Utilising symptom dimensions with diagnostic categories improves prediction of time to first remission in first-episode psychosis.
Any type of content formally published in an academic journal, usually following a peer-review process.
There has been much recent debate concerning the relative clinical utility of symptom dimensions versus conventional diagnostic categories in patients with psychosis. We investigated whether symptom dimensions rated at presentation for first-episode psychosis (FEP) better predicted time to first remission than categorical diagnosis over a four-year follow-up. The sample comprised 193 FEP patients aged 18-65years who presented to psychiatric services in South London, UK, between 2006 and 2010. Psychopathology was assessed at baseline with the Positive and Negative Syndrome Scale and five symptom dimensions were derived using Wallwork/Fortgang's model; baseline diagnoses were grouped using DSM-IV codes. Time to start of first remission was ascertained from clinical records. The Bayesian Information Criterion (BIC) was used to find the best fitting accelerated failure time model of dimensions, diagnoses and time to first remission. Sixty percent of patients remitted over the four years following first presentation to psychiatric services, and the average time to start of first remission was 18.3weeks (SD=26.0, median=8). The positive (BIC=166.26), excited (BIC=167.30) and disorganised/concrete (BIC=168.77) symptom dimensions, and a diagnosis of schizophrenia (BIC=166.91) predicted time to first remission. However, a combination of the DSM-IV diagnosis of schizophrenia with all five symptom dimensions led to the best fitting model (BIC=164.35). Combining categorical diagnosis with symptom dimension scores in FEP patients improved the accuracy of predicting time to first remission. Thus our data suggest that the decision to consign symptom dimensions to an annexe in DSM-5 should be reconsidered at the earliest opportunity.