Academic psychiatry leaders are all too familiar with being at the bottom of the food chain when working with carved-out behavioral health organizations. In this issue Fagan and colleagues describe the efforts of the Johns Hopkins department of psychiatry to join the handful of academic psychiatry departments that have taken capitated financial risk and medical management responsibility for a population.

The department was wise to "start slow and small," allowing a period in which to gear up service and management systems to deal with a capitated population and to help providers understand the incentives that guide the provision of care under capitation. It was also wise to start with a "narrow risk corridor," and the department was fortunate to have the financial cooperation of Johns Hopkins Medical Services Corporation (MSC).

Before taking responsibility for a population's care, it is vital to have data on previous use and costs as well as demographic data. The authors had data for the previous two years for 7,500 of their 22,000 patients, although it is not clear whether they obtained data for the entire population from the organization that previously held the contract. Thus it is hard to accept on faith that the data for 1998 to 2000 demonstrate improved management. Also, the authors acknowledge that two years is not long enough to determine whether the decreases in service use and cost were due to skill or to luck. Finally, despite the large positive differences between income and expenses under this contract, the department and MSC each made only $18,000 the first year and $29,000 the second. It is important to know whether the contract brought other advantages.

More demographic data on this atypical population would also have been informative. The population is composed of dependents of active-duty military personnel. This population tends to be relatively healthy and to use psychiatric services infrequently. It is unlikely to include a substantial Medicaid or disabled Medicare component. Also, it is unlikely that the same 22,000 individuals were covered by this contract for the entire period.

The decision to integrate mental health care with care provided by patients' primary care physicians was important. However, it is not clear who these physicians were and what incentives they were given to work closely with behavioral health providers.

Nevertheless, Fagan and colleagues are to be congratulated for accepting and successfully implementing a capitated mental health contract for an initial two years. Their department now has more of its fate in its own hands rather than in the hands of managed care organizations.