To the Editor: We thank Drs. Liberman and Kopelowicz for their interest in our article and for their comments. We wish to clarify several issues raised. We employed the following specific criteria for sustained remission. The patient 1) had previously met DSM-III-R or DSM-IV criteria for schizophrenia but not at present, 2) had received a clinical course specifier of "in full remission," 3) had been living independently of supervision by caretakers for the past 2 years, 4) had not had a psychiatric hospitalization for the last 5 years, 5) had presently reported psychosocial functioning within the "normal range," confirmed by a caregiver, and 6) was presently either not taking antipsychotic medications or taking no more than one-half of their highest daily dose since enrollment in our center. Although we did not specifically assess the financial and medication management abilities of our patients, the remitted patients seemed to manage their medications and finances by themselves. We used scores on psychopathology scales as dependent variables rather than as inclusion criteria. There was no difference in scores on positive, negative, and depressive symptoms between the remitted patients and the normal comparison subjects. In response to the suggestion by Drs. Liberman and Kopelowicz, we examined the BPRS scores of our patients. Ten of 12 of remitted patients did not score greater than 4 on any of the BPRS items.

It appears that the group described by Drs. Liberman and Kopelowicz was younger than ours. Their patients were all middle-aged, whereas ours were middle-aged and elderly. A majority of the patients in the elderly age range are unlikely to either go to school or to work half-time or more. We evaluated cognitive performance with an expanded version of the Halstead-Reitan Neuropsychological Test Battery (1, 2). We combined our measurements into a global neuropsychological T score (2) to reduce the number of analyses and to control for type I error. In view of the findings of Drs. Liberman and Kopelowicz, we compared our three groups of subjects on individual cognitive ability areas, including executive functioning, verbal fluency, and verbal working memory. There were no significant differences between the remitted and symptomatic patients in any of these areas, although there were significant differences between the patients and normal comparison subjects on some of them. Of course, as we mentioned in our article, these findings were limited by the small group sizes.

Heterogeneity is a hallmark of schizophrenia. It is to be expected, therefore, that there would be a spectrum of possible outcomes in schizophrenia. At one extreme lies complete recovery of the type described by Drs. Liberman and Kopelowicz, while at the other end is progressive deterioration. Most patients lie between these two extremes. In accordance with the notion that schizophrenia is a neurodevelopmental disorder, with studies showing neurocognitive impairment preceding the onset of symptoms by many years (3), it is not surprising that most patients would still have some cognitive impairments long after the remission of symptoms. Furthermore, the available treatments have little impact on the core cognitive deficits associated with schizophrenia. Clearly, more research needs to focus on developing better interventions that attack the primary neurocognitive abnormalities in this disorder. We hope that the type of complete recovery described by Drs. Liberman and Kopelowicz will one day become a rule rather than an exception.