinally co-follow complex medical patients. The student and resident see their collaborative patients together: the student is primarily responsible for running the patient visits while the resident directly observes and helps guide the encounter. The student then presents to the residentwho acts as preceptor for the visitin the presence of an experienced faculty preceptor who ensures the quality of patient care while observing and later providing feedback on the residents teaching skills. The students follow their collaborative patients over the course of the clerkship year, benefiting from frequent direct observation and feedback while learning principles of ambulatory medicine and serving as a point of continuity with patients while residents are out of clinic. The residents, who also complete an ambulatory resident-as-teacher curriculum, gain mentored experience teaching students outpatient medicine. Further work is needed to build the resident-as-teacher curriculum, generate guidelines for patient selection, and establish administrative processes to facilitate the complex schedules involved. EVALUATION: We are in the early stages of piloting this project. An initial pilot by one resident and two medical students has shown the project to be feasible within the resident and medical student clinic schedules, and other residents and medical students have expressed interest in participating. Residents and students will complete mid-year and year-end surveys assessing their comfort with ambulatory teaching and primary care medicine, respectively, and residents teaching skills will be periodically assessed by students and faculty. DISCUSSION/REFLECTION/LESSONS LEARNED: This program increases residents exposure to medical student teaching in a supervised environment in which they can explore outpatient teaching styles and techniques while receiving feedback and mentoring from experienced educators. Prior to this, ambulatory resident-as-teacher experiences in this residency were limited to resident-run peer didactics. Students benefit from near-peer teaching and direct observation, and are empowered by the opportunity to be an integral member of the healthcare team and serve a key role in maintaining patient continuity over the course of the clerkship year. Defining appropriate patientswho are medically and psychosocially complex enough to require frequent clinic visits and would benefit from continuity with a medical student, but not so complex that the student is overwhelmed and unable to run the visitshas proven to be an early challenge and is a main focus of program development for the remainder of the pilot year. MEDICAL STUDENTS AS PRIMARY CARE PROVIDERS: AN INTEGRATED MEDICAL STUDENT CLINIC THAT OFFERS A TRUE LONGITUDINAL PRIMARY CARE EXPERIENCE Serene I. Chen; Rebecca Liu; Pinar OraySchrom. Yale School of Medicine, New Haven, CT. (Tracking ID #1925854) NEEDS AND OBJECTIVES: Longitudinal clinical care experience is a rarely-met need in medical school education. The Wednesday Evening Clinic (WEC) at the Yale New Haven Hospital adult Primary Care Center (PCC) is a student clinic founded to address this need. It aims to: 1. Provide Yale medical students opportunity for a longitudinal clinical outpatient experience. 2. Offer students the opportunity to care for their own patient cohort and function as primary care providers (PCPs). 3. Provide outstanding medical care for an underserved patient population. SETTING AND PARTICIPANTS: The WEC is a weekly student-staffed clinic, fully-integrated into the adult PCC of Yale-New Haven Hospital. Annually, 15 medical students beyond their 3rd year participate for at least 1 year. DESCRIPTION: The WEC was founded in response to medical students desire for a longitudinal clinical experience. Modeled after group practice, students are divided into patient care teams, each supervised by a designated faculty attending and rotating volunteer attendings from the Yale and New Haven medical communities. As PCPs, students perform general health maintenance, chronic disease management, ageappropriate screening, counseling and patient-care coordination. Follow-up extends beyond clinic hours. Weekly attendance over 1 year, in conjunction with regular evaluations by faculty, satisfies the primary care clerkship requirement. The WEC is supported jointly by Yale-New Haven Hospital and Yale School of Medicine. EVALUATION: We surveyed WEC students from 2008 to 2013 on their experience. Throughout this time, over 60 medical students have served as PCPs. Survey results indicate that the WEC provides a unique longitudinal clinical experience not found in the traditional medical school curriculum. Students unanimously report increased comfort in providing ambulatory care after participating in the WEC. In addition, 38 % of WEC graduates pursued residency in primary care specialties versus 33 % among non-WEC graduates. We surveyed patients on their clinic perception and satisfaction. Results indicate that patient flow and wait time are areas for improvement. Prolonged visits and delays are likely due to the requirement for close supervision by attendings. DISCUSSION/REFLECTION/LESSONS LEARNED: Unlike other free or urgent care student clinics, the WEC is fully integrated into a well-established PCC of a teaching hospital, enabling students to follow results and care over time. The WEC is a unique experience in terms of degree of responsibility, ownership, and longitudinal exposure to patient care, with students functioning as PCPs. It provides insights into disease courses and opportunities for building patient-provider relationships, which are of increasing importance given the rising demand for primary care. The WEC fulfills primary care clerkship requirements, an additional innovation that builds structure for regular feedback and evaluation. The WEC is possible and sustainable by support from Yale School of Medicine, Yale-New Haven Hospital and dedicated volunteer attendings. Billing from encounters contributes to covering facility and staff costs. However, the long-term financial sustainability of the clinic depends on the overall financial health of the Yale-New Haven Health System. While similar opportunities are presently rare, this model can be scaled and implemented elsewhere, especially in light of increasing demand for primary care training. MOBILE DEVICE APPLICATION BASED EDUCATIONAL TOOL TO IMPROVE COST-CONSCIOUS CARE AMONG INTERNAL MEDICINE RESIDENTS Auras R. Atreya1; Behdad Besharatian1; Stefano Buccy2; Thomas Higgins1,3. 1Baystate Medical Center/Tufts University School of Medicine, Spring field, MA; 2University of Massachusetts, Amherst, MA; 3Baystate Medical Center/ Tufts University School of Medicine, Springfield, MA. (Tracking ID #1939067) NEEDS AND OBJECTIVES: Cost of care in the United States is substantially higher than other nations and health care over-utilization is an area of focus in todays economic environment. The American Board of Internal Medicine Foundation recently launched the Choosing Wisely initiative, aimed at reducing overuse of tests and procedures. Of the major drivers of overuse, one of the proposed physician-related factors is physician culture, which emphasizes thoroughness during medical education and training years. It is possible that this leads to increased test ordering in order to obtain an exhaustive evaluation. However, it has been previously shown that medical residents have poor knowledge of the cost of diagnostic tests. Given that costeffective care is part of the competencies required by the Accreditation Council for Graduate Medical Education (ACGME), there exists an opportunity to improve awareness of diagnostic tests and facilitate cost-effective care. Our objective was to develop an educational intervention that would raise awareness of medical residents with regard to frequently ordered diagnostic tests. SETTING AND PARTICIPANTS: Baystate Medical Center is an academic medical center in Springfield, MA that serves as the Western Campus of Tufts University School of Medicine. Educational curriculum for internal medicine residents and medicine-pediatrics residents on the inpatient medicine service includes a half-hour morning report that is based on clinical case-based learning. This is an interactive session where residents hone critical decision-making and are involved in determining appropriate work-up for each particular presentation. DESCRIPTION: We developed an android OS based application for mobile devices that would display the cost of commonly ordered tests. Cost of tests was based on feeschedule that was obtained from the Financial Department at Baystate Medical Center. The mobile device can be connected to a projector and the cost of the last ordered test is displayed along with total cost of tests ordered in real-time during morning report. Tests included in the application comprise of serum chemistries, hematological tests, cardiac enzymes, CSF, urine and stool studies. Frequently ordered radiological tests, echocardiography and stress tests were also built into the application. EVALUATION: Prior to implementing this application in morning report sessions, we conducted a pre-intervention survey where we asked all residents present at an academic half day (protected time to ensure attendance of all internal medicine and medicine-pediatrics residents) to estimate the value of 10 commonly ordered diagnostic tests. We will then educate residents to use the application during morning report. Since only the residents on inpatient medicine service attend morning report, we will have a group of residents exposed to the application in addition to those not exposed to the application. This will then allow us to conduct a post-intervention survey to compare the effect of the educational intervention in those exposed to the application, with the unexposed group serving as a control group. DISCUSSION/REFLECTION/LESSONS LEARNED: Although educational interventions to encourage cost-effective care among resident physicians have been developed before, this Android application is an innovative tool aimed at incorporating cost-consciousness into the initial work-up of patients admitted to the hospital without compromising thorough medical evaluation. Utilizing this tool during morning report will remain feasible since it does not involve a major change from current practices and it can become part of curriculum for residents on inpatient medicine service. Ultimately, the goal is to extend the use of this application in routine day-to-day patient care on the medicine wards and continue to teach costcontainment to medicine and medicine-pediatrics residents. MOBILE-CENTERED SPACED LEARNING: A NOVEL APPROACH TO GRADUATE MEDICAL EDUCATION IN THE DIGITAL ERA Robert J. Walter; Anantha K. Mallia; Ali Massoumi; Joshua Hartzell. Walter Reed National Military Medical Center, Bethesda, MD. (Tracking ID #1938605) NEEDS AND OBJECTIVES: Developing an optimal learning environment and curriculum in graduate medical education can be challenging, particularly in light of progressive work hour restrictions limiting the feasibility of traditional academic didactics that compete with patient care duties. Emerging literature has demonstrated the effectiveness of spaced learning within medical education as a viable adjunct to conventional pedagogy. This method of learning has been found to improve knowledge acquisition, increase long-term knowledge retention, and improve a learners ability to accurately self-assess their knowledge. However, spaced learning shares the same limitations as many other learning models with regard to accessibility, ease of use, and learner acceptance/adoption. Additionally, no studies to date have evaluated the usefulness of a mobile-based spaced learning curriculum within an internal medicine residency program. As such, the goal of our study was to assess the effectiveness and acceptance of a mobile-based spaced learning curriculum within a graduate medical education environment. SETTING AND PARTICIPANTS: Utilizing a mobile-based application, Qstream, that incorporates an adaptive spaced learning platform, we created an adjunct curriculum to supplement our existing academic didactics for housestaff during their critical care rotation at Walter Reed National Military Medical Center (WRNMMC). The curriculum consisted of 100 high-yield core knowledge questions within the area of critical care medicine. Each learner enrolled in the module received an email and/or RSS feed prompt daily to their mobile device of preference (smartphone, tablet) with 5 questions. Upon answering each question, a brief explanation of clinical reasoning was provided and the learner was anonymously shown their performance in comparison to their cohort. Questions that were answered incorrectly were asked again within 7 days until answered correctly twice, whereas questions initially answered correctly were asked again within 13 days. Questions answered correctly twice in succession (mastery of material) were retired. DESCRIPTION: We conducted a prospective, randomized, controlled study of 30 interns and residents rotating on the medical and surgical intensive care units at WRNMMC. All participants had their baseline knowledge assessed by a 25 question pre-test based upon the core knowledge topics within our spaced learning module. Sixteen learners were randomized to participate in the spaced learning curriculum (in addition to traditional didactics) and were enrolled within the module. Fourteen learners were randomized to the traditional academic didactics alone. Block randomization was used to minimize cross-communication between learners. All participants completed a 25 question post-test at the conclusion of their 4 week rotation. Learners who were randomized to participate in the spaced learning module answered an additional 10 questions regarding the perceived effectiveness and usability of the module. EVALUATION: Our cohort consisted of 18 PGY-1s (60 %), 5 PGY-2s (17 %), 6 PGY-3s (20 %), and 1 PGY-4 (3 %). Baseline characteristics (PGY level) were similar between randomized groups. Performance on the pre-test examination was comparable (61 % +/ 18 % vs 68 % +/ 17 %, p= 0.26). Post-test performance was superior within the spaced learning group (86 % +/ 11 % vs 71 % +/ 13 %, p = 0.002) as was the percent improvement from pre-test to post-test (18 % +/ 12 % vs 7 % +/ 13 %, p = 0.03). The spaced learning module received favorable reviews both in terms of overall educational benefit (average rating 4.44 on a 5-point Likert scale) and as an adjunct to traditional didactics and bedside teaching (average rating 3.56 on a 4-point Likert scale). The average time to complete the 5 daily questions was 4.25 min. The overall difficulty and number of questions administered daily were designated as appropriate by 81 % and 87 % of learners, respectively. DISCUSSION/REFLECTION/LESSONS LEARNED: In our cohort, we observed a statistically significant improvement in overall post-test performance within the spaced learning group. Furthermore, acceptance and perceived usefulness of the mobile-based platform was high among learners. This novel approach to traditional didactics may be useful as a means to supplement existing curricula. Additional measurements, such as in-service and ABIM exam scores, gathered from larger, multi-center trials are needed to confirm the overall utility of a spaced learning curriculum as an educational adjunct within graduate medical education. NOT TOO MUCH, NOT TOO LITTLE, BUT JUST RIGHT: GOLDILOCKS ROUNDS AT THE CAMBRIDGE HEALTH ALLIANCE Rachael E. Bedard1,2; Alison B. Rapoport1,2; Maren Batalden1,2. 1Cambridge Health Alliance, Cambridge, MA; 2Harvard Medical School, Boston, MA. (Tracking ID #1940038) NEEDS AND OBJECTIVES: The AGCME identifies Systems-Based Practice (SBP) as a core competency for internal medicine residents; sub-competencies described in this domain include providing cost-effective care, recognizing system errors and advocating for systems improvement. Through the new Clinical Learning Environment Review program, the ACGME expects training institutions to engage residents in efforts to ensure that their delivery system is providing safe, high quality, evidence-based care to patients. With these mandates, training programs across the country seek to build meaningful curricular opportunities for residents in this domain. The Goldilocks Rounds at the Cambridge Health Alliance provides an innovative forum for Internal Medicine residents to discuss cases involving the overuse, underuse and misuse of healthcare resources and to design and implement small tests of change. The objectives of these rounds are as follows: - To provide a forum for honest disclosure and discussion of clinical errors and system failures - To encourage residents to identify examples in their own practice of the overuse or inappropriate use of healthcare resources -To help residents articulate one improvement oriented action item from each rounds that residents have the power to implement in real time SETTING AND PARTICIPANTS: The Cambridge Health Alliance (CHA) is an integrated public healthcare system comprised of two community hospitals and an extensive primary care network. The internal medicine residency has 8 residents in each of the PGY-1, 2 and 3 years; there are an additional 8 psychiatry interns and 7 transitional-year interns who also train on one 32-bed inpatient medicine teaching unit. All residents are invited to monthly Goldilocks Rounds. The two chief medical residents facilitate rounds. Hospitalist teaching faculty are present to offer guidance; an open invitation is extended to nursing staff. DESCRIPTION: Goldilocks Rounds are scheduled during lunch hour at regular intervals throughout the year. Housestaff come to rounds prepared with recent stories from their own practice that exemplify the overuse or misuse of medical resources, situations in which a clinical error was made, or a system failure occurred. An unnecessary blood draw, a misunderstanding with a nurse: Trainees are encouraged not to accept these issues as byproducts of a complex health system but to identify them as areas for improvement and come up with solutions. During Goldilocks Rounds, participants select 12 pressing problems raised in the case discussion, brainstorm potential interventions, and commit to taking some individual or group action. Facilitators help formulate discrete, simple, positive interventions that are within the scope of the residents to enact. When applicable, chief residents lead the group in filing a formal occurrence report to model the importance of reporting errors. Residents are the primary physicians on this dedicated inpatient teaching unit; they interact consistently with the same group of nurses and ancillary staff and can exercise genuine leadership in changing the practice environment. Action items that have resulted from Goldilocks Rounds during this academic year include patient education about how to ask for PRN medications in the hospital and improved end of shift signout protocols between day and night resident teams. EVALUATION: Prior to the first Goldilocks Rounds all house officers were invited to complete a survey on their attitudes and perceived self-efficacy regarding SBP and quality improvement. Sixty percent of residents responded. The same survey will be administered at the end of this academic year to assess change in attitudes and selfefficacy in these domains. Residents are informally invited to offer feedback on Goldilocks Rounds to the chief residents throughout the year. Finally, the chief medical residents follow up each rounds with a synopsis of issues discussed and proposed interventions. Chief residents audit the implementation and efficacy of those interventions going forward, unless a specific resident champion is identified. DISCUSSION/REFLECTION/LESSONS LEARNED: Baseline survey results revealed that 43 % of respondents report feeling very comfortable discussing a medical error with a fellow housestaff. Only 21 % are very comfortable filing a formal occurrence report. Only 22 % currently feel very comfortable raising issues about healthcare costs and use of medical resources with their attending physician. Despite this, 72 % of respondents report that they strongly agree with the statement My role as a physician includes a responsibility to improve the system of care in which I practice. Certainly this enthusiasm for systems-improvement work is evident in the Goldilocks Rounds discussions where trainees are eager to volunteer experiences and generate improvement suggestions. The Goldilocks Rounds are a promising strategy for bringing systems-based practice into the residency curriculum in a practical, relevant way that impacts both resident learning and patient care. ONLINEMEDED: MOVING MEDICAL KNOWLEDGE OUT OF THE CLASS ROOM AND ONTO THE INTERNET Dustyn E. Williams. 