Six articles in this issue report on studies that involved participants with depression or bipolar disorder. In a two-part article, a team of Veterans Affairs researchers describes the development of a collaborative care model for persons with bipolar disorder (page 927) and presents data on outcomes over three years for 306 veterans randomly assigned to the intervention or to usual care. The cost-neutral intervention significantly reduced the number of weeks spent in an affective episode (page 937). An analysis of data for 1,531 participants in PRISM-E (Primary Care Research in Substance Abuse and Mental Health for the Elderly) found that depression outcomes at three and six months were similar for patients in two care models—an integrated model in which mental health services were co-located in primary care settings and a specialty referral model in which care was provided in a physically separate clinic (page 946). Of the first 1,000 patients to enroll in STEP-BD (Systematic Treatment Enhancement Program for Bipolar Disorder), 248 were in a depressive phase at study entry. Among the 60 percent of this group who received any psychotherapy over the next year, those with severe symptoms benefited most (page 959). Ten- to 15-year follow-up data are presented on use of mental health services by more than 300 individuals, nearly two-thirds of whom had a childhood diagnosis of major depression or an anxiety disorder. Compared with those who had no baseline mental disorders, participants with baseline depression were 13 times more likely to have received treatment in the follow-up period and those with anxiety disorders were six times more likely. Childhood anxiety was an especially strong predictor of later long-term treatment (page 966). The nine-item depression module of the Patient Health Questionnaire was an effective means of detecting depression when used to screen more than 3,400 low-income patients at a primary care clinic serving Chinese Americans, a group known to underutilize mental health services (see page 976).

Two studies—one in New York City and the other in San Francisco—examined the experiences of formerly homeless persons who entered housing programs. Carole E. Siegel, Ph.D., and her colleagues used statistical modeling techniques and propensity scoring to closely examine outcomes over 18 months for 157 persons entering either supported housing or more service-intensive community residences. Tenure in housing did not differ by housing type. Many in supported housing whose clinical profiles suggested more appropriate placement in community residences were successfully housed (page 982). In San Francisco, Tia E. Martinez, J.D., and Martha R. Burt, Ph.D., examined use of acute health care services among 236 formerly homeless individuals with dual diagnoses for two years after they entered a "low demand" housing program, which did not require treatment participation or abstinence. Providing permanent supportive housing to this group significantly reduced their use of emergency department and inpatient services (page 992). In both studies, substantial proportions of participants remained housed.

A survey of 51 state Medicaid agencies that sought to determine the extent to which states were using managed care programs for behavioral health services found that 31 states managed both mental health and substance abuse treatment and five managed only mental health care. M. Susan Ridgely, J.D., and Margaret A. Maglione, M.P.P., report that in 2000 a health maintenance organization model was most commonly used for managing behavioral health care, but 40 percent of plans were specialty carve-outs. Benefit designs were more inclusive than some advocates had feared, but large numbers of persons with the most severe illnesses were receiving care through fully capitated at-risk plans. The authors conclude by urging public-sector mental health authorities to aggressively seek opportunities to influence Medicaid policy at state and federal levels (page 1000).

• A second study by PRISM-E researchers reported in this month's issue evaluated two interventions that reduced at-risk drinking among elderly persons (page 954)

• The Law & Psychiatry column examines recent cases in which college students who sought treatment for depression were barred from returning to campus (page 914)

• An analysis of the Vermont Cancer Registry found that incidence among men under age 50 with serious mental illness was more than six times higher than among their counterparts in the general population (page 1032).