Rationale: Criteria for remission in schizophrenia have recently been presented. It is unclear how many acutely ill patients meet these criteria and how they compare with previously suggested definitions.
Objectives and methods: We re-analysed seven anti-psychotic drug trials (n = 1,708) of patients with schizophrenia to find out how many met the new remission criteria and their single components, how many met two previously used remission criteria, and how many met simpler measures of response (at least 50% Brief Psychiatric Rating Scale [BPRS] reduction, a Clinical Global Impressions [CGI] improvement score of at least 'much better' or a CGI severity score of 'mild or better').
Results: Thirty-seven percent/41% (last observation carried forward [LOCF]/completer analysis [CO]) of the initially acutely ill patients with positive symptoms met the severity criteria of remission at 4 weeks, and 27%/52% (worst case/CO) met the severity and time criteria at 1 year. Only 13%/21% (LOCF) met the severity criteria at 4 weeks/1 year when an item threshold 'at best very mild symptoms' was applied, and almost no patients were absolutely symptom-free. The psychotic symptoms component was more difficult to achieve than the negative component. The criteria were more stringent than 'at least 50% BPRS reduction' and than 'CGI improvement score of at least much better.' However, the definition 'CGI severity score mild or better' was of a stringency similar to the new remission criteria, which probably explains why fewer patients met previously defined criteria that included this scale.
Conclusion: The new remission criteria proved to be an achievable goal for clinical trials. A consensus on the application of their time component is still needed.