1Medical Education Online L3C, New Orleans, LA; 2Tulane University, New Orleans, LA. (Tracking ID #1896418) NEEDS AND OBJECTIVES: There is much to learn in medical school. The shift in recent years from primarily medical knowledge to promoting a more well-rounded physician has not lessened how much knowledge must be taught, only burdened medical schools and students alike; compressing more into less time. The need, then, is a means by which the medical knowledge component of medical school can be more efficiently delivered to medical students. The goal of this project was three fold. First, it was to consolidate and filter a vast amount of medical knowledge into the medical student level, as defined by the NBME and USMLE. Second, it was to create a method for dissemination that was both cost-effective and immediately scalable. Third, it was to introduce a mode of instruction that could prove fruitful for standardized testing examination as well as clinical practice, but in a completely novel way using the internet. SETTING AND PARTICIPANTS: Onlinemeded.org is a non-profit website that is available to everyone for free. Anyone is permitted to participate. There is no enrollment, no signup, and no cost. The only barrier is an internet connection that can support online streaming video. DESCRIPTION: 250 online videos, each approximately 15 min long, are displayed in a one-stop-shop access website. There are 19 categories that span across the core clerkships for the third and fourth years including neurology, psychiatry, medicine, surgery, pediatrics, and the subspecialties. Each page is dedicated to a single video and includes key concepts and errata, a discussion board, and an embedded YouTube Video. Each video is constructed to educate to a medical student, at the medical student level, including the use of advanced organizers, algorithms, and memory tools. EVALUATION: Google Analytics and YouTube Dashboard Manager were used. Each month over 6,000 unique visitors visit the site and over 24,000 videos are watched. The demographics are equal between men and women, and are primarily from the United States, with centers of use clustering around major cities: New Orleans, New York, and Chicago. Male and female users are equal. Feedback from users has revealed a tremendous amount of anecdotal success from the use of the site. Students comfort level on the wards increases, satisfaction in education increases, and scores on the USMLE Step 2 have shown substantial improvement from their Step 1 scores DISCUSSION/REFLECTION/LESSONS LEARNED: The demand for a resource like this is overwhelming as shown by the sheer number of users. The cost to maintain such a site is negligible; hosting videos on YouTube is free, and the cost of data for the actual site is minimal. The time cost to develop, film, and edit these videos is the primary barrier. However, once completed, they exist indefinitely until updates are required. This system provides a highly effective way of delivering medical knowledge content to medical students at extremely low cost, that is immediately scalable to every medical student in the world, and in a time-efficient manner, allowing medical schools to focus on the practice of medicine and cultivating future physicians rather than the emphasizing the basics of medical knowledge. ONLINE RESOURCE URL (OPTIONAL): www.onlinemeded.org OUTPATIENT MORNING REPORT IN THE EHR ERA: AN OPPORTUNITY TO APPLY ADULT LEARNING THEORY TO THE AMBULATORY CUR RICULUM James OKeefe. University of California, Los Angeles, Los Angeles, CA. (Tracking ID #1928226) NEEDS AND OBJECTIVES: The ambulatory curriculum for internal medicine residents is distributed across several venues, of which the Outpatient Morning Report (OMR) provides a unique opportunity to learn from cases from the resident clinics. Our objective was to enhance the learner-centered format of OMR by identifying multiple resident cases organized around a central theme for each session. SETTING AND PARTICIPANTS: OMR is a 45-minute conference that is attended by 25 residents (PGY2/PGY3) on ambulatory rotations. The conference is led by a chief resident and attended by a faculty preceptor. DESCRIPTION: The assigned resident chooses a case and notifies the chief resident of the topic. The chief resident performs an EHR search to identify similar cases from the clinics of residents who are attending the conference. Our EHR (Epic Systems) generates reports, with search criteria including demographics, diagnoses, medications, tests ordered, vital signs and lab values. Cases are printed and the relevant content is highlighted. During OMR, the assigned resident has 20 min to present their case(s) and lead a group discussion. Next, the chief resident distributes the surprise cases to individual residents, who are asked to present the highlighted portions of their cases to the group. The chief resident leads a nonjudgmental discussion of learning points from each case, provides an overview of relevant articles/resources, and allows for commentary by the faculty preceptor. Example: The assigned resident presented three patients for whom advance directives could have been (but were not) discussed in clinic: a 34 year-old with relapsed AML, a 76 year-old with dementia, and a healthy 57 year-old. The resident reviewed relevant articles from AFP and JAGS. Next, the chief resident led a discussion of surprise cases comprising several patients in which DNR status was documented in clinic but not acknowledged upon subsequent hospital admission. The group focused on potential QI projects to improve code status visibility in the EHR. EVALUATION: Conference topics are tracked in a spreadsheet to avoid duplication. High-yield topics are identified and repeated across groups of residents. DISCUSSION/REFLECTION/LESSONS LEARNED: We have applied adult learning theory by using our EHR to identify cases for OMR. We emphasize the relevance of GIM topics by drawing cases from resident clinics; we identify knowledge gaps during the process of chart review; and we direct teaching to the needs of the learners. By splitting conference time between the resident and chief resident, we maintain audience interest and balance topic selection. We are also able to include innovative topics (e.g. cost conscious care or systems-based care) in our curriculum. An unanticipated benefit of the format is near-100 % conference attendance, as residents do not know if one of their patients is chosen as a surprise case. The primary challenge is preparation time, which used to search for surprise cases (12 h) and prepare content (12 h). The amount of time has been lessened by asking residents to pick topics within the area of expertise of the guest faculty; this improves faculty engagement and reduces the amount of time for the chief resident to prepare content. A second limitation is that topics must be drawn in the resident practice and our EHR search for certain topics (e.g. refugee health) have yielded insufficient cases for a conference. PACT ICU - INTERDISCIPLINARY CARE CONFERENCES FOR HIGH RISK PRIMARY CARE PATIENTS William G. Weppner1,2; Janet V. Willis2; Amber Fisher2; Adam Brotman2; Rick Tivis2; Timothy Gordon2; C. Scott Smith1,2. 1University of Washington, Boise, ID; 2Boise VAMC, Boise, ID. (Tracking ID #1938812) NEEDS AND OBJECTIVES: Workplace interdisciplinary collaboration is recognized as an important part of quality primary care and favorable patient outcomes. However, many health professions training programs provide inadequate opportunities for trainees to develop the knowledge, skills, and attitudes needed for this. Interdisciplinary Case Conferences can address this gap by providing opportunities for participants from multiple professions to learn with, from and about each other through discussion of real patient cases and actionable strategies for care plans. SETTING AND PARTICIPANTS: The Boise VAMC Centers of Excellence in Primary Care Education developed the PACT ICU (Patient Aligned Care Team Interdisciplinary Care Update) to improve care of high risk patients in primary care. Based at our teaching clinic of approximately 32 medical residents and other trainees serving more than 2,000 patients, this offers a conference to develop coordinated, patient-centered care plans and reinforce principles of interdisciplinary collaboration. DESCRIPTION: On a weekly basis, rotating physician and nurse practitioner trainees lead discussion of a high risk patient selected from the top five highest risk patients as predicted by the VA-based Care Assessment Needs registry. A clinic-based registered nurse uses a web-based schedule and guide to facilitate selection and review of patients. The nurse and a physician lead coordinates an hour-long conference with interdisciplinary team members, including medicine residents and attendings, nurse practitioner trainees and supervisors, primary care nurses, pharmacists, psychologists and behavioral health staff, social workers and chaplain service. EVALUATION: Over the first 8 months of the PACT ICU, 27 trainees have participated, presenting 40 patients with 160 controls. There has been consistent involvement by different disciplines and trainees; surveys of 27 PACT ICU trainee participants suggest effectiveness at teaching PACT principles within a clinic team. The response to the statement My understanding of all of the elements (biological, social psychological) that must be considered in this patients care increased from 2.8 to 4.6 out of 5 (n=27, p<0.001). The response to the statement My understanding of the roles that each of the team can play in hard to manage patients like this one increased from 3.0 to 4.5 (p<0.001). Chart review of consult patterns suggests an increase in referrals to other members within the PACT team, such as pharmacy and behavioral health. Among patients that were presented in the PACT ICU, consults outside the team were 18 % lower than controls and consults to team members (pharmacy, psychology, social work, and RN case manager) increased. This consult pattern is emerging across all patients, suggesting generalizability of interprofessional collaboration. Using casecontrol methods, data from the first 6 months of this pilot indicates a non-significant trend towards decrease in patient utilization when compared to controls in the PCPs panel with similar risk scores for ER visits (0.94 vs. 1.20 per patient per year, p=0.48) and hospitalization (0.34 vs. 0.49 per patient per year, p=0.38). DISCUSSION/REFLECTION/LESSONS LEARNED: These findings suggest that PACT ICU has reinforced key patient-centered medical home concepts within an academic primary care setting, improving trainee understanding and referral patterns within the team. Overall, an interdisciplinary care conference such as PACT ICU can be a useful model for (1) teaching interdisciplinary care, (2) facilitating team-based care, and (3) improving patient consult management. ONLINE RESOURCE URL (OPTIONAL): Online Care Assessment Needs registry (VA-intranet based risk prediction tool) Online Care Management Tool 3.0.0 (VA- intranet based care management tool) Online PACT ICU schedule & tools: http://moodle2.boisevacoe.org PANEL MANAGEMENT IN ACADEMIC PRACTICES: WHAT IS IT THAT WE REALLY NEED TO TEACH? Kate Dube1; Rachel Willard-Grace1; Janhavi Athavale2; Thomas Bodenheimer1; Reena Gupta3. 1University of California, San Francisco, San Francisco, CA; 2Eastern Virginia Medical School, Norfolk, VA; 3University of California, San Francisco, San Francisco, CA. (Tracking ID #1939756) NEEDS AND OBJECTIVES: Panel management entails identifying and facilitating needed chronic and preventive care for a population of patients using clinical guidelines. In high-performing primary care practices, panel management is usually conducted by non-clinician members of the team, often medical assistants (MAs) or nurses who work closely with the clinician, relieving some of the routine patient care burden. Little is known about how panel management services are structured in teaching practices, which are charged not only with providing high quality care for patients, but also with teaching their residents requisite clinical and interdisciplinary leadership. We sought to understand how panel management is implemented in teaching practices. SETTING AND PARTICIPANTS: We conducted site visits at ten primary care residency clinics. Data was collected using a structured site visit guide and semi-structured interviews with clinic leadership, providers, trainees, and clinic staff. Site visit reports were coded and analyzed via matrix by two independent researchers through an iterative process. The research team reviewed these analyses to identify themes. DESCRIPTION: Two primary panel management models emerged: 1) Residents provide panel management to their assigned patients; or 2) Medical assistants and other allied health professionals provide panel management for the residents patients. The first model facilitates learning clinical guidelines, and it provides residents with an understanding of the process of panel management, such that they would understand what their MAs or nurses were doing as they entered the workforce upon completion of their training. On the other hand, the second model reinforces a teambased model of care that is more compatible with many contemporary primary care practices, and it provides an opportunity to build resident skills in leading a team. One site implemented a promising hybrid model that addressed both training residents in clinical skills and in team-based care. In this model, the residents and medical assistants share the panel management responsibility, work together to track patients who are in need of care and conduct outreach calls together. EVALUATION: Additional research is needed to understand how to define best practices and what patient and resident outcomes are associated with each model. DISCUSSION/REFLECTION/LESSONS LEARNED: Faced with the dual missions of training residents both in necessary clinical skills and in interdisciplinary team-based care models, teaching clinics have alternately entrusted panel management responsibilities to residents or MAs and nurses. A hybrid model, in which residents work collaborative with MAs or nurses, may provide a useful mode to bridge these models and achieve these dual missions. PAP EMPHASIS CLINICS: AVENUE TO IMPROVE CANCER SCREENING RATES AND TRAIN RESIDENTS IN WOMENS HEALTH Michelle Cleeves; Rebecca Hanratty; Sarah Christensen; Jennifer Adams. Denver Health and Hospital Authority, Denver, CO. (Tracking ID #1938721) NEEDS AND OBJECTIVES: Needs and Objectives: Though cervical cancer is easily preventable with regular screening, there are roughly 12,000 new cases of advanced cervical cancer annually in the United States. A significant cause of this failure to screen is that women are not offered cervical cancer screening by their primary care doctors, and studies have demonstrated this is especially likely when those care providers are recently graduated from residency. Though the ACGME and ABIM both require training and competency in preventive care counseling and specifically in delivery of pap smears, a deficiency persists in the provision of cervical cancer screening. By creating a pap emphasis clinic with specific curriculum and focused patient care for our residents, we aim to increase resident exposure to pelvic exams and cervical cancer screening with the goals of improving both resident education and cervical cancer screening rates in an urban underserved resident training clinic. SETTING AND PARTICIPANTS: Setting and Participants: PGY1, 2, and 3 internal medicine residents are scheduled to participate in the pap emphasis clinics at their continuity clinic site once during a 6 month intervention period. DESCRIPTION: Description: 14 patients are scheduled per session and residents perform pelvic exams, pap smears, STI testing, and family planning counseling as indicated. Residents also received a curriculum delivered by a GIM faculty member with expertise in womens health which includes review of pap screening guidelines, pathophysiology of cervical cancer development, preventive health care screening guidelines, and wet prep microscopy during their pap clinic. EVALUATION: Evaluation: We are assessing change in attitudes and medical knowledge with a 10-question survey administered to each resident before and after the curriculum is delivered. Topics contained in the survey included cervical cancer screening guidelines, pap smear results interpretation, and slide identification of common wet prep pathology. All 15 residents with continuity clinic at the pilot site will be scheduled to participate in the pap emphasis clinics throughout the study period. The session begins with a brief didactic session followed by direct patient care. We conclude the session with microscopy review of wet preps and a post-test. Although this project is ongoing, we have preliminary analysis demonstrating significant improvement in resident comfort with pelvic exams and medical knowledge in the areas tested. The residents also report high levels of satisfaction with the curriculum. At the same time, the pap rates increased in the pilot clinic sites demonstrating an impact on patient care. DISCUSSION/REFLECTION/LESSONS LEARNED: Discussion/Reflections/ Lessons Learned: Though cervical cancer screening is effective in prevention of invasive cervical cancer, there continue to be gaps in screening rates. Previous studies have demonstrated this is due in part to poor rates of physician implementation, especially among recent residency graduates. By creating a pap emphasis clinic, we were able to increase access to care for our patients and provide targeted teaching of pap smears and other preventive care topics in womens health, a skill set which will be carried forward into independent clinical practice. PATIENT SAFETY AT THE MOVIES Chaitanya Mandapakala1,2; Diane L. Levine1,3. 1Detroit Medical Center, Detroit, MI; 2John D Dingell VA Medical Center, Detroit, MI; 3Wayne State University, Detroit, MI. (Tracking ID #1938570) NEEDS AND OBJECTIVES: In the Institute of Medicine landmark report from 2001, To Err is Human and Crossing the Quality Chasm, the importance of teaching principles of quality improvement and patient safety throughout the continuum of undergraduate, graduate, and continuing education was emphasized. The Association of American Medical Colleges further recommends integrating these concepts as early as the first year of medical school. Efforts are presently underway to recognize important attributes students should possess at graduation from medical school called Core Entrustable Professional Activities for Entering Residency (CEPAER) for all entering residents to perform without direct supervision on day one of residency, regardless of specialty. Among these activities the ability to identify system failure and contribute to a culture of safety and improvement has been deemed as being critical. Wayne State University is the largest single campus medical school in the USA with 300 students per class. Third year medical students are assigned to a 2-month Internal Medicine (IM) clerkship at one of eight clinical sites. This makes teaching patient safety challenging. This year we initiated a patient safety half-day workshop to provide an opportunity for students to learn the principles of quality improvement and patient safety during their IM clerkship. Our medical school was very supportive of this endeavor, but was unable to provide financial support. So we developed a hands-on-experience utilizing low cost easily purchasable episode of a popular television show. SETTING AND PARTICIPANTS: Students on the clerkship were required to attend a half-day patient safety workshop in the middle of their clinical rotation unless they had their continuity clinic. DESCRIPTION: A pre-test was administered at the start of the workshop to test 7 basic concepts of quality improvement and patient safety. The session began with a case based presentation (morning report) providing an opportunity to discuss critical and latent medical errors, situations where errors are more probable while emphasizing concepts such as importance of proper transition of care and medication reconciliation. Students were encouraged to participate and identify how these errors could be avoided. Using a variety free resources demonstrating quality improvement from industry and personal stories the curriculum was delivered. Cases were examined to understand how the quality and patient safety department analyzes using various quality improvement techniques to strengthen the system. The high point of the day was the showing of a Greys Anatomy episode (cost $2.99) entitled I saw what I saw. Popcorn was distributed and lights were dimmed; students enjoyed the episode and participated in making a process map and a root cause analysis based on the events depicted in the episode. Possible solutions and PDSA cycles were described to further understand these principles. A post-test asking the same questions as the pretest was completed at the end of the workshop. Individual pre- post-test scores were analyzed. EVALUATION: To date three workshops have been conducted. A total of 132 medical students participated. Of the 132 students who attended, 120 (90.9 %) students completed pre and post tests. Analysis of pre-tests showed that 75 % of students were able to give examples of situations where medical errors are more probable. The majority of students (7599 %) showed knowledge deficits in the remaining concepts tested. All students showed knowledge improvement in all quality improvement and patient safety concepts discussed through this workshop (Swiss Cheese Model of Errors - 97.78 %, Situations where medical errors have higher probability to occur - 100 %, Sentinel Events - 98.75 %, construction of fishbone diagram - 97.48 %, RCA - 95.37 %, PDSA - 86.44 %). Based on feedback received after the first two workshops, the concept of human factors engineering was discussed in more detail during the third workshop with knowledge improvement of 96.42 %. Students also recorded 129 medical errors observed during their first month of clinical clerkship which were then classified into 10 categories. Medication errors (42 %) were the most common type. Interestingly only 1 student reported lack of hand hygiene as a patient safety concern. Finally, at the end of the workshop, 100 % of the students either agreed or strongly agreed to I am knowledgeable about patient safety. Qualitative analysis of the student comments indicated that the students found the workshop to be enjoyable and informative. DISCUSSION/REFLECTION/LESSONS LEARNED: The approach used targeted integrating quality improvement and patient safety thinking into daily workflow. This half day patient safety workshop has proved to be an accessible, feasible, inexpensive and successful way to introduce junior students to the concepts of patient safety and quality improvement. ONLINE RESOURCE URL (OPTIONAL): http://www.amazon.com/I-Saw-What/ dp/B003TTCZNO PATIENT-CENTERED MEDICAL HOME FOR SICKLE CELL DISEASE: IMPACT ON UTILIZATION OF ACUTE CARE SERVICES Temeia Martin; William P. Moran; Patrick D. Mauldin. Medical University of South Carolina, Charleston, SC. (Tracking ID #1940183) NEEDS AND OBJECTIVES: Patients with sickle cell disease (SCD) have a lifelimiting illness with a chronic relapsing course characterized by painful crises and recurring need for acute care for treatment. We sought to create a SCD outpatient medical home at the Medical University of South Carolina (MUSC) including acute care, individual care plans for each patient including acute treatment protocols and chronic medication regimens, and measure impact on emergency department (ED) visits and hospitalizations. SETTING AND PARTICIPANTS: University Internal Medicine (UIM) is an NCQA-certified level III PCMH and hospital-based clinic responsible for the care of almost 10,000 medical patients with diverse medical conditions, comorbidities, and behavioral risk factors. Half of the 30,000 annual patient visits occur in the Internal Medicine resident continuity clinic, and half in the contiguous faculty practice with the majority of SCD patients assigned to the resident clinic. We defined our SCD medical home population as a patient seen 2 more times in the UIM practice over the preceding 3 years. DESCRIPTION: An interdisciplinary team was brought together to provide focused care to SCD patients by way of a patient-centered medical home. Patients were tiered according to their patterns of utilization for enrollment in the SCD medical home. We sent letters asking them if they would like to transfer their care from the UIM practice to the SCD medical home and a follow-up phone call was made. Patients who consented attended a 1 h intake appointment and were scheduled for monthly visits thereafter. Follow-up visits were used to address issues related to SCD, co-existing chronic medical conditions, and health care maintenance. Patients were also given access to acute daily treatment of their pain with intravenous narcotics and hydration. We developed a referral process for medical home patients who presented to the ED. Plans of care for treatment of acute and chronic pain were individualized. A pharmacy clinic was developed for management and monitoring of Hydroxyurea usage in the patient population. We developed a transition plan with pediatrics to support matriculation to adult care. The population included in the SC medical home increased from a pilot of 10 patients to 78 over the first 20 months. The interdisciplinary team met weekly to coordinate ongoing development strategies for the clinic. EVALUATION: Baseline data demonstrated that SCD accounted for 1 % of the UIM practice and accounted for 15 % of ED visits, 10 % of hospital admissions, and 33 % of re-hospitalizations. Measures of success were utilization of ED and hospital, Hydoxyurea prescribing, and patient satisfaction. Three patients died before the program started and 2 have died since its initiation. Interrupted time series (ITS) prepost comparison design was used to assess change in utilization and data were limited to MUSC system. Statistically significant decreases in ED utilization (p < 0.0001) were achieved for the top 40 utilizers within the UIM PCMH Sickle Cell clinic following implementation of the clinic. Because of the small sample size, large fluctuations did not allow for significance to be determined for hospitalizations or rehospitalizations. To determine where the effect on ED utilization of the sickle cell clinic was the strongest, subgroup analyses were performed: top 7 utilizers, next 16 utilizers, and then the remainder of the top 40. Significant decreases in ED utilization were achieved for the top 7 utilizers (p = 0.0006), with the most significant decreases within the next 16 utilizers (p < 0.0001). There was no increase in Hydroxyurea prescriptions from baseline although there was an improvement in monitoring for toxicity and dose adjustments to target goals. Patient satisfaction data is gathered through qualitative and quantitative mechanisms including surveys and focus groups. DISCUSSION/REFLECTION/LESSONS LEARNED: Implementation of a patient centered model in SCD appears to have a positive impact on disease management. Greater access to acute and longitudinal medical services improves utilization of acute care services. Patterns of utilization correlate with severity of disease underscoring the need for enhanced models of care for SCD. ONLINE RESOURCE URL (OPTIONAL): NONE PRE-CLINICAL ELECTIVE TEACHES END-OF-LIFE COMPETENCIES: EFFECTS PERSIST INTO CLINICAL YEARS Melissa S. Keeport; Brian Cruz. