For a long time it was a problem of treatment research in schizophrenia that uniformly accepted definitions of response, remission and recovery were not available. The presentation will summarize recent reports on these issues and will come up with a number of suggestions.

Review of recent publications.

Response can be defined as a clinically meaningful improvement of the patient's psychopathology irrespective of whether he is still symptomatic at the end or not. When the BPRS or the PANSS are used for definitions of response, a cutoff of at least 50% reduction of the baseline score should be used for acutely ill, non-refractory patients and a cutoff of at least 25% reduction for refractory patients. A table presenting responder rates in 25% steps covering the whole range up to 100% has been suggested.

Remission is a state in which the patient is free of clinically significant symptoms. A definition based on 8 PANSS items rated mild or better for a duration of at least 6 months has recently been presented. The advantage of these remission criteria is that in contrast to the response cutoffs they show how many patients are still symptomatic at the end of a study or not. Their disadvantage is that they do not reflect the amount of change.

Both remission and responder rates could be indicated in future studies. The next challenges are the development of universally accepted definitions of recovery and relapse of schizophrenia.