The lead article reports results of a randomized trial that tested whether critical time intervention (CTI) would produce an enduring effect in preventing homelessness after inpatient discharge. CTI is a case management intervention to promote housing tenure during this critical transition in two ways: by strengthening a person's long-term ties to services, family, and friends and by providing emotional and practical support until the individual is settled with ongoing community-based services in place. In the study by Daniel B. Herman, D.S.W., M.S., and colleagues, 150 adults with a history of homelessness were discharged from a state psychiatric hospital in the New York City area to transitional residences on the hospital grounds and then to New York City. Participants were then randomly assigned to receive usual care or usual care plus CTI. A CTI worker made tailored arrangements in areas critical for community survival of that individual, and over nine months gradually transferred care to community resources. Participants' housing status was assessed every six weeks for 18 months. Those in the CTI group had significantly less homelessness in the final months of follow-up (odds ratio=.22) (page 713).

This month's issue feature reports from two groups of investigators who turned to large national data sets to answer important questions about antidepressant prescribing trends. In the first study, Rena Conti, Ph.D., and colleagues examined the question of whether U.S. adults overuse antidepressants. Concerns have been raised in recent years about prescription drug overuse, not only because of side effects and adverse health outcomes but also because of the need to make the health care system less wasteful—a policy imperative now that an estimated 57 million additional Americans will become insured under health reform. The data set included adults in the 2005 Medical Expenditure Panel Survey who self-reported antidepressant treatment (weighted N=23,026,608). Overuse was defined as off-label prescribing with limited or no scientific support. Dr. Conti and associates estimated overuse at 20%, with most concentrated in the newer-generation antidepressants (74% of overuse). The authors note that the 20% rate is much lower than the rate of 60% in an earlier report based on 2001 data (page 720). In the second study, Shih-Yin Chen, Ph.D., and Sengwee Toh, Sc.D., analyzed data from millions of office visits to explore changes in antidepressant prescribing practices after the 2003 U.S. Food and Drug Administration advisory that linked antidepressant use with an increased risk of suicidality among children. Data on outpatient visits between 1998 and 2007 with a diagnosis of depression—with and without an antidepressant—were from the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey. Although the authors found downward trends after the 2003 advisory in ambulatory visits with a diagnosis of depression and visits with an antidepressant prescription, children with major depression were no less likely to be prescribed an antidepressant (page 727).

Diffusion of innovations is a challenge in health care. The learning collaborative is a well-established framework for spreading and adapting evidence-based treatments and creating change in organizations to improve care processes. Two reports in this issue describe large, successful learning collaboratives—one to promote integration of care and the other to disseminate supported employment. In the first study, Steven D. Vannoy, Ph.D., M.P.H., and colleagues used mixed methods to evaluate outcomes of a one-year learning collaborative focused on integration of services between community health centers (CHCs) and community mental health centers (CMHCs). The initiative brought together CHC-CMHC pairs to improve treatment of affective disorders in CHCs and improve care of patients at risk of metabolic syndrome at CMHCs. All pairs increased capacity on one or more patient health indicators, and participants' satisfaction with the collaborative process was high (page 753). In the Best Practices column, Deborah R. Becker, M.Ed., C.R.C., and colleagues describe the structure, operation, evolution, and outcomes of a national learning collaborative undertaken by the Johnson & Johnson-Dartmouth Program to disseminate individual placement and support, the evidence-based practice of supported employment for people with severe mental illnesses. The mental health and vocational rehabilitation leaders in 12 states and the District of Columbia have been able to expand services and achieve good outcomes during a period that has seen the erosion of psychosocial services nationwide (page 704).