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1927107) NEEDS AND OBJECTIVES: Medical education inadequately addresses end-of-life (EOL) and palliative care. We previously reported that our interventiona preclinical electiveincreased students self-perception of readiness to discuss EOLrelated issues following completion of the course. The first cohort of intervention students to complete a significant clinical year was assessed to determine if their readiness continued 2 years after taking the course. Our aims were: 1) teach medical students basic EOL care competencies with an emphasis on confidence-building for addressing these topics with patients and their families; and 2) measure the impact of this elective from the pre-clinical (MS-1 and MS-2) years to the clinical (MS-3 and MS-4) years. SETTING AND PARTICIPANTS: We designed an elective open to first-year medical students at Tulane University School of Medicine. We have implemented it for three consecutive years, starting with the Class of 2014 in their MS-1 year. This cohort has now completed its first and main clinical year of medical school. We administered a survey to the entire Class of 2014 upon completion of their MS-1 year and upon completion of their MS-3 year to assess EOL competencies. Course participants responses were compared to control students who did not take the elective. DESCRIPTION: The elective is open to 25 medical students for 10 total course hours. Creative, active-learning modules are combined with published tools and techniques to explore the basics of hospice and palliative medicine. Multiple domains are addressed using a combination of didactics, group discussion, standardized patients, attendance at the hospital ethics committee meetings, and site visits. The elective utilizes student leadership of MS-3 and MS-4 students, expertise of Tulanes palliative care specialist on faculty, education by internal medicine residents pursuing fellowships in palliative care, and other multidisciplinary specialists. EVALUATION: Medical students who completed the elective in their MS-1 year said they felt more prepared to speak to patients and/or patient families about EOL issues than their counterparts, with 78 % of elective students responding prepared compared to only 32 % of controls; p < 0.001. These responses persisted 2 years later into the clinical training years, where 70 % of MS-3 students who had taken the elective responded prepared compared to only 36 % of students who had not taken the elective; p < 0.05. DISCUSSION/REFLECTION/LESSONS LEARNED: The increased comfort and ability to speak about EOL-related issues that students achieved following their participation in the MS-1 elective has persisted into their clinical years. This studentinitiated elective exemplifies the usefulness and high impact of teaching clinicallyrelevant skills early in the pre-clinical curriculum. ONLINE RESOURCE URL (OPTIONAL): http://tmedweb.tulane.edu/mu/eol/eolelective/ PREPARING RESIDENTS FOR THE PRIMARY CARE MEDICAL HOME: SELF-DIRECTED, LONGITUDINAL PROJECTS IN QUALITY IMPROVE MENT Carol Sprague2,1; Christine Mullowney1. 1Oregon Health and Science University, Portland, OR; 2Portland Veterans Affairs, Portland, OR. (Tracking ID #1938533) NEEDS AND OBJECTIVES: A core principle of the primary care medical home is ongoing engagement in quality and safety activities. Resident training needs to prepare young physicians for this work. We therefore designed and implemented projects in quality improvement with the following objectives: 1) To familiarize residents with the concepts and strategies of the patient-centered medical home 2) To foster the development of creative, systems-based solutions to challenges in clinical care 3) To practice use of the electronic health record in capturing population data and supporting the delivery of evidence-based care SETTING AND PARTICIPANTS: The Internal Medicine residency at the Oregon Health & Science University (OHSU) is a university-based, mid-sized program. The OHSU ambulatory curriculum includes a primary care clinic experience (for which we have three sites) as well as ambulatory based rotations, one of which is the Chronic Illness Management or CIM rotation. In addition to a team-based clinical experience, the CIM rotation includes a 2 h weekly session to learn the principles of the medical home. Residents are scheduled for 6 of these sessions over a 6 month period; interns are given a 3 month period to work on a QI project. DESCRIPTION: The initial sessions include interactive lectures on the patientcentered medical home and the PDSA tool. Residents then apply these concepts toward a self-directed quality improvement project. In deciding upon a project topic, residents account for the structure and patient demographics of their primary care site. Residents develop an aim statement and think through each step of a PDSA. They use the medical literature for guidance about benchmarks. Residents then obtain the best performance data available within their clinic setting. Clinic teams are involved in the improvements when possible. Project development is achieved through an interactive, iterative experience in which residents return on a monthly basis to reflect on their progress and to receive feedback from peers. At the final session, residents present their quality project to their peers and are asked to reflect on how they might accomplish their goals with a larger patient panel. EVALUATION: Resident evaluations of the rotation are positive. Residents appreciate applying the QI concepts to their own patient panels and clinic settings, a process that in one residents words helped me realize the potential . . . to make significant changes in my practice. Developing projects that are appropriate in three different clinical settings is challenging but helps residents learn about different systems of practice. Last academic year, three residents were selected to present their projects at a Medicine department M&M. The flexibility of the project allows residents to develop skills within the context of their interests and in keeping with their stage of development as a physician. DISCUSSION/REFLECTION/LESSONS LEARNED: The quality improvement project in the CIM rotation provides residents with a longitudinal, self-directed learning experience. Quality improvement concepts seem to be learned well when residents are given flexibility in determining a topic that will be meaningful to them. Our current structure is restricted by each clinics capacity to incorporate significant quality projects and by our ability to access accurate data in each setting. Ideally, for those residents with a particular interest, we would propose the development of a separate rotation that allows residents to contribute to clinic-wide, interprofessional efforts toward quality improvement. PROGRESS TESTING: AN INNOVATIVE APPROACH TO ASSESSMENT OF COMPETENCY-BASED MILESTONES IN CLINICAL SKILLS Lauren Block; R. Ellen Pearlman; Judith M. Brenner; Alice Fornari; Joseph S. Weiner; Marie Barilla-Labarca. Hofstra North Shore-LIJ School of Medicine, New Hyde Park, NY. (Tracking ID #1933537) NEEDS AND OBJECTIVES: Growing evidence suggests that medical student assessment should be developmentally sequenced, competency-based, individualized, and integrated across disciplines. Progress testing offers an efficient, individualized, comprehensive assessment across content areas and academic year. We sought to use a single standardized case as a progress test to compare developmentally-based clinical skills competencies among medical students. SETTING AND PARTICIPANTS: As the culmination of a longitudinal clinical skills curriculum, a single standardized case was administered in a clinical skills center to first and second year students at Hofstra North Shore-LIJ School of Medicine. DESCRIPTION: Students performed a history and physical examination on a standardized patient as well as a post-encounter exercise to assess competency in domains including communication, patient care, and professionalism. A committee of standardized patients and clinical faculty reviewed student performance based on proposed milestones expected for the students year in each domain. EVALUATION: 58 first year and 33 s year students completed the examination. 81 % (47) of first-year and 85 % (28) of second-year students met developmental expectations in communication, 90 % (52) and 94 % (31) in patient care, and 84 % (49) and 76 % (25) in professionalism. The 23 students who did not meet expectations were offered remediation through individual feedback, videotape review, and a repeat exam. The implications of these data will be discussed. DISCUSSION/REFLECTION/LESSONS LEARNED: A single standardized case served as a developmental progress test to identify students attaining milestones in clinical skills and highlight areas for feedback and remediation. Standardized patients and faculty cooperated to identify areas of weakness and offer individualized feedback to learners. As developmental milestones are proposed in medical education, it is critical to evaluate whether students meet the recommended competencies. We designed a single case to serve as an integrated clinical skills progress test. Future work will include prospective development of a competency-based instrument for longitudinal assessment in clinical skills. REACHING CONSENSUS AMONGST PHYSICIANS IN BUILDING AN EVALUATION TOOL TO IMPROVE DISCHARGE SUMMARIES AND OPTIMIZE TRANSITIONS OF CARE Karim Taha; David W. Frost. University Health Network (UHN), Toronto, ON, Canada. (Tracking ID #1936760) NEEDS AND OBJECTIVES: The discharge summary is an essential communication tool between hospitalists and primary care physicians. Despite advances in electronic transfer of discharge summaries as well as electronic templates, there remain challenges around optimization and standardization of its content. This study aimed to reach consensus amongst family physicians (or general practitioners, GPs) on the relative importance of various components of the discharge summary, and to determine the best areas to focus educational efforts. Based on these results, an evaluation tool for discharge summaries was developed for both educational and quality assurance purposes. SETTING AND PARTICIPANTS: Twenty-five family physicians (both academic and community-based) participated in our study employing the Delphi methodology of consensus building based on iterative surveying. This initiative was a partnership between the General Internal Medicine Division of the University Health Network in Toronto and family physicians in Toronto, Ontario and surrounding regions. DESCRIPTION: Three rounds of surveys regarding discharge summaries were completed. The first round allowed for family physicians to express open-ended opinions about the most salient aspects of discharge summaries. The second round asked participants to rank all opinions generated from the previous round in terms of importance and agreeability to their preferences. Finally, the last round addressed any outlier responses in the overall rankings. The culmination of these rounds served to establish a refined average in each discharge summary category. Findings: Specific instructions for the family physician in the follow-up section and a well-outlined discharge plan were areas of highest relevance to participants. These sections were also noted as least clearly filled out currently. Commonly reported areas requiring improvement included the need for a thorough medication reconciliation section, a concise overall report, and provision of specialist referral information. Participants expressed preference for inclusion of relevant lab results from the entire hospital stay duration, and suggestions were made on appropriate educational interventions to maximize discharge summary quality. EVALUATION: The tool developed will be subjected to an assessment of intended outcome (improved discharge summary quality), implementation and uptake levels, and determination of perceived value to educators and learners. Feedback generated will guide further revisions and dissemination decisions. DISCUSSION/REFLECTION/LESSONS LEARNED: Family physicians, as experts in the evaluation of discharge summary quality, purposely engaged with general internal medicine physicians in the successful development of an educational and quality improvement tool using rigorous consensus-building methodology. Deployment of the tool is currently under way in paper and electronic form. Evaluation of its utility in improving discharge summary quality is planned in the near future. With an almost 100 % response rate throughout three rounds of surveys, our ranked results are highly reflective of the collective opinions of partnering family physicians. RESIDENT AS PRECEPTOR: DEVELOPING SENIOR RESIDENTS AMBULATORY TEACHING SKILLS AS PART OF A RESIDENT AS TEACHER CURRICULUM Elena Lebduska1; Sudha Dubey2; Darlene LeFrancois1; Sheira Schlair1. 1Montefiore Medical Center, Bronx, NY; 2Beth Israel Medical Center, New York, NY. (Tracking ID #1937669) NEEDS AND OBJECTIVES: Fostering residents teaching skills can complement clinical skill development for internal medicine residents. While there are many opportunities for inpatient resident teaching skill development, opportunities to expand teaching skills may be underutilized in ambulatory settings. Our aim was to teach residents effective precepting skills in the ambulatory setting by providing a Resident as Preceptor curriculum followed by supervised precepting sessions with junior residents. SETTING AND PARTICIPANTS: Third year internal medicine residents on a Resident as Teacher elective, first and second year internal medicine residents, and attending internal medicine physicians in three ambulatory internal medicine sites at Montefiore Medical Center (Bronx, NY). DESCRIPTION: The Resident as Teacher elective is an intensive 2 week rotation focused on resident teaching skill development in which residents receive formal teaching skill instruction and are observed and given feedback in diverse teaching encounters. The Resident as Preceptor curriculum is a major focus of our Resident as Teacher elective. A two-hour learner-centered, multimedia curriculum focuses on adult learning theory, socratic method, the one-minute preceptor model, and RIME scheme. Following the didactic session, residents then precept for 12 half days. During these sessions the resident precepts a junior (PGY1 or PGY2) resident while an attending physician silently observes. All participants complete evaluation forms and give immediate feedback on the resident preceptors precepting skills. EVALUATION: Resident preceptors, junior residents and attending physicians are surveyed immediately following each precepting session to assess overall precepting effectiveness. Survey instrument outcomes are based on principles of the one-minute preceptor model and the RIME scheme with all questions scored on a 14 Likert scale (1 = strongly agree to 4 = strongly disagree) or open-ended questions. To date, 13 senior residents, 15 junior residents and 7 attending physicians have participated with the goal of 25 senior residents, 30 junior residents and 10 attending physicians having participated by Spring 2014. Of the initial participants, 66 % of residents and 62 % of attending physicians strongly agreed that the resident preceptors effectively elicited the junior residents reasoning, assessment, and plan. Seventy percent of residents strongly agreed that they were comfortable in their role as preceptor, and 80 % strongly agreed that the experience added educational value to their residency. Seventy-five percent of junior residents strongly agreed that they received behaviorally specific feedback. DISCUSSION/REFLECTION/LESSONS LEARNED: Interim analysis of survey and program evaluation data suggests that a learner centered curriculum focused on feedback and precepting skills, and observed precepting experience with immediate feedback on teaching skills, can add educational value to the residency experience. In addition, initial cohorts of resident preceptors seem to have been effective at applying lessons learned in the didactic portion of the curriculum, including principles of adult learning theory and learner assessment. Resident preceptors found it to be an overall rewarding experience. Incorporating resident precepting skill development into a Resident as Teacher curriculum can expand educational and assessment opportunities to the ambulatory setting. RESIDENTS JOIN THE HUDDLE: RESIDENT PARTICIPATION IN A QUALITY IMPROVEMENT PROJECT ON RN-MD COMMUNICATION DEVELOPS SEVERAL ACGME RESIDENT COMPETENCIES Rachel Hathaway1; Maren Batalden1; Priyank Jain1; Kathleen M. Clark2. 1Cambridge Health Alliance, Cambridge, MA; 2Cambridge Health Alliance, Cambridge, MA. (Tracking ID #1930039) NEEDS AND OBJECTIVES: Increased attention to developing resident proficiency in quality improvement has led to the inclusion of competency in quality improvement as an ACGME milestone. Quality improvement efforts emphasize interdisciplinary collaboration in clinical work; the capacity to work in teams is a competency required for the 21st century physician. This project involves residents helping to design and implement a structured, daily nurse-physician huddle on an inpatient teaching service to proactively communicate patient care plans at the start of each day. Resident leaders will be instrumental in structuring and evaluating the impact of this intervention. We intend to show that the active participation of Internal Medicine residents at Cambridge Hospital in a quality improvement project allows residents to meet the ACGME milestone of systematically analyzing their practice and engaging in practice improvement. Additionally, several other secondary competencies are strengthened through participating in this project. SETTING AND PARTICIPANTS: This pilot project will take place on a busy inpatient teaching service at Cambridge Hospital, a university-affiliated, community hospital in an urban, diverse, and largely underserved setting. The service is residentrun and exclusively geographically located on one floor of the hospital, allowing for consistent interactions between residents and a defined cohort of nurses. Participants in the nurse-physician huddle include 2nd and 3rd year residents, all floor nurses, the nurse manager, and the unit secretary. DESCRIPTION: Currently, MDs and RNs primarily communicate with one another via ad-hoc interactions. We intend to institute a daily, structured, and brief huddle between residents and nurses at the beginning of the day shift. The intent of this resident-led huddle is to address early in the shift any nursing or patient concerns, potential discharge complications, and to clearly communicate the patient care plan for the day to all relevant providers. Any patients identified as being sick or complex will prompt inclusion of the nurse on MD work rounds to ensure a cohesive team approach. All participants will be encouraged to be engaged and proactive in their roles. Summaries of each patient will be high-yield and concise. Real-time feedback will be solicited from both residents and nurses in order to rapidly improve care and anticipate problems. Objectives of the huddle include improved communication between residents and nurses, improved patient care, better teamwork, and improvement in both RN and MD satisfaction with their work environment. EVALUATION: There will be several tiers of evaluation. Residents will be quantitatively surveyed regarding their perceived growth in various ACGME competencies including quality improvement. Additionally, we will assess the value residents place on quality improvement and other related competencies as a result of participating in the project. Both residents and nurses will be evaluated regarding any perceived improvement in nurse-physician communication, teamwork, quality of patient care, and work environment satisfaction. Patient satisfaction will be surveyed via changes in Press Ganey scores. DISCUSSION/REFLECTION/LESSONS LEARNED: Finding meaningful ways for residents to develop competency in the realm of quality improvement is a priority. This is reflected in the current ACGME milestones. This quality improvement pilot project seeks to actively engage Internal Medicine residents at Cambridge Hospital in improving nurse-physician communication on a busy inpatient ward. Nurse-physician communication is improved by means of a structured, daily morning huddle to address active, patient-centered concerns. This resident-led initiative provides opportunity to develop and demonstrate competency in several ACGME milestones, including acting as patient advocates, striving for excellence in patient care, successfully participating in interprofessional teams, and practicing effective communication. Residents not only observe the inner workings of a quality improvement initiative but are critical players in the process. Residents will be encouraged to reflect on the impact that their work has had on their own development as leaders, on the culture of the unit in which we work and on patient care. RESTORING EMPHASIS ON AMBULATORY INTERNAL MEDICINE TRAINING - THE 3:1 MODEL. Joshua W. Harrison; Peter Cronkright; Astha Ramaiya. Upstate Medical University, Syracuse, NY. (Tracking ID #1910892) NEEDS AND OBJECTIVES: Inadequate ambulatory training in internal medicine (IM) residency programs may be partially responsible for the current shortage of primary care physicians. Without an adequate ambulatory care experience, IM residents fail to observe the rewarding nature of developing continuity relationships with their patients and do not gain the confidence and experience to be effective outpatient practitioners. With work hour demands and multiple reports citing concerns regarding ambulatory IM training, reform in this arena is necessary. SETTING AND PARTICIPANTS: A relatively novel 3:1 scheduling system that creates week long ambulatory blocks every fourth week was instituted for PGY-2/3 residents at our program. During this week residents participate in morning continuity clinic and afternoon subspecialty clinic training. Residents begin ambulatory days learning practice management issues, reviewing an ambulatory medicine article, covering a pharmacy topic, or reviewing a physical exam skill. Each quarter residents participate in a half day Teaching Activity allowing faculty to directly observe residents communicate with simulated patients, role play clinical teaching scenarios, or run simulated codes. DESCRIPTION: We hypothesized that this model would provide a more structured ambulatory educational atmosphere, better continuity of care, more exposure to subspecialty outpatient medicine, and eliminate conflict with inpatient duties. We hoped this would enhance resident satisfaction with ambulatory medicine and achieve our main goal of improving the ambulatory educational experience. EVALUATION: Surveys were created utilizing a Likert scale and distributed 9 months into this experience. Responses from PGY-1s maintained on a traditional day per week clinic schedule were compared to PGY-2/3 responses using a t-test. The PGY-2/3 group also retrospectively evaluated their experience in the 3:1 model compared to the prior years traditional schedule. These results were analyzed descriptively. Residents reported the 3:1 model had positively affected their satisfaction with residency training in general, their satisfaction with outpatient/ primary care training, and their outpatient/clinic educational experience. Residents perceived improvements in continuity of care with patients and in the quality of care they provided for patients. Their experience in ambulatory sub-specialty training was positive. PGY-1 s who remained on a traditional day per week clinic schedule were significantly less satisfied than PGY2/3 s with their ambulatory/primary care training (mean = 3.48 vs. 4.08, p = 0.01). Compared to PGY-2/3 s, PGY-1 s also perceived less continuity of care with their patients (mean = 3.91 vs. 4.38, p= 0.03), less adequate exposure to sub-specialty outpatient medicine (mean = 3.0 vs. 4.58, p 0.01), and were more distracted or worried about inpatient duties while at clinic (mean = 2.39 vs. 4.92, p 0.01). DISCUSSION/REFLECTION/LESSONS LEARNED: We present positive results following the institution of a relatively novel 3:1 scheduling model. This system