Dr. Cook and Colleagues Reply

To the Editor: We thank Dr. Rosenheck for his careful reading of our article, and in response to his question, we clarify that the definition of services integration that appears in the text of our article (p. 1950) is correct. The definition that appears in the footnote to Table 2 (p. 1954) lacks the detail that appears in the text. We apologize for any confusion this may have caused. Further, we agree with the excellent points that Dr. Rosenheck raises regarding the measurement of integrated mental health care and supported employment services. The analyses he suggests, including an examination of the effect of modest incremental change in the level of services integration on employment outcomes, would be of great interest to the field. Unfortunately, the measure of services integration that we used is not appropriate for such tests because it was created post hoc as a means of characterizing the variety of interventions included in the demonstration program based on available information. As such, we did not test its reliability or validity. The distribution of services integration with this measure was nonnormal, with modes at 0% and 100%. We feel that although the measure was able to differentiate high and low levels of integration of services in the Employment Implementation Demonstration Program multisite study, we cannot be sure that it taps the latent construct of integrated clinical and vocational services. Following Dr. Rosenheck’s line of reasoning, we agree that development of a more rigorous measure of services integration would be of great value to our field.