2013. Depression Screening: Comparison of Three Methods to Provider Referral in a VA Primary Care Clinic
JW Kanter, HSR&D Center of Excellence, Puget Sound Health Care System, A Epler, Department of Psychological Sciences, University of Missouri-Columbia, EF Chaney, Department of Psychiatry and Behavioral Sciences, University of Washington, SC Hedrick, Center of Excellence, Puget Sound Health Care System, C Liu, Center of Excellence, Puget Sound Health Care System, P Heagerty, Department of Biostatistics, University of Washington, P Lin, Center of Excellence, Puget Sound Health Care System

Objectives: Depression is under-detected and under-treated in primary care settings. The VA has responded to this problem by mandating annual depression screening for all enrolled primary care patients, but it is unknown which of the available screening procedures should be used. This study compared three separate case-finding procedures in terms of ease, intrusiveness, resource use, sensitivity, specificity, coverage, sustainability and overall effectiveness to the usual care system of provider referral.

Methods: We analyzed existing data from a completed randomized team trial of collaborative care depression treatment in which patients were recruited using three separate screening procedures: (1) a two-stage waiting room screening interview (WRI), (2) a two-stage self-report mail survey (MSR), and (3) a waiting room self-report screen consisting of two-items embedded in a larger survey (WRSR). Data on patients identified through these procedures were compared to data on patients referred to the larger study by primary care providers (PCPR).

Results: WRI obtained 25 patients of which 96% were diagnosed with Major Depression; MSR obtained 81 patients (77% diagnosed); WRSR obtained 341 patients (58% diagnosed); and PCPR obtained 276 patients (84% diagnosed). Patients identified through self-report (either MSR or WRSR) were more often single and were less educated than were other patients, and MSR patients also were more likely to have higher chronic disease scores.

Conclusions: Patients were not randomly assigned to screening procedures in this study, restricting causal inferences about the different screening procedures. In general, different screening procedures obtained samples with different characteristics, and some (WRI, PCPR) appeared significantly more sensitive in detecting Major Depression.  Different results across the methods suggest different procedures may be optimal for different resource levels and patient characteristics.

Impact: Determining a case-finding procedure involves compromise and consideration of multiple factors. For example, the most sensitive procedure in this study, a two-stage waiting room interview, also bears the highest cost and intrusiveness and has limited coverage unless it can be sustained. On the other hand, the need for follow-up assessment is reduced by this procedure because it is so sensitive, offsetting follow-up costs of less sensitive procedures.