effectively eliminates the concurrent demands of inpatient and outpatient duties, provides increased time for educational sessions, has improved resident satisfaction in several arenas, likely promotes continuity of care, and provides important exposure to subspecialty outpatient medicine. We believe these changes enhance our residents ambulatory educational experience. Given the resoundingly positive results the 3:1 model has since been instituted for all categorical residents at our program. SEALS: A PROCESS AND INTERMEDIATE OUTCOME EVALUATION OF AN INNOVATIVE PIPELINE CURRICULUM Cassandra D. Fritz; Monica Vela. University of Chicago, Chicago, IL. (Tracking ID #1936651) NEEDS AND OBJECTIVES: Pipeline programs are one proposed method to increase the number of under-represented minority (URM) students matriculating to medical school. Although pipeline programs have demonstrated success in increasing matriculation of URM students, a review conducted by the Department of Health and Human Services in 2003 notes the lack of appropriate evaluation measures for pipeline curriculum. Therefore, our goals were to 1) develop an evaluation mechanism using process and intermediate outcome measures for pipeline programs and 2) apply the evaluation to a novel pilot pipeline program. SETTING AND PARTICIPANTS: The SEALS summer pipeline program is a novel curriculum established to promote 5 competencies among minority medical students including: Socialization and Professionalism, Education in science learning tools, Acquiring finance literacy, Leveraging Mentorship and Networks, and Social expectations and resilience. Local academically successful URM and/or underserved rising first and second year college students were recruited to apply through the web. IRB approved pre/post program surveys included short answer, likert scale responses, and personal statement writing and were administered to the 13 SEALS students. In addition, this survey was administered to a cohort of ten students in a comparable, research-based pipeline program at our institution during the same summer session. This cohort served as a control. DESCRIPTION: The pre and post survey instrument included qualitative and quantitative measures designed to assess students knowledge, skills, attitudes, and quantify hours of programmatic activities. Students knowledge of (1) the medical school admissions process, (2) requirements for medical school, (3) AMCAS application service, (4) Health Care Disparities, and (5) Anatomy and Physiology of systems related to greatest health threats to underserved populations was assessed. Skills of how to (1) shadow a physician effectively, (2) seek out research opportunities, (3) engage and reflect upon service opportunities, (4) develop a personal statement of interest in medicine, (5) interview and communicate, and (6) effectively complete of a financial aid form were also assessed. The survey gauged the students attitudes toward the following: (1) service, (2) research, (3) medicine as a profession, and (4) commitment to medicine as a career. Finally, the total number of hours spent on each activity and students assessment of the quality of the activities was obtained. EVALUATION: Our evaluation consisted of process and intermediate outcome measures. Process measures included who participated, quantity of activities, and quality of those activities. Intermediate outcome measures quantified students change in knowledge, skills, and attitudes. The evaluation was also administered to the control cohort, who did not differ significantly in any demographic when compared to the SEALS cohort. Process Outcomes. One hundred percent of students agreed/ strongly agreed that the Meet the Professor series, Health Care Disparities series, and Pre-Med 101 workshops were instrumental in motivating students to purse their passion for medicine (11 of 11). Furthermore, these curriculum components inspired students to advocate for minority health issues. Intermediate Outcomes. The preSEALS to pre-control analysis did not reveal any important statistically significant difference between the two cohorts. A comparison of pre-SEALS to post-SEALS found statistically significant changes favoring post- SEALS: knowledge of MCAT components (p = 0.007), ability to ask for letter of recommendation (p = 0.04), attitudes toward research being an important determinant of success (p = 0.02) and appearing confident during an interview (p = 0.04). The post-SEALS to post-control did uncover a number of statistically significant values including, but not limited to ability to communicate with patients (p = 0.046), understanding of health disparities (p = 0.04), understanding of major disease entities (p < 0.01), and students resolve to become a physician (p < 0.01). Pre and post- SEALS personal statements were compared to the control groups AMCAS personal statement using a 5-point scale. The SEALS cohort average score increased by nearly 1 point. (pre = 2.09 to post = 2.95). These averages were then compared to the control cohort average, which was determined to be 2.40. DISCUSSION/REFLECTION/LESSONS LEARNED: Two major lessons about pipeline curriculum emerge. First, defining best practices in pipeline curriculum requires process and intermediate outcome measures. Outlining process and intermediate outcome measures of pipeline programs will allow for systematic improvement of program practices and dissemination of successful pipeline practices. Second, use of a control group revealed a gap in professional development training for URM students participating in research-based pipeline programs. SHARED DECISION MAKING IN CHRONIC CONDITIONS: A CURRICU LUM IN USING DECISION AIDS FOR INTERNAL MEDICINE RESIDENTS Leigh H. Simmons1,2; Lauren Leavitt2,1; Alaka Ray1; Blair W. Fosburgh1; Karen R. Sepucha2,1. 1Massachusetts General Hospital, Boston, MA; 2Massachusetts General Hospital, Boston, MA. (Tracking ID #1936903) NEEDS AND OBJECTIVES: In the current medical environment, there is heightened attention to training physicians in communication skills that will enhance patient understanding of treatment recommendations and reduce preference misdiagnosis. The aim of this study was to develop and implement an educational workshop for residents to promote shared decision making (SDM) with their patients during clinical interactions, with a specific focus on decision making for four chronic conditions (diabetes, depression, hypertension, and hyperlipidemia). SETTING AND PARTICIPANTS: We conducted our study in a large academic internal medicine residency; 120 residents (from all years of training) participated during their ambulatory rotation. DESCRIPTION: We used a mixed methods approach to understanding our residents perspectives on SDM, improving resident knowledge about SDM, and implementing a novel paper decision aid for resident training. We ran focus groups with residents to elicit challenges of decision making with patients for chronic conditions and to get feedback on a new tool, the Choice Report. The Choice Report is a single-page print decision aid that highlights six steps of shared decision making, including FAQs about treatment options. Using the focus group insights, we developed a workshop curriculum including the following: a didactic presentation about risk communication, introduction of the Choice Report to guide conversations about chronic conditions, and role-playing exercises for a chronic condition using the Choice Report. Participants were shown how to download the Choice Reports from the hospitals primary care-focused intranet (accessible by all clinicians). Participants completed surveys after the session and we tracked the number of times that Choice Reports were downloaded. EVALUATION: 120 residents attended the workshops over a 4-month period. The interns rated the workshop higher than the residents (81 % very useful content & 62 % excellent overall rating vs. Forty-six percent very useful content & 41 % excellent overall rating). This trend has been observed in our previous workshops, in which interns have generally found the content of shared decision making discussions more useful and rated them more highly than the 2nd and 3rd year residents. The Choice Reports on diabetes, depression, hyperlipidemia, and hypertension were downloaded 1500 times in the 8 months following their introduction. DISCUSSION/REFLECTION/LESSONS LEARNED: The workshop introduced residents to a SDM communication strategy for routine clinical encounters, with a focus on decision making for chronic conditions. This curriculum was novel in that it introduced a SDM tool to clinical topics not previously recognized as candidates for a shared decision making approach. The Choice Reports also generated significant interest from our general medicine clinicians and have continued to be used regularly in the course of routine primary care. We had initially planned to include a component of direct observation, with preceptors completing a mini-CEX of a visit using a Choice Report. However, we found that coordinating direct observations of eligible encounters was prohibitively difficult. Because this tool has been accepted and used by our staff and resident clinicians, we are developing new Choice Reports, to include cancer screening decisions, additional chronic conditions, and multiple language translations. NEEDS AND OBJECTIVES: Spanish speakers with limited English proficiency (LEP) represent a large and growing part of the population in the United States, with 80 % of hospitals providing care to LEP patients regularly. Having a languageconcordant provider is associated with improved patient satisfaction and improved processes of care. Previously described medical Spanish curricula have improved language skills among students and practicing physicians, but these curricula are incompatible with resident schedules. In 2009, a needs assessment at our institution showed that residents saw Spanish-speaking patients regularly and wanted a medical Spanish curriculum designed for residents. Objective: 1) Demonstrate improvement in medical Spanish for a diverse group of residents 2) Improve cultural competency SETTING AND PARTICIPANTS: Our institution is a 1,541 bed tertiary-care hospital with over 500 resident physicians, in a city of 129,585 people, of whom 27.4 % self-identify as Hispanic. In 20102011, we created a pilot curriculum with 20 participants from internal medicine, pediatrics, and combined med-peds programs. The curriculum was well received, so in 20112012 we expanded our learner-base, offering the curriculum to residents in their second year or above from all programs at our institution. Participants also were required to have at least a conversational level of Spanish. We had 63 participants, 33 (52 %) of whom were from procedural specialties. DESCRIPTION: The year-long, self-directed curriculum contains nine month-long modules, with 3 months that residents can take off during particularly demanding clinical rotations. The curriculum includes practice with a variety of skills including grammar, vocabulary, reading, writing, listening comprehension, speaking and cultural competency, in keeping with the principles of adult second-language learning, which suggests that incorporating multiple learning modalities is most effective. Each month-long, systems-based module includes: a grammar/vocabulary activity; a health-related DVD program; patient cases with a Spanish-speaking tutor; and community practice sessions. The community practice sessions involve having participants prepare health education talks on various to practice with English as a Second Language (ESL) students at a local non-profit, getting the chance to practice vocabulary, speaking and listening comprehension skills, while improving understanding of cultural concerns that would arise from the patient population that our residents serve. EVALUATION: We conducted pre-intervention and final assessments of language skills and cultural competency using previously validated self-report questions. We also collected open-ended feedback from participants and tutors. Of 63 participants, 13 (21 %) responded to the final survey. Of those, nine (71 %) were satisfied or very satisfied with the curriculum, but only five (38 %) completed a majority of the modules. Every respondent cited time as a barrier. There were correlations between number of modules completed and self-reported fluency scores (p = 0.025), change in self-reported receptive language ability (p = 0.025), and self-reported understanding of the health-related cultural beliefs of Spanish speaking patients (p = 0.047). Residents in cognitive specialties were more likely than residents in procedural specialties to complete the post-course evaluation (p = 0.03). DISCUSSION/REFLECTION/LESSONS LEARNED: We developed and implemented a novel medical Spanish curriculum for residents in order to address the prevalent health disparities issue of communication with Spanish-speaking LEP patients. Residents at our institution recognized the need for specific training to improve language concordance and expressed a desire for a medical Spanish curriculum that fit their schedules. Our institution, recognizing the need, supported our intervention. Our study demonstrated that residents who completed more of the curriculum had better self-reported fluency scores and cultural competency skills. The majority of residents who provided feedback were satisfied with the YRSI and several were able to complete significant parts of the curriculum. However, while residents from multiple specialties were initially enthusiastic about the availability of the curriculum, they ultimately found it difficult to participate, primarily citing time constraints. For future participants, we plan greater clarity about time required, more documentation of completion of curricular materials, frequent reminders, and largegroup events to encourage a sense of team. We will also include more surgeryspecific activities. Ultimately, measurement of the curriculums impact on patient outcomes will also be important. While curricular revision is necessary, we believe that the implementation of a flexible, self-directed curriculum is a feasible way to incorporate Spanish language education into a traditional residency schedule, at our institution and others. ONLINE RESOURCE URL (OPTIONAL): http://www.yalepediatrics.org/education/residency/spanish/index.aspx NEEDS AND OBJECTIVES: Needs : Humanism in medicine is heavily stressed in undergraduate medical education, particularly within the first and second years of medical school. Yet, when residency begins, there is a gradual erosion of cognitive attention to humanism and the values that initially attracted physicians in training to the profession of medicine are set aside in favor of acquisition of knowledge, patient management, and efficiency. The reduction in duty hours, opportunities for continuity of care, and a relatively young faculty mentor group only accentuate the erosion of attention to humanism and lived expression of meaningful values. In addition, there is a need for physicians in training to be challenged to progress in clinical problem solving abilities along a continuum toward expertise, and, learning to tap into episodic memories of actual patients has been demonstrated to serve this progression. Objectives- 1. To revitalize the humanistic aspect of medicine within Internal Medicine residents 2. To encourage deep reflection and meaningful attention and focus on the primary values of medicine, which when lived frequently, translate to increased career fulfillment 3. To encourage the practice of episodic memory formation which is utilized in advanced clinical problem solving SETTING AND PARTICIPANTS: Setting & Participants The curriculum has been instituted in the St Josephs Internal Medicine residency program since 2010. It is a series of 34 conferences at the end of the academic year. Each third year resident actively participates in the storytelling DESCRIPTION: All residents anticipate and plan to present a patient or patients that touched them during their training experience. The patients they present are framed as those that will never be forgotten for the teaching and learning they inspired. Residents receive a list of learning objectives and a bulleted list of reflection prompts to aid their reflection process. The presentation format is left entirely up to the residents-they may use a Powerpoint or engage in narrative-story telling (the latter is encouraged). The audience includes multi-disciplinary teams from both the inpatient and outpatient care settings, hospital staff, administrators, educators, nurses, hospital leadership and medical students. Following the residents shared reflections and narrative, the audience is encouraged to comment and share their thoughts about the impact the story telling had, memories and learning recalled, and similar patient care experiences. These sessions are highly anticipated as a culmination/capstone to the residency training program. Resident participants as well as the audience engage in collective humanism and discussion of the value that form the core of medical professionalism. Residents also reflect on how the learning from their patient has shaped their decision-making and how they anticipate the memory will become significant in their careers as physicians. EVALUATION: The sessions are evaluated on the response that it generates in the residents and the audience. Both groups are asked to share in writing the value(s) that were evoked by the capstone presentation. Several of the essays have been published by residents in different venues. Audience comments and the narratives are collected and we plan to publish a collection of the best essays in reflection. We also plan to start a Story-Corp collection of the narratives to archive these collections. DISCUSSION/REFLECTION/LESSONS LEARNED: As the practice of medicine and residency training is transforming in response to the societies call for decreased duty hours, there is great danger of the loss of humanism among physicians in training due to the toll of rapid cycle change. In addition, amongst junior and even seasoned faculty, we are observing disillusionment and burnout leading to career attrition. This capstone curriculum is designed to re-invoke the humanism in medicine and bring together team members who may not see the patient but are invested in the success of the enterprise: leaders as well as physicians and the multi-disciplinary team STORYTELLING WITH INPATIENTS Katherine C. Chretien1,2; Rebecca Swenson3; Bona Yoon1; Ricklie Julian4; Jonathan Keenan2; Raya Kheirbek1,2. 1Washington DC VAMC, Washington, DC; 2George Washington University, Washington, DC; 3Dartmouth-Hitchcock Medical Center, Lebanon, NH; 4Montefiore Medical Center, Bronx, NY. (Tracking ID #1938337) NEEDS AND OBJECTIVES: Empathy is an essential component of patientcentered care and improves patient outcomes. A recent systematic review of interventions to teach empathy to medical trainees identified a need for future interventions to be built upon a foundation of relationship-centered care. Narrative competence can be defined as the ability to acknowledge, absorb, interpret, and act on the stories and plights of others. There are few published educational interventions that address empathy while involving actual patient-provider interactions, and/or address narrative competence. We sought to incorporate a brief experiential narrative medicine curriculum within the third-year medicine clerkship with the goals of developing narrative competence, practicing attentive listening, and stimulating reflection while providing patient-centered care for hospitalized patients. SETTING AND PARTICIPANTS: Single institution with third-year basic medicine clerkship students on a 4 week rotation and inpatients on the acute (non-ICU) medicine wards. DESCRIPTION: The narrative medicine curriculum consisted of 1) Introductory session where students were introduced to narrative medicine and practiced attentive listening to storytelling in pairs; 2) Patient activity where students elicited illness narratives from patients, attentively listened, wrote their version of the story, and then read these back to patients. Students also asked patients to select a piece of art to help tell their story, chosen from a digital collection stored on an iPad. 3) Debriefing/ reflection session where students wrote reflectively and participated in a facilitated discussion about their experiences with the activity. EVALUATION: To evaluate and help assess acceptability and feasibility of the curriculum, five focus groups were held between July 2011 and March 2012. Thirtyone students participated. Transcripts were qualitatively analyzed for themes by two independent researchers; disagreements were resolved through discussion. Students discussed their patients reactions to the narrative medicine activity, students own experiences, the student-patient dynamic, and challenges. They also identified what they felt they learned, including seeing the patient as human, being open, and clinically relevant insights. Students felt that the activity improved the student-patient relationship and, in some cases, powerfully so. Response to the art component was mixed; for some student-patient pairs, the art the patient chose brought new insights into patient understanding. For other pairs, the technology and/or image library created barriers. Interviews with patients revealed generally positive response, although some patients did not recall the activity. Several patients asked their students to be their primary care provider. Comparison of audiotaped patient stories to the corresponding written narratives penned by students is ongoing to assess students narrative competence. DISCUSSION/REFLECTION/LESSONS LEARNED: A brief, experiential narrative medicine curriculum was feasible and acceptable to students and patients. From a time and resource perspective, the curriculum took 2 h of faculty time per month. Both patients and students appear to have yielded benefits - for patients, being attended to and being heard; for students, gaining deeper appreciation of the human side of medicine, of knowing patients better. The impact on the student-patient relationship was often positive, at times powerfully so. Ongoing focus groups helped to make continual improvements to the curriculum, including the retirement of the iPad art component and refinement of how the patient activity was structured to maximize comfort of both students and patients. TEACHING INTERPROFESSIONAL TRAINEES TO WORK IN CLINICAL CARE TEAMS: THE SAN FRANCISCO VA CENTER OF EXCELLENCE IN PRIMARY CARE EDUCATION Rebecca L. Shunk1,3; Bridget OBrien1; Maya H. Dulay1,3; Terry Keene2,4; Denise L. Davis1,3; Shalini Patel1,3; Melissa Bachhuber1,3; Calvin L. Chou1,3; Kristen Weaver2,4; Christina Kim2,4; Lynn OBrien2,4. 