Response to White

To the Editor: Dr. White correctly states that collaborative care models have traditionally involved collaboration between primary care providers, care managers, and consulting psychiatrists. This collaboration typically includes weekly supervisory meetings between care managers and psychiatrists to discuss patients failing treatment, followed by treatment recommendations to the primary care provider when appropriate. We agree with Dr. White that psychiatric supervision and consultation are integral elements of effective mental health programs. In the telemedicine-based collaborative care arm of our study, we adapted this traditional model for small rural primary care clinics by establishing an off-site multidisciplinary depression care team (psychiatrist, psychologist, pharmacist, and care manager) to collaborate with the on-site primary care providers.

Dr. White is also correct that in the practice-based collaborative care intervention arm of our study, depression treatment was delivered by an on-site primary care provider and an on-site care manager, without the involvement of a consulting psychiatrist. This comparison group was chosen for three reasons. First, most primary care practices do not have access to an on-site psychiatrist. Second, two high-quality randomized trials had previously demonstrated that depression outcomes can be improved in small primary care clinics lacking on-site mental health specialists by training depression care managers (without psychiatric supervision) to support primary care providers (1, 2). Third, this comparison group replicated the Depression Health Disparities Collaborative that was being disseminated in federally qualified health centers by the Health Resources and Services Administration (HRSA). For the depression collaborative, HRSA specifically recommended: “Establish linkages with key specialists to assure that primary care providers have access to expert support,” and our practice-based care managers were encouraged to use mental health resources available in the community (e.g., community mental health centers). However, for federally qualified health centers located in medically underserved areas, access to psychiatric care is challenging, and obtaining psychiatric consultation and care manager supervision is especially difficult.

Because this was a pragmatic rather than an explanatory trial, we cannot isolate the mechanism(s) of action that led to better clinic outcomes in the telemedicine-based arm compared with the practice-based arm. Certainly, the depression care manager in the telemedicine-based arm exhibited higher fidelity to the care manager protocol than the care managers in the practice-based arm. Higher fidelity may well have been better in the telemedicine-based arm because of the psychiatric supervision and the expectation that the care manager would be required to report patient outcomes to the psychiatrist on a weekly basis. Ideally, we believe that supervision should focus on fidelity to a specific evidence-based treatment protocol and should use data regarding process and outcomes of care to systematically identify patients who are not getting better. Unstructured supervision, like unstructured treatment, may have little clinical benefit. Use of a decision support system (such as NetDSS) facilitates this type of process and outcomes-focused supervision. In addition, both our experience and our view of the evidence to date indicate that such structured supervision is more easily accomplished in centralized treatment programs than in locally grown ones.