To the Editor: Dr. Liberman suggests that "the surprising and counterintuitive finding of poorer outcomes in the patients in the active control group (versus the group with naturalistic follow-up and the group undergoing cognitive adaptation training) could be explained on the basis of bias by unblinded assessors who were also involved as clinicians in the interventions." We, too, were somewhat surprised by this observation. However, the assessments were conducted by research personnel who were unaware of the subjects’ treatment groups, so that bias seems an unlikely explanation.

Dr. Liberman also says that it is possible that the concerns of patients were referred to indifferent treatment teams and that, because our clinicians did not help directly, referral to their clinical treatment team may have been experienced by the patients as rejection. Although this is a possibility, our therapists were trained to conceive of any interaction with the patients as therapeutic and to address their concerns by making note of the seriousness of their problems and by providing the telephone numbers of caseworkers and clinics. In the case of severe symptom exacerbation, therapists for the control group and the group undergoing cognitive adaptation training actually encouraged the patients to contact the clinic. We think that the poorer reported results for the active control group in our study had to do with these prompts for the patients to contact the treatment team. It was likely that contact was made with the team and that the exacerbation of symptoms was noted in the charts of these patients. In the group with naturalistic follow-up only, although significant exacerbations may have occurred between assessment visits without suggestions to actively seek out the treatment team, it is possible that these exacerbations were managed by other means and happened without documentation of their occurrence.

In our most recent study of stable outpatients, therapists made fewer referrals to the clinical treatment team for severe exacerbations. In that study, the patients in this same active control condition as the one described in the article performed similarly to those in naturalistic follow-up. These recent data do not support the notion that our therapists were behaving in ways that caused the patients to feel rejected.

Finally, Dr. Liberman suggests that some information on what happens when supports are discontinued should be made available. He makes the important point that, in other types of programs, such as the Program of Assertive Community Treatment, which rely on supports, regression to baseline levels of functioning and symptom levels is common when treatment ends. Compensatory treatment provides both supports and training in the use of supports, and anecdotal information suggests that approximately one-half of the patients who undergo cognitive adaptation training are able to maintain the gains made during treatment over the next 6 months. However, as in other supportive treatments, many patients return to baseline levels of functioning when treatment is stopped. We have several studies ongoing or proposed that will attempt to identify the characteristics of patients who are able to maintain gains in treatment versus those who need continued assistance.