1UCSF, San Francisco, CA; 2UCSF, San Francisco, CA; 3San Francisco VA Medical Center, San Francsco, CA; 4San Francisco VA Medical Center, San Francisco, CA. (Tracking ID #1936850) NEEDS AND OBJECTIVES: Primary care clinics at the San Francisco VA (SFVA) transitioned to an interprofessional team-based model of patient care called Patient Aligned Care Teams (PACTs) in 2010. In 2011 the SFVA was awarded a Center of Excellence to develop and implement an innovative model of patient-centered, interprofessional education in a team-based ambulatory clinic. The core educational domains addressed in our center include sustained relationships, performance improvement, patient-centered communication, and interprofessional collaboration. SETTING AND PARTICIPANTS: In July 2011, we added trainee squads (2 s-year internal medicine residents and 1 s-year nurse practitioner student) to each of 8 preexisting teamlets (1 registered nurse, 1 clinical associate, 1 clerical associate) in three SFVA primary care clinics. Each trainee acts as primary provider for a personal panel of patients but performs cross-coverage for squad members. Teamlets and squad members discuss team patients in daily huddles or brief meetings along with other interprofessional trainees. DESCRIPTION: Curricular design for the core domains includes didactics, workplace reinforcement and reflection. For example, we set the foundation for interprofessional collaboration through interactive seminars on team members roles, conflict negotiation and huddling, then reinforce these skills in the workplace through daily huddles and interprofessional clinical practice. Team-building retreats provide opportunities to learn and practice skills together; to learn about each other both professionally and personally; and to reflect on teamwork, all requirements of highfunctioning teams. EVALUATION: To assess interprofessional collaboration, members of all trainee teams were asked to complete the Team Development Measure (TDM). The TDM is a tool created by Peace Health and used by the VA for PACT evaluation nationally scored on 0100 scale. Team members evaluate how their team is functioning in terms of cohesiveness, communication, role clarity and goals and means clarity. Team members complete the TDM early and again later in the academic year and have an opportunity to review and discuss their results during a facilitated session. Scores improved from early year to late year (Year 159.4 early/64.6 late) and they started and ended up slightly higher in year 2 compared to year 1. (Year 262.4/70.4 late) In June 2013 the first cohort of internal medicine residents graduated and were asked to rate their experience on a 5 point Likert scale with 5 being strongly agree. They rated effectiveness of working with trainees from other health professions (i.e. pharmacy, psychology, social work, nutrition) to provide quality patient care 4.5 and the educational value of opportunities to work with trainees from other health professions 4.19. Patient surveys indicate a high level of satisfaction equivalent to other VA providers locally and nationally. When asked using any number from 0 to 10, where 0 is the worst provider possible and 10 is the best provider possible, what number would you use to rate this provider? 73 % and 76 % of residents and NP students patients respectively rated their provider either a 9 or 10. DISCUSSION/REFLECTION/LESSONS LEARNED: Trainee clinical care teams can learn about and improve team functioning through a curriculum involving didactics, workplace reinforcement and reflection. Interprofessional trainees can successfully work and learn together to deliver team-based care creating positive experiences for trainees and patients alike. TEACHING PATIENT-CENTERED ELECTRONIC MEDICAL RECORD (EMR) USE TO MILLENNIAL LEARNERS: ARE WE PREACHING TO THE CHOIR? Wei Wei Lee; Lollita Alkureishi; Jeanne M. Farnan; Vineet Arora. University of Chicago, Chicago, IL. (Tracking ID #1937618) NEEDS AND OBJECTIVES: Studies demonstrate that Electronic Medical Record (EMR) use in exam rooms can prevent providers from focusing on patients and be detrimental to patient-doctor communication. Despite rapid EMR adoption, few curricula address how to teach patient-centered EMR use to tech-savvy millennial learners. In 2013, we implemented a Patient-Centered EMR Use curriculum for second-year medical students (MS2s), which consisted of a lecture and Observed Structured Clinical Exam (OSCE). We aim to assess the impact of our curriculum. Our objectives are to: 1) Compare OSCE performance of MS2s, who received the lecture, to third-year medical students (MS3s), who did not receive training on this topic. 2) Compare MS2 and MS3 post-OSCE survey scores on self-assessed knowledge, attitude and skills. SETTING AND PARTICIPANTS: We implemented a Patient-Centered EMR use curriculum for MS2s which consisted of a 1 h lecture and an OSCE to allow for skills practice. MS3 students served as historical controls; they did not receive the lecture but did participate in the OSCE as part of a required year-end multi-station OSCE exercise. DESCRIPTION: The MS2 lecture was integrated into a required Clinical Skills course. The OSCE required students to: 1) log into the EMR and review relevant records; 2) address the patients chief complaint and; 3) use the EMR to counsel the patient on obesity. MS2s participated in a group OSCE (1 student interacted with the Standardized Patient (SP), 3 students observed). MS3s participated in individual OSCEs, where each student directly interacted with the SP. An SP evaluation tool was developed to assess OSCE performance. The SP evaluated every student he directly interacted with during the OSCE. Post-OSCE self-assessment surveys were distributed to all students. MS2s performance on the OSCE and self-reported knowledge, attitude and skills was compared to the historical control group of MS3s who did not receive training. EVALUATION: 45 MS2s attended the lecture, 88 participated in the group OSCE (n = 20 students who directly interacted with the SP; n = 68 student observers), and 80/ 88 (91 %) post-OSCE surveys were analyzed. Ninety-six MS3s participated in the OSCE and 88/96 (92 %) post-OSCE surveys were analyzed. 96 % (77/80) of MS2s and 85 % (75/88) of MS3s agreed (3 on 5 point scale) that the topic was at least moderately important to their current training. 55 % (44/80) of MS2s rated their knowledge as good or excellent compared to 19 % (17/88) of MS3s. 39 % (31/80) of MS2s rated their training as good or excellent compared with 14 % (12/88) of MS3s. 87 % (70/80) of MS2s reported moderate confidence (3 on 5 point scale) compared with 73 % (64/88) of MS3s. Overall, 85 % (68/80) of MS2s and 70 % (62/88) of MS3s agreed or strongly agreed that training should be required for all students. The SP evaluation tool consisted of a 16-item checklist (score range: 1580 points). The SP ratings were significantly higher for MS2s (n = 20) than for MS3s (n = 96) [70.8 (SD = 4.3) v. 58.1 (SD = 13.1), p < 0.001]. Overall, 95 % (19/20) of MS2s were rated as good or excellent in their ability to use the EMR to enhance communication compared with only 53 % (51/96) of MS3s. DISCUSSION/REFLECTION/LESSONS LEARNED: Medical educators may assume that millennial learners are inherently tech-savvy and adept at integrating the EMR into clinical care. Interestingly, we found that MS2s who received a short lecture on patient-centered EMR use performed significantly better on an OSCE than MS3s who received no training on this topic. Our study shows that at baseline, millennial learners may not have the skills necessary to integrate the EMR in a patient-centered manner and should be taught patient-centered EMR use. TEACHING QUALITY IMPROVEMENT USING A CASE-BASED LEARNING FORMAT Georgia N. McIntosh; Nathan Schwartz; Steven Bishop; Michelle N. Brooks; Bennett Lee. Virginia Commonweallth University, Richmond, VA. (Tracking ID #1933579) NEEDS AND OBJECTIVES: Despite years of innovation in business, and aviation, the field of QI with regards to medicine is relatively new. As such, there has been little research into the most effective teaching method. The purpose of this project is to pilot a case-based learning format for delivering QI knowledge. This interactive format lends itself to discussion and brainstorming different solutions to problems, the essential piece of developing a QI project. Through this format we aimed to teach residents the components o QI: develop a process map, define change items, create an aims statement and measures and implement and interpret data from a PDSA cycle. SETTING AND PARTICIPANTS: This educational model is geared towards internal medicine residents but can be used for residents in any field, medical students or faculty members new to quality improvement. DESCRIPTION: Our case-based learning format occurred over two 2-hour sessions. For the first session, participants first completed a pre-test self- assessment rating their comfort level in different aspects of QI. Residents were divided into groups of 67, for a total of 6 small groups. Participants were introduced to a case- a trend of intern duty hour violations during a specific rotation. After a brief overview of QI principles in healthcare, the groups were tasked with creating their own aim statements based on the given case. Using the gallery walk concept where groups post their work on the wall and each group is allowed to view and comment on each proposal, the large group then critiqued and voted on which aim statement to implement. Next, residents were given a brief review of process improvement tools (process mapping, mind maps, Ishikawa diagrams), and were tasked with mapping out a typical day for the rotation in question. After a brief review of the concept of change items and measures development, groups developed their own change items based on their aim statements. The large group again used the gallery walk concept and voted on the change items to implement. In the second session, groups were reminded of the aim statement and change items that were voted on during the previous session. They were tasked with developing their measurement items in their small group work and again voted in the large group format. Participants were then given simulated data based on the intervention they voted on. From this, the concepts of run sheets and variations in data (common cause variation, special cause variation) were introduced. Based on their analysis of this data, groups developed next action items in their PDSA cycles and voted on these. EVALUATION: Forty-four residents completed the pre-test and 20 completed the post-test survey. To the question how comfortable are you in your current skills with the following aspects of quality assessment and improvement, participants answered 12 questions with the following choices: 1) not at all, 2) slightly, 3) moderately or 4) extremely. We dichotomized the responses into either 1 and 2 or 3 and 4. For all questions, using Fischers exact test, respondents reported improved comfort in their quality improvement skills, most notably with using measurement to improve their skills (38 % to 85 %, p = 0.003), identifying whether a change leads to an improvement (37 % to 70 %, p= 0.03), and in using the PDSA model as a systematic framework for trial and learning (10 % to 55 %, 0.004.) DISCUSSION/REFLECTION/LESSONS LEARNED: Case-based learning is an engaging learning technique that can easily be applied to teaching QI. Because casebased teaching is discussion-based, it lends itself well to the ability to brainstorm different scenarios that are essential to crafting a project. With our format, we were able to effectively introduce the concepts of quality improvement in an active and engaging learning platform, as evidenced by improved self- reporting of their comfort levels. One obvious caveat to our study was the reduction in post-test survey responses, partly due to variability in conference attendance due to clinical duties as well as a lack of perceived importance of residents. We also recognize that this reports comfort level, not actual knowledge or skill. Next steps would include a means to assess knowledge of QI and the subsequent skill of performing a true project. This educational model has the potential to be used for practitioner new to quality improvement regardless of level of training. TEACHING RESIDENTS SCREENING, BRIEF INTERVENTION AND REFERRAL TO TREATMENT SKILLS FOR ALCOHOL USE: USING CHART-STIMULATED RECALL TO ASSESS CURRICULAR IMPACT Maria A. Wamsley1; Nathaniel Gleason1; Michelle Guy1; Katherine Julian1; Patricia S. OSullivan2,1; Scott Steiger1; Jason Satterfield1. 1University of California, San Francisco, San Francisco, CA; 2University of California, San Francisco, San Francisco, CA. (Tracking ID #1937193) NEEDS AND OBJECTIVES: It is widely recognized that there is a critical need for more screening, brief intervention and referral to treatment (SBIRT) for alcohol use disorders (AUDs) in primary care settings. Barriers include inadequate provider skills and confidence as well as systems-related factors such as time pressures in the clinical setting and lack of referral resources. The optimal format for curricular interventions to address physician-related barriers remains unclear. One potential solution to systems-related issues is to better support SBIRT for alcohol use disorders using tools in the electronic health record (EHR). We implemented a 3-hour curriculum to instruct primary care internal medicine (IM) residents in SBIRT skills. We also developed a set of tools in the EHR to serve as a scaffold for the SBIRT curricular content, facilitate documentation of alcohol use, and to provide patients with tailored resources. SETTING AND PARTICIPANTS: This study took place at an academic primary care residency program. Twenty PGY2 and PGY3 primary care IM residents participated in the curriculum and 16 participated in the evaluation. DESCRIPTION: The curriculum consisted of two 1.5-hour sessions. We developed charting tools for the EHR (EPIC) to prompt residents to consider curricular content and to facilitate appropriate and efficient documentation of alcohol use. These included alcohol use history (HPI) and assessment and plan (A/P) charting tools and patient information resources (PI) with a list of referral resources. We introduced charting tools to learners after SBIRT curriculum completion. EVALUATION: We evaluated the SBIRT curricular intervention utilizing chart review and chart-stimulated recall (CSR). Six months after curriculum completion, we provided residents a list of their patients seen in the preceding 12 months who were drinking at or above the recommended drinking limits and asked the residents to select up to 3 patients for review with a faculty member. Faculty reviewed charts to determine whether patients were seen in the 6-month period after curriculum completion and used a 25-item checklist to assess resident use of SBIRT skills and charting tools. Faculty met with residents individually, performed a CSR for each patient using a structured interview form and provided feedback to the residents. Residents subsequently evaluated the CSR process. DISCUSSION/REFLECTION/LESSONS LEARNED: Sixteen residents participated in the chart review; 39 charts met criteria for our study. Residents did reasonably well at documenting alcohol use in the medical record (79 %) including quantity and frequency of use (64 %) and documenting their recommendation for patients to reduce their alcohol use (59 %). Residents correctly diagnosed the patient, documented an appropriate follow-up and plan only about half of the time. Resident reported barriers included time, resident discomfort, competing issues in the visit (medical/mental health), and perceived willingness of the patient to engage. Only 3 charts (8 %) included the use of the electronic HPI tool, 1 chart used the A/P tool, and 3 (8 %) used the PI tool. Resident reported barriers to EHR tool use included lack of awareness of the tools and lack of use of electronic charting tools in their normal workflow. Overall satisfaction with the CSR process was high; residents felt that the CSR reinforced SBIRT curriculum and provided valuable feedback and planned to make changes in their clinical practice. A more intensive curriculum with opportunities to practice skills and receive feedback over time may be required to improve skills. For EHR tools to be useful, additional reinforcement including simulation to practice integration of the tools into the visit workflow is essential. TEACHING SAFE TRANSITIONS: A POST-DISCHARGE FOLLOW-UP PILOT FOR THIRD-YEAR MEDICAL STUDENTS Kara Bischoff; Hemali Patel; Larissa Thomas; Karen E. Hauer; Cindy Lai. University of California, San Francisco, San Francisco, CA. (Tracking ID #1934559) NEEDS AND OBJECTIVES: Adverse events are common during transitions of care (TOC), including discharge from the hospital. About 20 % of patients experience adverse events after discharge and nearly 20 % of Medicare fee-for-service patients are readmitted within 30 days. Acknowledging the critical role that the healthcare system plays in determining quality of care, medical school curricula are focusing increasingly on systems-based practice. By engaging medical students in analyzing their patients transitions out of the hospital, we aimed to develop students awareness and skills regarding TOC. SETTING AND PARTICIPANTS: All third year medical students during their core medicine clerkship at University of California, San Francisco. DESCRIPTION: Students attended a one-hour, interactive didactic session about TOC. We emphasized common adverse events that occur during TOC, risk factors for these adverse events, and evidence-based strategies to prevent them. Students reviewed the electronic medical record and called at least one patient whom they discharged from the hospital. Students used a templated note to record any common risk factors for adverse events around the time of discharge and whether the patient experienced any adverse events, including inability to obtain medications, inability to adhere to recommendations, uncontrolled symptoms, lack of intended follow-up, emergency visits, or readmissions. Students used a semi-structured phone call format using the Teach-Back Method to assess patients comprehension of their medical conditions and treatment plan, as well as their adherence to the plan. Students reinforced healthy behaviors and communicated with primary care providers (PCPs) about any adverse events identified. Students completed a structured, written reflection about ways they will change their clinical practice based on this exercise. They identified systems issues in need of improvement. They shared their work with their inpatient team and sent a summary e-mail to PCPs. Students received quantitative and qualitative feedback from an evaluation committee, which was intended to be constructive and did not contribute to their clerkship grade. EVALUATION: Students completed surveys before and after this curriculum. So far, 32 students have completed the curriculum. Preliminary results from 32 students who completed the pre-survey and 14 students who completed the post-survey show that the percentage of students who feel knowledgeable about how TOC occur at the time of hospital discharge increased from 31 % before the curriculum to 86 % after. The percentage of students who feel confident in their ability to identify risk factors for adverse events during TOC increased from 38 to 93 %. The percentage of students who feel confident in their ability to prevent or mitigate adverse events during TOC increased from 6 to 64 %. Eighty-six percent of students feel that this exercise fills an educational gap, and 86 % recommend continuing the exercise. The most common criticism, cited by three students, is the time required to complete this exercise. The rate of successful completion of two multiple choice questions that tested students ability to identify and mitigate adverse events during TOC increased from 16 % and 94 % prior to the exercise to 36 % and 100 % after the exercise, respectively. DISCUSSION/REFLECTION/LESSONS LEARNED: This pilot demonstrates that a didactic and experiential curriculum can fill an educational gap and improve students knowledge and confidence regarding TOC. TEACHING SHARED DECISION MAKING: IMPROVING RESIDENT COMMUNICATION THROUGH CURRICULAR INNOVATION AND OB SERVED VISITS Maryann K. Overland1,2; Linda Pyke1; Anne P. Poppe1; Joyce E. Wipf1,2. 1VA Puget Sound Health Care System, Seattle, WA; 2University of Washington, Seattle, WA. (Tracking ID #1927018) NEEDS AND OBJECTIVES: Communication skills are foundational for primary care physicians and central to patient-centered care. The literature is rich with evidence that shared decision making (SDM) is underutilized among primary care physicians, and there are no best practices for how to teach it. To strengthen the ability of medicine residents and students to engage patients in their care plan, our VA Center of Excellence in Primary Care Education (COEPCE) designed a curriculum to improve mastery of SDM. Our curriculum consisted of direct observation and brief didactics, production and review of several videos of expert clinicians, and standardized patient (SP) encounters. SETTING AND PARTICIPANTS: In the spring of 2013, our VA COEPCE developed and implemented a half-day seminar on SDM for an interprofessional mix of trainees consisting of internal medicine residents and doctorate-level nurse practitioner students (DNPs). The session was preceded and followed by observed clinical encounters with direct feedback from continuity clinic mentors using a validated SDM tool. DESCRIPTION: First, faculty mentors were trained to use the OPTION tool and observed trainees during a clinic visit to assess the trainees baseline skill in SDM. Then, trainees participated in a half-day seminar that included the following: 1. Didactic on the spirit and skills of SDM 2. Review of videos of unscripted encounters featuring local expert clinical faculty and SPs engaging in SDM. 3. A brief review of clinical guidelines and provision of fact sheets on six cases: diabetes, hypertension, breast cancer screening, depression, atrial fibrillation, and goals-of-care in terminal cancer. 4. SP encounters to practice SDM around the six cases. 5. Wrap up and feedback To maximize participants comfort, practice time, and clinical knowledge acquisition, we created three stations where SPs alternated between two unique roles. Medicine residents and DNP students rotated through each station in pairs, practicing and observing each other. The observer and SPs provided immediate and direct feedback. After the seminar, faculty mentors again observed resident clinical encounters to evaluate SDM skill acquisition. EVALUATION: Faculty observers reported improved use of SDM during patient encounters after the seminar. Trainees reported a meaningful increase in confidence in their ability to use SDM (per visual analog scale) and a high likelihood to use the skills acquired during this seminar (4.5/5 on a 5-point Likert Scale). They were very enthusiastic about the standardized patients and working in pairs (4.7/5 on a 5-point Likert Scale). Trainees highly valued the video demonstrations of brief encounters by experienced clinicians. DISCUSSION/REFLECTION/LESSONS LEARNED: SDM is an important part of every primary care visit, but is a skill that is difficult to teach, often taking a lifetime to master. In this seminar, we achieved our dual-mission of providing a safe setting for practicing challenging communication skills and embedding it within clinical knowledge attainment. The design of this seminar successfully minimized knowledge differences among interprofessional learners. This innovative curriculum utilizing standardized patient encounters and real-life clinical scenarios allowed primary care physicians to strengthen communication skills around engaging patients in shared decision making. TEACHING TEAM-BASED CARE: THE USE OF INTERDISCIPLINARY TEAM MEETINGS IN THE AMBULATORY INTERNAL MEDICINE RES IDENCY CURRICULUM Jonathan Arend1; Stefani Russo1; Eva Waite1; Deborah Korenstein3; Lauren Peccoralo1; Aparna Sarin1; Alex Federman1; Linda Pagan2; Theresa Soriano1. 