Letter

In Reply: We appreciate the comments of Drs. Rosenheck, Neale, and Mohamed on our study of client and program factors associated with clients' entry into the VA health system's ACT program (MHICM) during fiscal year (FY) 2004. It is important to recognize the VA's recent expansion of this program and the ongoing implementation of a pilot program in rural sites. Also, it is helpful to consider the VA's specialized case management services for homeless veterans, which may explain the lower likelihood of initiation of sustained ACT services in FY 2004 among veterans who were homeless.

We note that the study criteria for measuring program eligibility—mental illness diagnosis and high inpatient psychiatric utilization in the prior year—were specified in VHA Directive 2006-004 regarding the MHICM program. It was not possible for us to assess the three other criteria listed in the directive—having severe functional impairment, being inadequately served by conventional clinic-based treatment, and being clinically appropriate for outpatient treatment. We agree that administrative data do not fully capture the complexity of clinical decisions affecting program initiation. It is therefore likely that some of the individuals identified in our study did not satisfy all five eligibility criteria. However, the VA directive specifies that "programs are intended … for veterans who meet all of the … five criteria." Therefore, the suggestion that the study failed to include some veterans who could be considered program eligible but who had not been recently hospitalized seems to reflect concerns regarding the VA directive rather than the study's reliance on administrative data. We understand that sometimes clinical assessments of patients' needs may supersede national directives.

Research is needed to assess the correspondence of indicators drawn from administrative data with on-the-ground clinical assessments. Our study is part a larger evaluation of VA program inclusion criteria, and we operationalized eligibility measures using the hospitalization criterion specified in the VA directive. We look forward to publication of the study that reports MHICM program use by a substantial number of elderly veterans. Certainly, further work is needed to evaluate changes in factors associated with ACT program initiation in the VA health system in recent years given ongoing program expansion and development.