The association between payment model and specialist physicians' selection of patients with diabetes: a descriptive study.

OPEN Research The association between payment model and specialist physicians’ selection of patients with diabetes: a descriptive study Amity E. Quinn PhD, Alun Edwards MB BChir, Peter Senior MBBS PhD, Kerry A. McBrien MD MPH, Brenda R. Hemmelgarn MD PhD, Marcello Tonelli MD SM, Flora Au MA, Zhihai Ma MSc, Robert G. Weaver MSc, Braden J. Manns MD MSc Abstract Background: As the number of people with chronic diseases increases, understanding the impact of payment model on the types of patients seen by specialists has implications for improving the quality and value of care. We sought to determine if there is an association between specialist physician payment model and the types of patients seen. Methods: In this descriptive study, we used administrative data to compare demographic characteristics, illness severity and visit indication of patients with diabetes seen by fee-for-service and salary-based internal medicine and diabetes specialists in Calgary and Edmonton between April 2011 and September 2014. The study cohort included all newly referred adults with diabetes (no appointment with a specialist in prior 4 yr). Diabetes was identified using a validated algorithm that excludes gestational diabetes. Results: Patients managed by salary-based physicians (n = 2736) were sicker than those managed by fee-for-service physicians (n = 21 218). Patients managed by salary-based specialists were more likely to have 5 or more comorbidities (23.0% [n = 628] v. 18.1% [n = 3843]) and to have been admitted to hospital or seen in an emergency department for an ambulatory care sensitive condition in the year before their index visit, probably reflecting poorer disease control or barriers to optimal outpatient care. A higher proportion of visits to salary-based physicians were for appropriate indications (65.2% [n = 744] v. 55.6% [n = 5553]; risk ratio 1.17, 95% confidence interval 1.09–1.27). Interpretation: Salary-based specialists were more likely to see patients with a clear indication for a specialist visit, while fee-forservice specialists were more likely to see healthier patients. Future research is needed to determine if the differences in types of patients are attributable to payment model or other provider- or system-level factors. A s the number of people with chronic diseases increases, it is critical to determine the optimal models to improve the quality and value of chronic disease care. Chronic disease management models have focused on the role of primary care,1 although specialists are also key members of the chronic care team, providing additional support and care to patients with more complex health care needs.2 Fee for service (FFS) is the dominant physician remuneration model in Canada: 72.1% of physician payments in Canada3 are reimbursed under FFS. FFS financially rewards physicians who have more patient visits and more clinical activity. Studies in both primary care and specialty care settings have found that FFS payment results in increased health care utilization. 4,5 Understanding the © 2019 Joule Inc. or its licensors Competing interests: Braden Manns, Brenda Hemmelgarn, Marcello Tonelli, Kerry McBrien, Alun Edwards and Peter Senior are employed under an academic alternative relationship plan. Brenda Hemmelgarn reports receiving grants from Amgen, outside the submitted work. Peter Senior reports receiving personal fees from Abbott, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck and Sanofi and research support for his institution from AstraZeneca, Novo Nordisk, Prometic and Sanofi, outside the submitted work. He has served as an officer with Diabetes Canada, outside the submitted work. No other competing interests were declared. This article has been peer reviewed. Correspondence to: Braden Manns, CMAJ Open 2019. DOI:10.9778/cmajo.20180171 CMAJ OPEN, 7(1) E109 OPEN Research impact of payment model on the types of patients seen by specialists is important, because health care payers presumably want to prioritize care for patients with the highest needs who are at the highest risk of complications and hospital admissions and therefore have the highest chance of benefiting from specialist care. Insofar as physicians can select who they provide care to, FFS payment might incentivize specialists to select more complex patients who need more health care,6 or it might induce selection of patients with less complexity (because they can see patients in less time) although the clinical value of such visits might be uncertain. In primary care, physicians with more complex patients have been found to be less likely to select a payment model other than FFS,7 but it is unclear how payment model might affect patient selection in specialty care. We aimed to compare demographic characteristics, illness severity and visit indication of patients with diabetes newly referred to FFS and salary-based specialists to determine if there was an association between specialist physician payment model and patients seen. Methods Overview A salary-based remuneration model called Academic Alternative Relationship Plans (AARP) was implemented across Alberta in 2004 with the goal of promoting innovative ways to provide patient care in a more efficient and accessible manner (for instance, outreach clinics to First Nations and other rural communities), and it is currently used by over 700 specialist physicians (one-quarter of all medicine specialist physicians in Alberta). AARP pay physicians on a contractual basis and provide a mechanism to compensate physicians for clinical, administrative, teaching and research contributions.8 While physicians remain independent contractors, the AARP model is most similar to a salary. When specialist physicians in both groups receive consultation requests, they can choose to provide written feedback to the referring physician (not reimbursed by FFS) but not see the patient formally, book a phone call to discuss the patient with the referring physician (a reimbursable FFS service), or book an in-person visit for the patient (reimbursable at a higher FFS rate). Data sources We used the Interdisciplinary Chronic Disease Collaboration Data Repository,9 which includes laboratory and administrative health data (including vital statistics; prescription drug data; physician claims; data on hospital admissions, emergency department visits and outpatient visits; and all health care costs) for all Albertans from 1994 to 2015. This data set has been used for many observational studies10–13 and for assessing outcomes in a randomized controlled trial of over 20 000 people.14 The study cohort included newly referred adults with diabetes, identified using a validated algorithm based on 2 or more physician claims for diabetes. 15,16 Compared with chartconfirmed diagnoses in primary care, this definition is relatively accurate (sensitivity and specificity of 86% and E110 CMAJ OPEN, 7(1) 97%).10 All patients with diabetes who were seen by inte (...truncated)