1Icahn School of Medicine at Mount Sinai, New York, NY; 2Icahn School of Medicine at Mount Sinai, New York, NY; 3American College of Physicians, Philadelphia, PA. (Tracking ID #1939032) NEEDS AND OBJECTIVES: Amid increasing enthusiasm for team-based models of patient care, incorporation of team-based care into residency curriculum is starting to emerge. As academic primary care practices embrace the Patient-Centered Medical Home (PCMH) model and residency programs seek ways to satisfy Next Accreditation System (NAS) milestones related to team-based care, educators must develop methods for teaching the skills needed to operate effectively in interdisciplinary care teams. As part of a comprehensive PCMH curriculum, we instituted an outpatient interdisciplinary team meeting structure to support team-based care and team-based quality improvement. We sought to determine the feasibility of team meetings and assess resident satisfaction with interdisciplinary care before and after institution of the meetings. SETTING AND PARTICIPANTS: In our academic primary care practice at the Icahn School of Medicine at Mount Sinai in New York, NY, internal medicine residents are assigned to 1 of 3 firms, in which they provide patient care and work with firm-based attending physicians, medical assistants, registered nurses, nurse practitioners, social workers, care coordinators, patient navigators, and front desk registration staff. In July 2013, we implemented weekly team meetings in each firm, attended by members of each of the above disciplines. All residents attended their respective team meeting during ambulatory blocks. DESCRIPTION: We developed the format of the team meetings based on literature review, collaboration with other academic medical centers, and local priorities. Prior to initiating the firm-based meetings, we held several all-staff meetings, in which residents, general internal medicine faculty, and clinic staff were trained in principles of teamwork, effective meeting skills, and quality improvement methodology. We redesigned the resident schedule and clinic staffing to provide 1 h per week of protected time for the meetings. The meetings were co-led by a registered nurse and an attending physician and focused on team communication, clinical workflows, practice initiatives, and quality improvement. EVALUATION: Before implementing the meetings, we surveyed all residents to assess their satisfaction with interdisciplinary care at the practice. Sixty-six surveys were returned (51.5 % response rate). When asked about the practices team-based functioning, 8 % of the residents were very satisfied, 42 % were somewhat satisfied, 42 % were somewhat dissatisfied, and 8 % were very dissatisfied. Levels of satisfaction (somewhat satisfied) for components of team-based care, such as comanagement of patients medical conditions, communication about patient care, shared responsibility of patient care, and personal support from team members, ranged from 37 to 48 %. Six-month follow-up survey results are pending. The regular meeting schedule was maintained without disruption of clinical care, losses in productivity, or significant use of additional funding sources. DISCUSSION/REFLECTION/LESSONS LEARNED: Weekly interdisciplinary team meetings are a feasible method of introducing internal medicine residents to team-based care in an ambulatory setting. Ongoing evaluation will determine whether participation in such meetings alters attitudes toward team-based care. The role of team meetings in influencing ACGME core competencies or other performance metrics requires further study. TEACHING THE PATIENT-CENTERED MEDICAL HOME Elisha L. Brownfield; Benjamin Clyburn; Kimberly S. Davis; William P. Moran; Patrick D. Mauldin. MUSC, Charleston, SC. (Tracking ID #1936971) NEEDS AND OBJECTIVES: The consensus report from the Alliance for Academic Internal Medicine Education Redesign Task Force calls for transformative ambulatory training for Internal Medicine. This would include the training of residents in highly functional ambulatory care settings and allowing residents to take on leadership roles within multidisciplinary clinical care teams while minimizing the conflict between outpatient and inpatient responsibilities. To date, there are few published educational models for Internal Medicine residency programs. In July 2012, our Internal Medicine (IM) residency program redesigned ambulatory training. Our objectives included: 1) improvement of care continuity, 2) increased exposure to common outpatient clinical problems and non-Internal Medicine specialties, 3) expansion of the ambulatory curriculum, 4) separation of inpatient and outpatient responsibilities, and 4) improved resident satisfaction of the clinic experience. SETTING AND PARTICIPANTS: The University Internal Medicine (UIM) is a hospital-based, General Internal Medicine teaching clinic with approximately 11,000 patients and an NCQA-certified Level 3 Patient Centered Medical Home. Since July 2012, a total of 92 categorical Internal Medicine and combined (Med-Peds and MedPsych) PGY 23 residents have rotated through the new ambulatory curriculum. The UIM is staffed by 15 Internal Medicine faculty members and a complement of registered and licensed practical nurses, certified medical assistants, PharmD/certified diabetes educators, social worker, and administrative support staff. DESCRIPTION: In July 2012, the IM program transitioned all of its PGY 23 categorical residents from a weekly half-day clinic schedule into a monthly clinic block occurring every 4th month throughout the year. During these blocks, residents have no ward responsibilities and spend 67 half days per week in the UIM clinic while the remainder of the week rotate through a variety of other experiences including gynecology, dermatology, orthopedics, otolaryngology, allergy and sleep medicine. Each resident also has an assigned academic half day during the week. Residents are grouped into teams of approximately 15 residents all sharing the same nurse case manager. Significant curriculum changes include enhanced didactic conferences and required completion of an online curriculum and mini-CEX evaluations. A weekly interdisciplinary meeting focusing on systems-based care of resident panel patients was introduced and most recently centers on patients with poorly controlled diabetes and hypertension. EVALUATION: Cumulative resident evaluation scores for the UIM rotation were comparable between pre-intervention and post-intervention time periods (1207 responses: 3.47, 102 responses: 3.50 Likert scale 14). In a May 2013 survey of 28 residents, the majority felt the interdisciplinary meeting lead to improvements in clinic operations (57.1 % strongly agree/agree). These residents were less sure of their own roles within the PCMH (46.4 % strongly agree/agree; 39.3 % disagree/ strongly disagree) or the roles of team members (32.1 % strongly agree/agree, 57.1 % disagree/strongly disagree). Follow-up surveys are pending. Resident continuity rates for 6788 visits (Oct 2012Jan 2013) were 45 % for individual Primary Care Physician and 8889 % for patient care teams. No pre-intervention comparison is available, and published rates are rare, but generally are below 50 %. Although not an explicit goal of the curricular change, there is an upward trend ambulatory medicine career choice (average 20042013:4, 2014:6). DISCUSSION/REFLECTION/LESSONS LEARNED: Our categorical Internal Medicine residents now experience complete separation of inpatient and outpatient responsibilities and an enhanced PCMH curriculum. Patient panel audits are provided to each resident, and the completion of an online curriculum and mini-CEX requirements are ensured. As our program transitions to ACGME milestone reporting, the ambulatory rotation provided early experience with this new requirement. Patient continuity within resident teams is high, but remains <50 % per individual resident. Overall, resident satisfaction with their outpatient experience is high. Strengths of the new program include: the ability to train residents within a certified PCMH, resident reflection of a patient-care data audit involving an interdisciplinary team, experience with other related specialties, and an excellent opportunity to evaluate residents over time. Our greatest challenges continue to be an increase in the administrative and teaching burden for faculty and staff, clinic staff shortages and turnover, cultural shift changes for all working in the PCMH, patient care continuity in the resident panels, and data management needs for our population management. THE (ELECTRONIC) CHART WAR: RESIDENT AND FACULTY PERCEP TIONS AND PRACTICES IN OUTPATIENT DOCUMENTATION AND CHART MANAGEMENT Sarah A. Tilstra; Peggy Hasley; Anu Munshi; Brian S. Heist; Shanta M. Zimmer; Jaishree Hariharan. University of Pittsburgh Medical Center, Pittsburgh, PA. (Tracking ID #1939144) NEEDS AND OBJECTIVES: The electronic health record (EHR) has simplified communication among providers and improved clinical decision support while creating complexity in documentation for medical trainees to master. There is no consensus on how to best teach outpatient documentation and chart management to residents. Last year, we reported on current resident practices in outpatient documentation which confirmed that residents excel in documentation within the progress note but routinely fail to update the overall EHR portal. This can impact quality, medical errors, communication with providers, efficiency, and reimbursement. Thus, an ideal resident curriculum should not just focus on the appropriate components of a progress note but also on how to best navigate, update, and utilize the online patient portal. We set out to define residents and faculty opinions regarding outpatient chart management by surveying our current residents and teaching faculty. Our objectives are: 1. Determine the perceived efficiency and accuracy of outpatient clinical documentation 2. Describe the barriers to effective documentation 3. Identify curricular needs for an outpatient resident documentation/ chart management SETTING AND PARTICIPANTS: Two faculty/resident clinic sites at the University of Pittsburgh Medical Center where 72 categorical residents and precepting faculty hold their weekly continuity clinic. Both practices have a fully integrated EHR (EpicCare). DESCRIPTION: Individual surveys for teaching faculty and residents were designed by the curricular team. Twenty-one questions were exactly the same, but appropriately worded for residents or faculty. Two additional questions were included in the faculty survey. Surveys were vetted for clarity and appropriateness by the residency program administration and faculty leadership prior to distribution in May 2013 via Survey Monkey. EVALUATION: Overall, 39/72 (54 %) residents and 14/19 (74 %) eligible teaching faculty completed surveys. Only 39 % percent of residents report satisfaction with their training in outpatient documentation and chart management, 21 % received formal training in medical school, while 57 % of faculty received training during their career. When compared to faculty, residents have a harder time completing outpatient notes within 24 h (100 % vs. 87 %), are less confident knowing the required documentation elements for a given clinical encounter (86 % vs. 56 %), are less comfortable determining the appropriate billing for a patient encounter (100 % vs. 44 %), are less likely to know how their quality of care is measured (93 % vs. 61 %), are more likely to cut-and-paste from their last note (7 % vs. 13 %), and less likely to use templates for documentation (67 % vs. 54 %). Sixty-two percent of residents, compared to 7 % of faculty, report spending >20 min on documentation per patient encounter, with 50 % of these encounters taking more than 25 min per patient. Faculty and residents are equally comfortable using the electronic tabs within the EHR, but faculty more often update the tabs to reflect appropriate clinical information. Eighty-six percent of faculty members have received feedback on documentation whereas only 54 % of residents routinely receive feedback. Preceptors routinely give residents feedback 43 % of the time, but do review and make changes to the documentation without giving direct feedback 79 % of the time. Residents mostly struggle with completing documentation on time. Knowing what to include in a note for specific patient encounters was second, followed by identifying appropriate billing and updating the tabs in the electronic chart. Nearly 50 % of residents identified more training in the electronic chart with frequent refreshers as a mandatory step in moving forward, with a subset of residents asking for easier to use templates to streamline documentation. Faculty also reported that time was a major struggle in completing outpatient documentation and that better templates could make this easier. Training in billing requirements and electronic-portal maintenance should be vital components in the outpatient curriculum DISCUSSION/REFLECTION/LESSONS LEARNED: Residents are not well trained in outpatient documentation and electronic chart management, and struggle with efficiency, navigating/updating the electronic portal, identifying needed components for documentation and billing, and do not receive adequate feedback. Going forward, we plan to develop a longitudinal outpatient documentation and chart management curriculum using minimum standards that are tailored to each PGY level, develop easy-to-use templates, construct an objective chart evaluation tool, and outline regular feedback intervals for residents. THE EFFECT OF GENDER ON ATTITUDES TOWARD THE HOMELESS AMONG PRIMARY CARE INTERNAL MEDICINE RESIDENTS Frances Norlock; Laura Sadowski; Meghan Kapolnek. Stroger Hospital of Cook County, Chicago, IL. (Tracking ID #1938650) NEEDS AND OBJECTIVES: The prevalence of homelessness is increasing. Few resident programs offer formal training in how to provide care for the homeless. Educators implemented a homeless medicine rotation to decrease residents stigma associated with homelessness and improve the clinical care they provide to homeless patients. Investigators hypothesized that resident gender would affect ones attitude toward the homeless in that female residents would have lower stigma than their male counterparts prior to the educational intervention, and in turn, male residents would have greater improvement of their stigma scores than female residents by the end of the rotation. SETTING AND PARTICIPANTS: PCIM residents (n = 12) participated in a 2 week homeless medicine rotation based at an urban, public hospital focused on collaborating with partners in the community. Residents participated in group seminars on campus, engaged in tours and clinical experiences in the community including working with patients in shelters, clinics and residential housing programs. DESCRIPTION: Curriculum consisted of small group seminars on the epidemiology, chronic disease management, health outcomes, and lifestyle issues affecting morbidity and mortality in this population. Residents worked in community-based clinics that served the homeless including those with addictions and the recently incarcerated. Learners were assigned evening shifts to care for those in shelters with medical issues as well as residential facilities which cared for women and children. Residents also heard life stories from homeless individuals in small group settings. Residents attitudes were assessed before and after the course through an 11 item Attitudes Toward Homelessness Inventory (ATHI) consisting of 4 subscales which attribute homelessness to societal cause; personal cause; having a willingness to affiliate with the homeless; and belief that homelessness is a solvable problem. A higher summation score of the 4 subscales equates to a lower stigmatizing attitude of homelessness. A meaningful difference score of 5.5 was set a priori. EVALUATION: Results from the ATHI showed that women did have lower levels of stigma toward the homeless at the onset of the rotation with a higher total summation score of 45.5 compared to the male summation score of 38.5 (p-value 0.14) consistent with a meaningful difference in gender prior to the rotation. The improvement in total scores overall was 1.75 among women and 7.6 points among men consistent with a meaningful change in scores between men and women (5.85 points), and a meaningful change in score for male residents (7.6 points) post-rotation. The greatest mean change in score between men and women was an improvement of 2.6 points in the personal cause subscale. Lastly, women had lower stigma scores for all 4 subscales pre- and post-rotation with the greatest improvement of 2.5 points in the societal cause score. DISCUSSION/REFLECTION/LESSONS LEARNED: Educators showed higher stigmatizing attitudes toward the homeless associated with gender, but with a dedicated homeless medicine rotation male residents did have a meaningful improvement in their attitudes by the end of the rotation with the greatest improvement in the personal cause subscale. Findings suggest the need for residents to be trained in caring for the homeless to improve attitudes and lower stigma when caring for this vulnerable population. THE NEW KID ON THE BLOCK: TEAM BASED LEARNING CURRICULUM IN AMBULATORY EDUCATION, A PROMISING ALTERNATIVE Raghav Sharma; Sandy Balwan; Alice Fornari; Tomoko Ouchi. North Shore LIJ Health System, Great Neck, NY. (Tracking ID #1938778) NEEDS AND OBJECTIVES: Team Based Learning (TBL) was developed to promote active learning, problem solving, and concept application in college classrooms and is a key strategy in undergraduate medical education. Compared to didactic teaching, TBL provides an active learning strategy that strongly engages participants in the learning process. There is little data however, on the use of TBL in Graduate Medical Education (GME). In 2012, we replaced our didactic based ambulatory medicine curriculum with a TBL curriculum. We hypothesized that residents would report better engagement and that faculty would report more resident involvement. We also hypothesized that residents medical knowledge would improve based on their in-training exam scores. SETTING AND PARTICIPANTS: All ambulatory clinic sites of a large urban internal medicine residency program participated. Each site hosted 12 residents for 35 weeks. A total of 258 Resident surveys and 34 faculty surveys were reviewed. DESCRIPTION: Residents were divided into two groups prior to initiating TBL sessions. Ambulatory faculty members prepared 7 distinct TBL modules: Diabetes, hypertension, hyperlipidemia, perioperative evaluation, asthma, dizziness, and GERD. TBL sessions were held for 1 h, 3 days per week. At the beginning of the TBL sessions, residents underwent a closed book quiz on a previously emailed article, followed by a group quiz, and concluded with a group discussion of the answers. During the second session residents received a clinical vignette on which they self assigned learning objectives (LO), which were discussed in a round table fashion at the close of the weeks TBL session. At the conclusion of the TBL sessions residents were given anonymous peer evaluations and separate surveys assessing peer performance, self-engagement, and faculty member facilitation. The faculty and residents received 10 and 12 survey questions respectively. Responses were measured on a 5-point Likert scale. The results account for the mean of the numeric value of each question. The assessment of medical knowledge of the residents is ongoing, pending the results of the in-service exam scores. EVALUATION: Most residents and faculty agreed to strongly agreed that coresidents and themselves were actively involved in the TBL session and contributed meaningfully to the discussion. Residents and faculty further responded that residents were paying attention most of the time to the TBL session and residents were active learners during discussions. Residents agreed that they would like more TBL sessions and found the faculty members to be effective facilitators. Residents neither agreed nor disagreed whether the facilitator assumed a lecture model. DISCUSSION/REFLECTION/LESSONS LEARNED: The Integration of TBL into ambulatory residency curriculum fosters proactive resident engagement and personal accountability of the learner, leading to a more dynamic learning environment and enhanced retention. However, mild barriers remain as facilitators may default to traditional teaching styles while adjusting to a TBL based curriculum. Overall, if appropriately implemented in an ambulatory setting, TBL would help enrich the teaching and learning climate in residency programs throughout the country. THE RESIDENT AS TEACHER: IMPLEMENTATION OF AN EVIDENCEBASED, BEST-PRACTICES APPROACH TO CLINICAL EDUCATION THROUGH A TEACHING ELECTIVE FOR RESIDENTS Bret Simon1; Christopher Moreland1,2; Kanapa Kornsawad1,2. 1Univ Texas Health Science Center San Antonio, San Antonio, TX; 2South Texas Veterans Health Care System, San Antonio, TX. (Tracking ID #1940759) NEEDS AND OBJECTIVES: Residents make essential contributions to medical education through their work with fellow residents and medical students. However, like many faculty, residents typically have little to no formal training in educational principles and practices. The clinical-educator is typically more clinician than educator. In an effort to improve educational practices generally, and to introduce trainees to current research and practice in clinical teaching, we designed and implemented a one-month elective rotation on The Resident as Teacher. SETTING AND PARTICIPANTS: The course is open to R2 and R3 residents at a large teaching hospital in the south central US. Offered all months except December, and open to no more than two residents per rotation. DESCRIPTION: The rotation is founded on core tenets of adult learning theory (Knowles) and stresses autonomy, self-identification of learning goals, collaboration in identifying methods of achieving self-identified learning goals, and self-evaluation of progress toward those goals. Reflective writing is a major component of the rotation, as residents are asked to respond in writing to 1215 current readings from the professional literature. We believe that this practice of writing to learn is underutilized in medical education and represents a significant source of learning in this course. Residents also participate in facilitated or guided activity during the rotation through teaching activity, shadowing teams during rounds, giving feedback to students, and observing other educational practices. Readings are selected to parallel resident goals, and reflections are a way to integrate knowledge and direct experience in the teaching activities. A critical component of the course is weekly meetings between residents and course directors. These loosely structured, collegial discussions are a way of promoting critical thinking and self-insight in learners. EVALUATION: One of the novel aspects of this course is the use of concept mapping as a method of learner evaluation. Residents are provided with resources and examples at the beginning of the rotation and told they will be asked to illustrate the process of clinical teaching via a concept map by the end of the course. Residents share their current concepts and connections during weekly meetings and share their current concept map with course directors. This process encourages deep-learning and the integration of principles and practices. It requires analysis and synthesishigher order cognitive processesand helps assure long-term retention and use of content. DISCUSSION/REFLECTION/LESSONS LEARNED: Residents come to this rotation with a variety of teaching experiences, but the content and methods in this course are clearly new to all residents to date. The course is scholarly in nature, relying heavily on the professional literature and not simply a how-to approach to clinical teaching. This practice mirrors the evidence-based approach to medicine advocated in clinical practice, as one of our goals is to help learners recognize that there is body of evidence relevant to teaching as well. We have been somewhat surprised at the learning goals identified by residents, as the same goals are frequently reported: feedback skills, motivating learners, teaching styles, and being a better presenter. The utility of the concept map as a method of self-assessment and summative evaluation has been less clear. We find that some residents enjoy the activity, but many are also perplexed by it. By limiting ourselves to this visual form of organization and synthesis, we may fail to recognize the range of learning approaches residents may take. Thus, we are currently considering other synthetic activitiesnarrative, checklistsas alternative demonstrations of competence. THE USE OF SOCIAL MEDIA TO SUPPLEMENT RESIDENT MEDICAL EDUCATION Panagis Galiatsatos; Fernanda Porto Carreiro; Jennifer Hayashi; Sammy Zakaria; Colleen Christmas. Johns Hopkins Bayview Medical Center, Baltimore, MD. (Tracking ID #1928160) NEEDS AND OBJECTIVES: Since residents work at variable times, new approaches to asynchronous learning are needed to ensure all residents receive a comprehensive education. The use of social media tools may be useful, because they can be accessed at a time convenient to the learner. The objective of this study is to assess if the use of Twitter for medical education impacts the attitude and behavior of physicians-in-training towards using social media for medical education. SETTING AND PARTICIPANTS: Pre-intervention and post-intervention surveys. Internal medicine resident physicians were surveyed before the launching of the Hopkins Bayview IM Twitter page on August 1, 2013, and again 135 days later, to determine their use of the twitter application, web