Transdiagnostic Clinical Global Impression Scoring for Routine Clinical Settings.

behavioral sciences Concept Paper Transdiagnostic Clinical Global Impression Scoring for Routine Clinical Settings Boadie W. Dunlop *, Jaclyn Gray and Mark H. Rapaport Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, 12 Executive Park Drive, 3rd Floor, Atlanta, GA 30329, USA; (J.G.); (M.H.R.) * Correspondence: ; Tel.: +1-404-727-8474; Fax: +1-404-727-3700 Received: 13 February 2017; Accepted: 21 June 2017; Published: 27 June 2017 Abstract: Although there is great interest in the improving the ability to track patients’ change over time in routine clinical care settings, no standardized transdiagnostic measure is currently available for busy clinicians to apply. The Clinical Global Impression (CGI) scales are simple measures widely used as outcomes in psychiatric clinical trials. However, the CGI suffers from poorly defined scoring anchors. Efforts to improve the anchors by enhancing the anchor descriptions have proven useful but are limited by being disease-specific, thereby acting as a barrier to the routine clinical adoption of the CGI. To inform the development of more broadly applicable CGI scoring anchors, we surveyed 24 clinical trial investigators, asking them to rank-order seven elements that inform their CGI-Severity (CGI-S) scoring. Symptom severity emerged as the most important element in determining CGI-S scores; the functional status of the patient emerged as a second element. Less importance was given to self-report symptom scores, staff observations, or side effects. Relative rankings of the elements’ importance did not differ by investigators’ experience nor time usually spent with patients. We integrated these results with published illness-specific CGI anchors to develop the Transdiagnostic CGI (T-CGI), which employs standardized scoring anchors applicable across psychiatric illnesses. Pending validity and reliability evaluations, the T-CGI may prove well-suited for inclusion in routine clinical settings and for incorporation into electronic medical records as a simple and useful measure of treatment efficacy. Keywords: mood disorders; psychotic disorders; anxiety disorders; pharmacotherapy; psychotherapy; rating scale 1. Introduction Fundamental to the treatment of psychiatric disorders is the ability of clinicians to determine whether an intervention is helping a patient recover from their illness. In clinical trials, accurate and consistent measurement of improvement is crucial for determining the potential efficacy of a new treatment. However, routine clinical settings have lagged in adopting standardized measures of change, despite the demonstrated value of measurement-based care for improving patient outcomes [1–3]. Less than 20% of psychiatrists routinely use symptom rating scales as part of their clinical practice [4]. This low rate of adoption stems from the time-consuming and disease-specific rating scales that are not practical options in busy clinical settings, as well as clinicians’ perceptions that existing rating scales have limited clinical utility [4,5]. The development of a transdiagnostic measure that easily and reliably captures illness severity and change over time would address this unmet clinical need. Historically, symptom rating scales have been the primary tools used to assess efficacy, because they have well-established psychometric properties and evaluate a range of symptoms. However, several concerns about the reliability and validity of the various measures commonly employed in clinical trials exist, and rating scale scores may be confounded by side effects of medications Behav. Sci. 2017, 7, 40; doi:10.3390/bs7030040 www.mdpi.com/journal/behavsci Behav. Sci. 2017, 7, 40 2 of 10 (such as changes in appetite or sleep) that are scored as new symptoms by masked raters or by patients on self-report questionnaires [6,7]. Moreover, meaningful assessment of treatment efficacy requires consideration of factors beyond symptom change alone, such as quality of life and level of functioning [8,9]. How patients perceive their own quality of life is only partially explained by symptom rating scale scores [10], and the scales fail to fully capture changes in functional status [11]. Although functioning typically improves with symptom reduction, these concepts are not always concordant, and functional changes [12] and quality of life gains [13] often lag behind symptom change. The Clinical Global Impression (CGI) scales were developed as simplified global measures to reflect the clinician’s overall impression of a patient’s condition (CGI-Severity, CGI-S, rated 1–7 from “normal” to “among the most extremely ill” and change over time (CGI-Improvement, CGI-I, rated 1–7 from “very much improved” to “very much worse” [14]. The appeal of the CGI measures is their easy translation to clinical care; they represent a common heuristic used by clinicians in evaluating patients and making treatment decisions [15]. However, an important limitation of the original CGI scales is their lack of well-defined anchor points [15]. Revisions of classic symptom rating scales have identified ambiguous or absent anchor point descriptions as a significant source of unreliability in scoring, addressed by adding explicit descriptions for the numerical scores [16–18]. For the CGI, the rater’s experience may also be a source of variability, given the scale’s instruction to “consider his [sic] total clinical experience with the given population” in making the rating [14]. More detailed anchor point descriptions for the CGI scales are therefore necessary to improve inter-rater reliability, and have been developed by several groups for illness-specific versions of the CGI [19–26]. In addition to concerns about inter-rater reliability, psychometric evaluations of the CGI have identified potential problems with validity, scaling, and test-retest reliability, in some disease populations [27–30]. Despite these concerns, Leon and colleagues found the CGI measures to have good internal consistency and concurrent validity, which could be further improved by more rigorous rater training and more well-structured anchor points [31]. Indeed, the CGI scales have been used as the primary outcome measure in clinical trials for a variety of conditions, including major depression [32], social phobia [33], post-traumatic stress disorder [34], panic disorder [35], binge-eating disorder [36], and complicated grief [25,26]. A recent modification of the CGI that aimed to improve scoring reliability in clinical trials, the Structured Interview Guide for Global Impressions [37], requires approximately ten minutes to complete. Although ten minutes to administer a scale during a clinical trial visit is not excessive, this would amount to 33–50% of the time allotted for outpatient psychiatric appointments, making it impractical for routine clinical settings. In summary, the CGI scales appear to have potential utility as a rating tool for routine (...truncated)