DESCRIPTION (provided by applicant): Background/Rationale: Depression is a major cause of morbidity and mortality in primary care patients. Randomized trials have shown that the effects on patient health can be ameliorated by instituting collaborative care models (CCM) for depression in primary care, and CCM is currently mandated by VA national policy. However, implementing, sustaining and spreading CCM as part of routine primary and mental health care requires translation from a research innovation to a naturalistically implemented program. Site-level contextual factors introduce potential for drop in CCM effectiveness, yet the potential drop has seldom been investigated. To ensure value for VA's continuing investment in CCM, it is critical to gain a better understanding of if and when a potential drop occurs. This project develops new longitudinal electronic depression care quality measures and uses a rigorous, innovative quasi-experimental approach (stepped design) to evaluate the effectiveness of CCM as naturalistically implemented, sustained, and spread between 2001 and 2010. This period spanned a series of researcher-initiated and policymaker-initiated depression quality improvement efforts culminating in the VA-CCM mandate for national implementation in 2007 including Translating Initiatives in Depression into Effective Solution (TIDES)-CCM. The stepped design makes it possible to integrate more information about site-level context than do alternative non- randomized designs comparing implementation versus non-implementation sites. The design's potential for CCM program evaluation has not been previously assessed. Objectives: The proposed project will 1) develop longitudinal electronic population-based measures of depression quality of care and validate the measures conceptually and operationally using modified Delphi panel judgment by VA and non-VA experts and 2) demonstrate the feasibility of using the stepped design approach with the new measures for evaluating TIDES-CCM. We will use expert panel methods to assess the overall acceptability and feasibility of the measures and the stepped design for evaluating CCM, future implications for VA's CCM program implementation, and optimal statistical features for broader use. Methods: We will develop and validate longitudinal electronic population-based depression care quality measures. These measures will be used with contextual features captured in the TIDES-CCM historical implementation database to test the utility and validity of a stepped design for comparing TIDES-CCM primary care practice sites to usual care sites on depression care quality. The stepped design, based on standard statistical principles and established stepped wedge design applications, has the potential to maximize the precision of comparisons between intervention and control sites when randomized designs are not feasible or not desirable. Expert panel judgment will be used throughout the project. The development of measures in this project, using expert panel and data validation methods, will meet a widely-recognized need for longitudinal measures of the course of depression clinical care from screening through treatment. If successful, the results of this project will inform the sciences of depression qualit measurement, comparative evaluation design for naturalistically spread interventions, and on-going national implementation of CCM.

This proposed project will provide critical information about the comparative effectiveness of collaborative care models (CCM) as implemented and spread by VA VISNs, medical centers, and primary care practices that will help inform national on-going implementation efforts. It will provide the basis for a new generation of population-based depression care quality measures based on electronic VA data and it will inform the scientific field of program evaluation in VA by testing a design that makes full use of VA's electronic data system and measurement capabilities. Project results will inform the sciences of depression quality measurement, comparative evaluation design for naturalistically spread interventions, and provide important information for the on-going national implementation of CCM within the VA.