page, and social media for medical education in general. Participants included residents at an internal medicine urban academic training program. DESCRIPTION: All residents within our training program were administered webbased surveys, immediately before and then 135 days after creating a Twitter page devoted to medical education. The surveys assessed views and frequency of use of social media for medical education purposes, and consisted of ten Likert scale questions. Each answer consisted of a data-point on a 15 scale (1 = not useful, 3 = useful, 5 = very useful). The final question was open ended, and asked for general comments. EVALUATION: Thirty-five of 50 residents (70 %) completed the pre-survey and 37 (74 %) participated in the post-survey. At baseline, 34 out of 35 residents used social media and 9 used Twitter specifically. However, only 22 (63 %) used social media for medical education, and 3 specifically used Twitter. After the establishment of the Twitter page, the percentage of residents using social media for educational purposes increased (31 of 37 residents, 84 %), 22 specifically used Twitter (p = 0.03 for the change), and their frequency of using the application at least once a week increased from 11.5 to 56.8 % (p < 0.05). Almost all residents (35 of 37) felt that social media could be useful as a medical education tool, which numerically increased from 29 out of 35 in the pre-intervention survey, but these percentages were not statistically significant (p = 0.37). DISCUSSION/REFLECTION/LESSONS LEARNED: Residents believe social media could be used for medical education. After we launched a Twitter page for medical education, there was a significant increase in the use and frequency of Twitter for resident medical education. Further research should be performed to see if social media can impact overall medical knowledge and patient care. USING A WEB LOG TO ENHANCE COMPETENCY IN MEDICAL KNOWLEDGE AND PRACTICE BASED LEARNING AND IMPROVE MENT Matthew Hollon. Sacred Heart Medical Center, Spokane, WA. (Tracking ID #1940872) NEEDS AND OBJECTIVES: Literature suggests that role modeling is a nuanced, deliberate learning strategy. We postulated that faculty transparency in how we learn from uncertainty in our everyday clinical work would lead residents to embrace and incorporate the core principles of the competency of Practice Based Learning and Improvement (PBLI) into their professional development. The faculty sought to model the steps of PBLI on a weekly basis through the use of a web log or blog by posting a clinical scenario where we were faced with uncertainty about the best way to proceed then translated that into a clinical question and pursued an answer. Simultaneously, we sought to share our acquired medical knowledge with our whole program through the blog. SETTING AND PARTICIPANTS: A community-based, university-affiliated internal medicine residency program with a total of 21 residents and 8 core faculty. DESCRIPTION: Prior to the start of the project, we surveyed residents about actions consistent with the fundamental steps of PBLI. Specifically we asked residents the frequency that they felt uncertain in clinical situations over the last month, how often they searched for answers when faced with uncertainty, and how often they successfully found answers. Subsequently, once each week for a period of 24 weeks, 1 of 8 core faculty members wrote a summary of a situation where he or she faced clinical uncertainty, a description of how he or she went about finding an answer, and a discussion of the medical knowledge learned. Prior to the summary being posted to the blog and accessible to everyone in the program, we distributed electronically a single pre-test question to assess baseline knowledge in the area covered by the summary. One week following distribution of the pre-test, we closed access to the question then posted the written summary. Approximately 1 month after the end of this 6 month project, we distributed a final survey. This survey reassessed medical knowledge of topics covered in the faculty posts using the same 24 single pretest questions. The survey also re-evaluated actions consistent with the PBLI steps. EVALUATION: All residents completed both the initial and final survey and completed, on average, 77 % of the single pretest knowledge questions. The frequency that residents encountered clinical situations where they felt uncertain declined over the course of the 24 weeks from a mean frequency of uncertainty of 36 % to 28 % (p = 0.008). The frequency with which residents sought answers when faced with uncertainty prior to the project was high at mean of 81 % and had not changed significantly 24 weeks later remaining at a mean of 78 %. The range of the frequency of answer seeking when faced with uncertainty was between 30 % and 100 % with a total of 3 residents reporting on the final survey that they sought answers less than half the time when faced with uncertainty while 15 residents sought answers at least 80 % of the time. Success in finding an answer to questions generated by clinical uncertainty did not change significantly from before to after the project, remaining at a mean of approximately 75 %. Medical knowledge improved significantly (p = 0.001) with the use of faculty posts to the blog. Prior to posting answers to clinical questions, residents got a mean of 52 % of knowledge questions correct. When reassessed 1 month after the end of the 6 month period, residents got a mean of 65 % of questions these correct. Lastly, 95 % of residents felt that faculty should continue to post general clinical content to the blog. DISCUSSION/REFLECTION/LESSONS LEARNED: We noticed a slight decline in clinical uncertainty over the course of 6 months consistent with expectations. However, we failed to show improvement in frequency of searching for answers when faced with uncertainty. Notably, this is already happening at high level for most, which is reassuring, making it difficult to show large improvement. We did note that some residents were less engaged in the fundamentals of PLBLI. It may be that efforts to encourage self-reflection, recognition of clinical uncertainty, and answer seeking are best targeted to this group. Interestingly, strategies for finding answers did not seem to change substantially. Despite faculty role modeling using primary data found through PubMed searches, residents continued to most frequently use general resources when seeking answers to clinical questions. We did demonstrate that faculty posting to a blog led to modest improvements in medical knowledge. This was particularly noteworthy given that retest occurred up to 7 months after the initial posts of answers. Finally, residents truly appreciated this intervention. As the blog became increasingly valued as a resource for clinical information, it even inspired some residents to do their own posting of medical knowledge content. In summary, faculty role modeling of PBLI through a blog was valued by residents and led to modest improvements in medical knowledge but did not alter behavior that was already taking place frequently. USING COLLABORATIVE EVALUATION TO IDENTIFY CRITERIA FOR THE ASSESSMENT OF ATTENDING PHYSICIANS IN HOSPITAL MEDI CINE Bret Simon1; Luci Leykum1,2; Jane E. ORorke1,2; Megan Freeman1. 1Univ Texas Health Science Center San Antonio, San Antonio, TX; 2South Texas Veterans Health Care System, San Antonio, TX. (Tracking ID #1940677) NEEDS AND OBJECTIVES: Various approaches to the assessment of attending physicians have been reported in the literature. Externally determined criteria (e.g., ACGME, ABIM) are often used to ensure that measures required for accreditation are addressed. Such tools are useful for making comparisons across institutions, but they may fail to identify institution-specific needs. These measures may be particularly inadequate in cases where rapid curricular changes are being implemented or professional development activity results in extensive and significant changes in the way attendings and residents interact. Nominal Group Technique can be used as a means of identifying institution-, program-, or specialty-specific evaluation issues that are particularly salient or newly emergent. We sought to use NGT as a method for obtaining stakeholder input on attending performance in the inpatient general medicine setting. SETTING AND PARTICIPANTS: Two teaching hospitals (community and Veterans Administration) serving as primary training sites for a large medical school in the south central US. Participants included three categories of stakeholders whose roles were principally in medical education in the inpatient setting. The three groups included hospitalist faculty, administrators (departmental, hospital, and graduate medical education), and residents representing all 3 years of training. A total of 24 individuals participated. DESCRIPTION: The Nominal Group Technique is designed to solicit input around a particular problem, issue, or concern from a range of stakeholders. The process is transparenta feature particularly important when participants represent a range of organizational levels (dean, faculty, resident). Participants offer their input as short phrases or words in response to a central question, written on white-boards so they are viewable by all participants. In this case, What criteria should be used by housestaff to assess attending performance? The method is a three-step process: 1) solicitation of input from participants in an unedited, round-robin fashion until all views are expressed, 2) group review of the content to consider consolidation or elimination of content that is duplicative, 3) individual ranking of the final list of words or phrases. The ranking process can vary, but in this case participants were asked to identify their top five items. Points were assigned according to rank, with five points assigned to the top rank and one point assigned to the lowest rank. The final product is a list of rank-ordered items, produced by all stakeholders, to address the question posed. EVALUATION: Participants identified nearly 90 items related to the evaluation of attending performance. After some consolidation, 12 items were eliminated. Despite the large number of items, when they were summed and ranked, only 11 were scored high enough, or were selected by enough participants, to achieve a score of more than 10 points. These items served as the core content for a review and modification of the existing attending evaluation form. Items included: professionalism; clinical knowledge; organized approach to teaching; provides useful, timely, formative and summative, written and oral feedback; clear expectations of all team members; communication skills; ability to convey thought process; ability to provide proper autonomy and supervision; interpersonal skills with learners, patients, and families; quality of teaching; and promotes critical thinking. DISCUSSION/REFLECTION/LESSONS LEARNED: We found that many of the items identified as important criteria for the assessment of attending performance were already, to some degree, part of the existing evaluation form. Nonetheless, we have been able to utilize the input of our stakeholders to help clarify and operationalize behaviors associated with broad concepts like professionalism and communication skills. Our intent is to continue to use the nominal group technique on a semi-annual basis to assure that our attending evaluation process is adequately representing the needs and concerns of stakeholders. USING HUDDLES TO INTEGRATE MEDICAL RESIDENTS INTO AMBU LATORY TEAMS Joshua M. Liao; Andrew J. Van Wieren; Brigid M. Dolan. Brigham & Womens Hospital/Harvard Medical School, Boston, MA. (Tracking ID #1927565) NEEDS AND OBJECTIVES: Harvard Medical Schools Center for Primary Care recently launched the Academic Innovations Collaborative to catalyze widespread innovation and transformation at affiliated primary care clinics. While the initiative requires participating sites to focus on team-based care and incorporate trainees into them, many sites have found it difficult to integrate medical residents - frontline physicians and large portions of the clinical workforce - into care teams. SETTING AND PARTICIPANTS: Recognizing the importance of resident involvement in team-based care, we selected an affiliated community health center in Southern Jamaica Plan staffed by practicing primary care providers, a cohort of full-time MAs, and resident physicians from the Brigham & Womens Hospital Internal Medicine Residency. We chose the site because of strong leadership support for this work, low turnover rate among its MAs, and the presence of large proportions of full-time providers with prior experience in huddling with MAs. DESCRIPTION: We framed our project within the period when residents were physically at their half-day continuity clinic sessions. We incorporated Lean principles to study existing processes and root causes by observing in-clinic communication between physicians - both full-time and resident physicians - and medical assistants (MAs). Based on our mapping and analysis, we used an impact/ effort matrix to evaluate and prioritize potential ideas, ultimately deciding on residentMA huddles as a high yield intervention. We utilized a Plan-Do-Study-Act approach to implement pre- and post-clinic resident-MA huddles. These were initially done under the supervision and guidance of experienced physician preceptors, but eventually residents and MAs ran independent huddles. We collected pre- and postintervention surveys from practicing physician preceptors, residents and MAs assessing their perceptions of resident integration into care teams. EVALUATION: Using verbal feedback and written post-intervention surveys, we gathered data on process measures (adherence to huddles) and outcome measures (effect on perceptions of resident integration into care teams). The frequency of both pre- and post-clinic huddles increased over our study period. Our work revealed that while pre-clinic huddles were relatively easy to adopt, post-clinic huddles were considerably more difficult to organize. Interestingly, resident physicians and physician preceptors perceived residents as becoming more integrated into care teams over the study period. In contrast, MAs felt that resident integration did not change significantly. This was in part due to their perception that despite huddles, residents still spent very limited time in the clinic. Another important issue was that MAs felt that while full-time clinic physicians were able to communicate plans iteratively with staff, there was frequently no formal communication plan established between residents and MAs. DISCUSSION/REFLECTION/LESSONS LEARNED: This project taught us several important lessons. First, it supported the idea that huddles can be used effectively to increase perceptions of resident integration into clinic teams. However, it also revealed that these perceptions can vary based on an individuals role on the clinical team. These two findings underscore the reality that teamness is based on more than just the execution of pre- and post-clinic communication. Perhaps not surprisingly, they show that the content and format of communication is important. They reveal that how and when information is communicated may be more crucial for building a sense of team than the act and consistency of communication. Our data also suggests that to clinic staff, the activities that occur outside of clinic sessions may also contribute significantly to perceptions of resident integration. The ability for staff to interact with full-time physicians iteratively throughout full workdays may be important for clarifying plans, improving communication, building relationships, and contributing to a sense of team integration. Because in contrast, huddles only add an additional few minutes to a residents total time in clinic, staff may perceive them to be less integrated even despite huddles. Going forward, educational changes such as the ambulatory burst model will likely increase the amount of contiguous clinic time, and thereby increase the chance that residents can interact iteratively and naturally with MAs like full-time practitioners frequently do. Regardless, huddles will still play important roles in sustaining communication between resident physicians and clinic staff. We plan to maximize adherence to and impact of resident-MA huddles by adjusting our huddle format, ensuring that individuals convey content that is important to team workflow, and developing this into even more of a joint effort among all clinic staff via dashboards in staff meetings and reports of current huddle rates and resident integration scores. We are also exploring expansion of this project to other clinic sites. NEEDS AND OBJECTIVES: Healthy People 2020 acknowledges a need to address the health disparities endured by many lesbian, gay, bisexual, transgender, and queer/ questioning (LGBTQ) individuals. Such disparities are rooted in discrimination, violence, civil rights inequalities, and poor access to health care. Sequeira et al. (2013) noted: The Institute of Medicine recognizes that few practicing physicians are knowledgeable about and sensitive to the needs of LGBT patients. A recent survey of medical school deans found that medical school curricula devote a median of 5 h covering LGBT-related content across the 4-year curriculum. The same group indicated that improving access to curricular material would be the best way to increase the amount of content presented on this subject. The need for increased LGBTQ knowledge is clear, and it is necessary that individuals in an academic medical setting learn about health disparities endured by many lesbian, gay, bisexual, transgender, and queer/questioning (LGBTQ) individuals. The Safe Space program aims to correct this lack of knowledge as well as aims to identify the LGBTQ population and their perceptions of the healthcare communities knowledge base. This interactive presentation will provide an educational resource for colleagues and future healthcare professionals about this important topic and methods to evaluate its effectiveness. Objectives 1. Discuss how the UTHSCSA Safe Space is educating students, faculty, staff, and clinicians about social, environmental, and physical factors that influence the health of lesbian, gay, bisexual, transgender, and queer (LGBTQ) individuals. 2. Describe Safe Space outcomes to date 3. Discuss Survey model for identifying LGBTQ community demographics and opinions. SETTING AND PARTICIPANTS: The UT Health Science Center Safe Space program is designed to raise cultural competency and professionalism of all students, faculty, and staff at the Health Science Center to include more knowledge of the LGBTQ population. DESCRIPTION: Participants in the program undergo one lecture workshop approximately 90 min in length. The lecture and discussions provide information on the following topics:knowledge of LGBT terminology, stereotypes, the coming out process, suicide risk, knowledge of bias, LGBTQ education, LGBTQ resources, Suicide prevention resources, and preparedness for dealing with LGBTQ issues or scenarios. Workshops include opportunities for participants to open a dialogue on topics such as conflict management, employment regulations, adoption practices locally and nationally, substance abuse, suicide, depression and transitions between genders, all concerning LGBTQ persons. Safe Space participants are required to complete a pre and post-training survey that assesses baseline knowledge of LGBTQ terminology, the coming out experience, and suicide resources The LGBT Access to healthcare survey will provide information about the Rio Grande Valleys LGBTQ communities and their access to healthcare in the valley. The survey aims to sample a population of the LGBTQ population in the Rio Grande Valley, the presence of supportive populations, the level of comfort when accessing healthcare in the Rio Grande Valley, the perception of bias from healthcare personal in providing treatment, and the perception of the LGBTQ population on the education status of healthcare workers on LGBT topics. The survey will be anonymous. A copy of the LGBTQ community surveys will also be distributed to the attendees for their review and comments. EVALUATION: Analysis following a two tailed T-test showed significant improvements in knowledge of LGBT and Suicide resources, the coming out process, suicide risk assessment and most importantly feeling prepared with dealing with LGBT issues. DISCUSSION/REFLECTION/LESSONS LEARNED: The UT Health Science Center Safe Space Program met with a great amount of success during its initial implementation between 2012 and 2013. The data showed significance in increasing the fund of knowledge of students faculty and in regards to certain topics discussed. Some major feedback that is provided is that participants would like longer sessions to increase discussion times, as well as have more members of the LGBTQ community present to aid in the presentation. The topic is a rapidly developing field and would do well to establish itself as a permanent area of medical education. ONLINE RESOURCE URL (OPTIONAL): 1. Lesbian, Gay, Bisexual, and Transgender Health. Healthy People 2020. 22 Mar 2012. http://healthypeople.gov/ 2020/topicsobjectives2020/overview.aspx?topicid=25. 2. LGBT (Lesbian, Gay, Bisexual, Transgender) Safe Space Training. Office of Diversity, Northwestern U n i v e r s i t y F e i n b e r g S c h o o l o f M e d i c i n e . 6 M a r 2 0 1 2 . h t t p : / / www.feinberg.northwestern.edu/diversity/programs/safe-space/index.html. 3. Gina M. Sequeira, Chayan Chakraborti, Brandy A. Panunti. Integrating Lesbian, Gay, Bisexual, and Transgender (LGBT) Content Into Undergraduate Medical School VHAS COMPREHENSIVE WOMENS HEALTH SCAN-ECHO PROGRAM: LESSONS LEARNED Kristina M. Cordasco1,3; Jessica Zuchowski1; Alison B. Hamilton1,4; Herschel Knapp2; Joann O. Saavedra2; Donna L. Washington1,3. 1VA Greater Los Angeles Healthcare System, North Hills, CA; 2VA Greater Los Angeles Healthcare System, Los Angeles, CA; 3The University of California, Los Angeles, Los Angeles, CA; 4The University of California, Los Angeles, Los Angeles, CA. (Tracking ID #1939313) NEEDS AND OBJECTIVES: Although women are among the fasting growing segment of the Veterans Health Administration (VHA) patient population, many VHA primary care providers (PCPs) have relatively small female caseloads, making it challenging for them to build and maintain their womens health knowledge and skills. Therefore, as part of VAs SCAN-ECHO (Specialty Care Access Network Extension for Community Healthcare Outcomes) program, we implemented a Comprehensive Womens Health SCAN-ECHO program with the objective of building and maintaining VHA womens health PCPs knowledge in the evaluation and management of conditions specific to, more common in, or with special considerations relevant to women (e.g., gynecology, breast issues, depression, osteoporosis). SETTING AND PARTICIPANTS: Our participants included womens health PCPs across three VHA healthcare systems. DESCRIPTION: The program consisted of monthly group PCP-specialist clinical tele-videoconferencing sessions in which PCPs obtained virtual specialist consultation on their cases and received serial patient-based education. Prior to each session, PCPs submitted consultation requests on patients for whom they wanted diagnostic and/or therapeutic management advice. During sessions, the specialist at a hub facility and PCPs at multiple spoke sites engaged in real-time group discussions about the cases. In addition to giving recommendations, the specialist discussed the underlying evidence or rationale. Content also included patient communication strategies (e.g., explaining contraceptive options). The PCP who submitted the consult, as well as other PCPs participating in the session, interacted with the specialist and with each other, asking clarifying questions. The specialist also provided a brief didactic relevant to the cases. Continuing Medical Education credit was provided. Each session was held at the noon hour in the hosting sites time zone. EVALUATION: We used participation logs, participant surveys, semi-structured interviews, and field notes from key stakeholder meetings to evaluate our program. We conducted 14 sessions over 16 months. Overall, 65 PCPs in three VHA health care systems participated in one or more sessions, with an average of 11 participants per session. Participation rates were much higher in the one healthcare system where providers were provided with time protected from their clinic responsibilities (rather than meeting over lunch). Attempts to expand to two additional VHA healthcare systems, where the session would have been in the lunch hour, were not successful secondary to conflicting meetings or educational programming. Of the 58 postsession surveys received, in 45 (78 %) the participant strongly agreed, and in 7 (12 %) the participant agreed, that the information provided in the session would influence their patient care. Of these 52, 74 % indicated that the information would change their ordering of diagnostic studies; 83 % their treatment plans; 58 % their prescription of medications; and 68 % their overall interactions with patients. Among the 10 PCP interviewees, all reported that they found SCAN-ECHO to be useful for building and maintaining their knowledge on womens health topics. However, all interviewees also reported that the session being conducted during their lunch hour was a limiting factor for consistent participation, with morning clinics often running late and participants having competing clinical, administrative, educational and personal priorities for that time. Interviewees and discussions with key stakeholders revealed that the rotating specialists and topics across the breadth of womens health, although appealing in its comprehensiveness, also limited submission of cases as PCPs were unlikely to have active case-based questions coincident with the monthly specialists area of expertise. DISCUSSION/REFLECTION/LESSONS LEARNED: VAs SCAN-ECHO program is a promising modality for building and maintaining PCP knowledge on womens health topics, with attendees indicating that it influences their patient care. However, PCPs receiving time protected from clinical responsibilities is essential for robust and consistent participation. Further, narrowing in on a consistent womens health area, such as gynecology, although sacrificing the programs comprehensiveness, may facilitate PCPs having active cased-based questions for the sessions. ONLINE RESOURCE URL (OPTIONAL): http://vaww.portal.gla.med.va.gov/ sites/PAC/SCAN/Shared%20Documents/Gynecology.aspx VIDEOTAPING HOME VISITSFIRST-YEAR MEDICAL STUDENTS BRING PATIENTS WORLDS TO THE CLINICAL TEAM Barbara Ogur1,2; Marie-Louise Jean-Baptiste1; Jeremy Keller1; Nancy Blum1; Andreas Mitchell1; Margaret Fallon1; Elorm Avakame1; Ethan Balgley1; Betty Erfe1; Grace Chao1; Jose O. Gutierrez1. 1Harvard Medical School, Boston, MA; 2Harvard Medical School, Boston, MA. (Tracking ID #1934147) NEEDS AND OBJECTIVES: Early exposure to patients is recognized as very valuable in providing students with opportunities to ground their scientific studies in clinical practice. According to learning theory and corroborated by observation, students learn more and feel motivated to learn more deeply when their learning is directly connected to real patients experiences. Many programs begin clinical exposure in the first year, through a course in interviewing patients. These courses are highly valued by students. We describe a pilot to extend these curricula using the hypothesis that students will learn even more powerfully if their learning experiences provide them an opportunity to not just learn from patients but to actually contribute to their patients care. SETTING AND PARTICIPANTS: At Harvard Medical School, first year students have a year-long course in interviewing, Patient-Doctor 1, which emphasizes human connection and the understanding of patients experiences of illness by providing students opportunities to interview patients and to reflect on their experiences in small groups with experienced clinician preceptors. DESCRIPTION: We planned and implemented a pilot program to explore ways for first-year students to both learn from patients and also contribute meaningfully to patients care. In one section of the course, as part of a required home visit, we asked one group of students, with written consent of their patients, to videotape the home environment. Students then provided their video and the written report of their home visit, including what special insights they learned about the patient from being in the home, to their patients ambulatory clinical care team. EVALUATION: We surveyed students, preceptors, and the clinical team caring for the patient to determine their feelings about this exercise. Samples of the videos will be available for review. Students reported that these home visits provided them with an expanded sense of the patient as a person with important connections to family and friends, with a role in the community, and with a coherent set of beliefs about health and diet perhaps not totally aligned with the beliefs of the medical world. Students found experiencing the physical adaptations for patients with mobility constraints very informative and directly witnessed possible environmental triggers for a patient with asthma. Clinical team members learned new information about patients belief systems, adaptations, and social supports. DISCUSSION/REFLECTION/LESSONS LEARNED: We piloted a small innovation of a medical school interviewing course that both provided students with a powerful learning experience in contextualizing patients illnesses through the lens of patients lives in their homes and communities and also through the lens of performing a valued function of bringing this broader view of the patient back to the clinical team. We believe this exercise warrants a trial with a larger cohort. We learned that students, preceptors, and clinical teams found the exercise useful and inspiring. We believe that first-year students can bring a valuable, humanizing perspective on the patient to over-stretched clinical teams as they also learn to be caring physicians attuned to their patients real experiences. VISUAL THINKING STRATEGIES: AN ARTS-BASED INTERVENTION TO TEACH TEAMWORK TO SENIOR MEDICAL STUDENTS Elizabeth Karwowski1,2; Alexa R. Miller3; Suzanne Sarfaty1. 1Boston University Medical Center, Boston, MA; 2Boston VA Healthcare System, Boston, MA; 3Brandeis University, Waltham, MA. (Tracking ID #1939008) NEEDS AND OBJECTIVES: In recent years, Boston University School of Medicine (BUSM) recognized a local need to develop a formal curriculum to teach teamwork. The BUSM curriculum committee simultaneously advocated for increased exposure to the arts and humanities throughout medical school training. In order to meet these needs, the BUSM fourth year Advanced Internal Medicine clerkship developed a new curriculum using Visual Thinking Strategies (VTS) to facilitate teaching teamwork skills. VTS has been shown to successfully teach observation skills but has been studied infrequently outside of teaching observation. Objectives of the curriculum include: *Identify characteristics of a highly functioning team *Listen actively and encourage ideas/ opinions of other team members *Practice respectful dialogue and expressing differences in opinion *Explore conflict resolution skills in a group setting *Integrate viewpoints of all team members SETTING AND PARTICIPANTS: The VTS and Teamwork curriculum was developed for the BUSM Advanced Internal Medicine clerkship, a mandatory fourweek clerkship completed during the final year of medical school. The clerkship includes 1 day of classroom-based didactic teaching per week with a focus on teaching advanced communication skills. The VTS and Teamwork curriculum was inserted into these didactic sessions. DESCRIPTION: Development of the VTS and Teamwork curriculum has been an iterative process with changes in response to student feedback. The curriculum was first implemented in June 2013, evolved over the next several blocks, and finalized in December 2013. The curriculum is divided into three distinct workshops: 1) Introduction to VTS and Teamwork: Speak up, 2) Listening as a Team Member, and 3) Listening as a Leader. VTS is a well-described process that involves the facilitated discussion of artwork to build skills in observation, communication, and listening. During a VTS session a trained facilitator guides learner discussion of a work of art using three questions: what is happening here? What do you see that makes you say that? What more can we find? Each workshop includes at least one VTS session. Additionally, two of the workshops include modified VTS type activities. The second workshop incorporates a small group activity to enhance listening skills. The third workshop uses a separate small group activity to allow students to practice the role of group facilitator. In addition, students also receive formal teaching on teamwork. The curriculum uses the model described by Amy Edmondson called Teaming on the Fly, which describes five attributes of highly functioning teams: speaking up, listening intensely, integrating different facts and points of view, experimenting iteratively, reflecting on your own ideas on actions. EVALUATION: There are two methods of curriculum evaluation. First, the students complete a pre and post survey. Both surveys include the TeamSTEPPS - Teamwork Attitudes Questionnaire, a validated survey developed by the Agency for Healthcare Research and Quality to study self-rated attitudes towards teamwork. The change in score after completion of the curriculum will be analyzed. The post survey includes an additional set of questions to determine student perceptions of the curriculum. Second, the students take part in a focus group at the completion of the clerkship to further explore their attitudes towards the curriculum. The BUSM Institutional Review Board approved this study. Data collection began in late fall 2013. DISCUSSION/REFLECTION/LESSONS LEARNED: Preliminary data reveal that students largely value the addition of VTS to their curriculum as a novel method of medical education. At times they struggle to link VTS with learning teamwork skills and continue to seek an explicit set of techniques to improve their teamwork skill set. A subgroup of students express striking discontent with the curriculum. A subgroup analysis is planned to attempt to determine unique attributes of students dissatisfied with the curriculum. Thus far it seems that using an art-based intervention to teach skills beyond observation alone is generally well received and valued by students. Future attempts to develop novel arts-based curricular interventions need to focus on an explicit connection between the activity and the clinical skill to be learned. WELCOME TO THE NEIGHBORHOOD: TEACHING THE SOCIAL DETERMINANTS OF HEALTH Jada C. Bussey-Jones; Maura George; Stacie Schmidt; Jennifer E. Bracey; Mehul Tejani; Schuyler D. Livingston. Emory University, Atlanta, GA. (Tracking ID #1934052) NEEDS AND OBJECTIVES: New physicians require broad medical knowledge and a practiced set of procedural skills. Training in these areas would suffice in decreasing disease burden if medical care were the only factor relevant to patient health outcomes. Research suggests, however, that multiple complex social determinants (e.g. neighborhood characteristics, social policies, and available resources) play a significant role in producing or mitigating health outcomes and racial/ethnic disparities. Studies suggest that fourth year students were no more knowledgeable about health care access than first-year students, and that students attitudes toward the underserved become more negative throughout medical school. Further, most current educational environments do not prepare learners for engagement with broader health issues. This course was designed to 1) describe the complexity and impact of community and other social determinants of health on patient health outcomes, and 2) state ways, to connect community/public service and advocacy with learners clinical and academic work. SETTING AND PARTICIPANTS: Emory University School of Medicine month long elective course available to interprofessional learners (nursing, medical, PA students) at Emory and Morehouse Schools of Medicine, as well as visiting students from outside institutions. DESCRIPTION: Our curriculum uses a combination of didactic lectures, readings, experiential learning, direct service activities and personal reflection to equip future physicians to practice medicine in a socially complex world. We help students better understand and cooperatively address the problems facing patients and communities. Examples of student activities include assessment of community resources such as green space, fresh vegetables, pharmacy supply, safety, tobacco, alcohol stores/ billboards. Learners also participate in simulation experiences (e.g. navigate public hospital as a patient - assessing wait times, barriers, etc.). Participants engage with community partners sites using an asset model approach to collaborate and participate in health related interventions. Our students also engage several essential community resources that provide context for the health of many of their patients including, homeless shelters, nursing homes, and prison medical facilities. Finally, they are trained asked to complete advocacy tasks such as visiting the state capital, meeting with legislators, writing letters to the editor or position papers. This course also utilizes several means of reflection including small group discussions, journaling, photo journaling. EVALUATION: The course was evaluated in several ways 1. Learners attitudes were compared to controls through use of Medical Student Attitudes toward the Underserved questionnaire (MSATU), a validated survey instrument administered at 0, 1 and 12 months. Our preliminary results reveal a significant difference in MSATU scores between controls and course participants at -month 0. However, there was no significant change in scores when comparing 0 to 1 month survey results within social med participants or controls. This suggests that there may baseline attitudes among self-selected participants that led them to this course. 2. Semi-structured interviews were completed at the end of the elective to characterize the learners and identify best received components of the elective. The curriculum was very well received, and we had a marked increase in the number of requested participants in the course over the 2 years of implementation. 3. Ongoing analysis is underway, including an electronic database for longitudinal analysis of elective participants. DISCUSSION/REFLECTION/LESSONS LEARNED: This presentation will highlight the development and implementation of the Social Determinants of Health curriculum at Emory University School of Medicine. We will describe the core components of our curriculum, share tools, and discuss challenges. We will also specifically discuss strategies to forge relationships with community partners, collaborate with educational leadership to broaden our curricular impact, increase core faculty knowledge and expertise, and evaluate and modify curriculum. This elective seems to attract a selected group of individuals with positive baseline attitudes about caring for the underserved. Qualitative interviews suggest these positive attitudes were bolstered through service learning. Learners in our course expressed a desire for more social medicine concepts and didactics to occur earlier in undergraduate medical education. An increased number of learners and ongoing evaluation efforts will direct future course development and improve our understanding of the personal characteristics, qualities, and experiences that lead to positive attitudes about care of underservedand careers in community engagement, service, and health advocacy. WORK-PLACE BASED ASSESSMENTS: LEVERAGING CLOUD-BASED MOBILE TECHNOLOGY FOR CAPTURING NARRATIVE COMMENTS Churlson Han; Gary S. Ferenchick; David Solomon. Michigan State University, East Lansing, MI. (Tracking ID #1939243) NEEDS AND OBJECTIVES: Direct observation and feedback are essential components of medical education. Immediate, specific feedback can be vital to learner development. Valued feedback is frequent, relevant to observed behavior, specific, and timely. The purpose of this exploratory study was to calculate the percentage of work-place based assessments, completed with a Cloud-based mobile just-in-time evaluation system (i.e. JIT, described below), that included any written comments (i.e. comment-enhanced evaluation); and to calculate the percentage of comments that were general, specific, or specific with detail, regarding students clinical performance; and to describe evaluator variables affecting comment specificity. SETTING AND PARTICIPANTS: Between July 2012 and September 2013, 256 students at Michigan State Universitys College of Human Medicine were required to obtain ten directly observed evaluations. Using JIT, all evaluator comments were electronically captured in a permanent, cloud-based record, and were subsequently coded on three levels: general (Good physical exam skills or Practice more), specific (Excellent neuro exam or Better organization), or specific with detail (Listened to axilla for radiation or Percuss lung fields & compare sides). Evaluators were not provided any training on providing narrative comments. Three raters independently coded all narrative comments after establishing inter-rater reliability. DESCRIPTION: We developed, implemented and assessed a Cloud-based clinical evaluation application (i.e. Just In Time Medicine or JIT) for internet-enabled devices including smart phones, which displays customized competency checklists for 17 CDIM training problems for evaluators to use when implementing work-place based assessments of students clinical skills in authentic clinical settings. JIT includes discretionary sections for evaluators to provide narrative written comments pertaining to: 1) one area the student did well and 2) one area the student can improve upon. EVALUATION: 2658 CEX assessments were completed among the 256 students, with 227 students (88.7 %) having at least one comment-enhanced evaluation. The range of comment-enhanced evaluations per student was 011, with an average of 4.9. Out of a total of 1285 comments, 288 included only well done comments, 44 included only needs improvement comments. Nine hundred fifty-five included comments for both. For well done feedback, 17.2 % were general, 57.2 % specific, and 22.7 % specific with detail. For needs improvement feedback, 14.8 % were general, 53.4 % specific, and 31.8 % specific with detail. Resident feedback was compared to faculty. For well done comments, faculty feedback was 28.0 % general, 56.0 % specific, and 16.0 % specific with feedback, while resident feedback was 18.2 %, 58.4 %, and 23.4 % respectively. For needs improvement comments, faculty feedback was 11.5 %, 50.3 %, and 38.2 %, while resident feedback was 16.4 %, 55.3 %, and 28.3 % respectively. DISCUSSION/REFLECTION/LESSONS LEARNED: When directly observing and assessing learner performance in authentic clinical settings, we determined that written comments accompany a high percentage of evaluations using Cloud-based technology delivered via mobile devices. These comments provide a high level of specificity. Evaluators displayed more detailed feedback for needs improvement items. We did not observe any variance in well done comments between faculty and residents, but faculty provided more specific with detail feedback on needs improvement items. This use of technology to capture narrative comments can provide specific and relevant student feedback. ONLINE RESOURCE URL (OPTIONAL): www.justintimemedicine.com/mobile UN: PW: testuser WORKING TOWARDS MEANINGFUL RESIDENT INVOLVEMENT IN QUALITY IMPROVEMENT AT TWO UNIVERSITY CONTINUITY CLINICS Danielle F. Loeb; Susan M. Nikels; Rachel Swigris; Suzanne Brandenburg. University of Colorado Denver, Aurora, CO. (Tracking ID #1938666) NEEDS AND OBJECTIVES: The ACGME requires that residents are integrated and actively participate in interdisciplinary clinical quality improvement (QI) and patient safety programs. In 2012 we initially introduced a QI curriculum in the ambulatory setting for all interns. Summative assessment by QI faculty and leadership revealed that lack of faculty exposure to QI principles and a non standardized approach to project organization led to high variability in project quality and resident involvement. The objective of this project was to improve integration of interns into QI projects in their continuity clinic settings. SETTING AND PARTICIPANTS: This educational innovation took place in two resident ambulatory internal medicine clinics associated with one university hospital. It included 19 interns in the two clinics. DESCRIPTION: Drawing on our prior experience and a literature review of best practices, we developed a combined didactic and experiential curriculum (Figure 1). The primary improvements include: 1) IHI (Institute for Health Care Improvement) QI curriculum required for interns and QI faculty 2) Each intern developed a personal QI project to practice fundamental QI skills 3) Interns integrated into designated clinic team QI projects 4) Dedicated FTE for QI faculty to lead clinic team QI projects 5) Personal QI projects will be reviewed with a standardized evaluation including: SMART aim statement, outcome measures, systematic data collection, PDSA cycle completion, presentation of data, and completion of IHI curriculum. 6) Each intern will submit summary of team QI project including: SMART aim statement, literature review, fishbone diagram or process maps, outcome measures, and PDSA cycle completion. Figure 1. Quality Improvement Curriculum by Calendar Month July IHI Open School (IHI OS) QI 101103 completed through GME August IHI OS QI 105: The Human Side of QI Completion of QIKAT Pre-test September Create Personal Improvement project Conduct Personal PDSA cycle October Review concepts from IHI OS QI 101 Begin to compile list of possible QI projects November IHI Open School QI 102 and 103 Choose QI projects (determine teams, etc.) December IHI Open School QI 104 Start project charter Determine clinical team dedicated to mentor project January Review project charter with faculty, revise PRN Create Aim Statement February Review Aim statement with faculty, revise PRN Determine sources of data Start SIPOC analysis March Revise SIPOC analysis Start drafting Fishbone or Process map April Define metrics Propose PDSA cycles May Conduct PDSA cycles (if possible) June Submit QI H&P and Reflection summary to MedHub Create Next Steps Summary Present to Clinic EVALUATION: To date, 10 of 19 interns have presented personal QI projects with satisfactory review. These will be completed by February 2014. Of those who have presented personal project, all have satisfactorily completed their SMART Aim statement, balancing measures, systematic data collection, graphed results, PDSA cycle completion for their personal QI projects. Interns have completed 88 % of their assigned IHI modules. The interns will start integration into clinical teams in February with completion of team QI summary report in June 2014. DISCUSSION/REFLECTION/LESSONS LEARNED: The primary lessons learned include: The didactic curriculum should be offered in conjunction temporally with the experiential curriculum (personal QI project) Supervising faculty for QI projects need adequate education in QI Residents should be integrated into projects that align with clinic goals to ensure dedicated QI faculty time and other support needed for success Flexibility and willingness to adapt on the fly are important approaches to implementing new curricular elements in clinical settings In conclusion, involving residents in QI continuity clinic projects offers important opportunities to integrate them into their clinics while providing them with essential QI skills. Steps toward achieving meaningful integration include providing faculty and resident didactics, offering an experiential curriculum and securing support for key faculty supervisors.


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2Fs11606-014-2834-9.pdf

Abstracts from the 37th Annual Meeting of the Society of General Internal Medicine, Journal of General Internal Medicine, 2014, 1-545, DOI: 10.1007/s11606-014-2834-9