Predictive validity of clinical variables in the "at risk" for psychosis population: international comparison with results from the North American Prodrome Longitudinal Study.

BACKGROUND

There has been recent optimism with regard to improving the predictive validity of those individuals who develop a psychotic disorder from the "Ultra High Risk" (UHR) or putatively prodromal population using combinations of clinical variables. We aimed to test the recent results from a large collaborative consortium in an independent cohort from the PACE (Personal Assistance and Clinical Evaluation) clinic in Australia.

METHOD

The North American Prodrome Longitudinal Study (NAPLS) consortium study reported 5 important clinical predictive variables within their US sample of UHR patients: genetic risk with functional decline; high unusual thought content score; high suspicion/paranoia score; low social functioning and history of substance abuse. We examined the predictive validity of these variables using data from a cohort of 104 UHR patients from the PACE clinic in Melbourne, Australia. Cox regression was used to explore the relationship between these variables at baseline and transition to psychosis by 28months.

RESULTS

Three of the five variables were found to be associated with transition in our sample: high unusual thought content scores; low functioning; and having genetic risk with functional decline. A combination of two out of three of these features produced a reasonable predictive validity (positive predictive value (PPV) 65.4%, sensitivity 37.3%, and specificity 87.2%) but with overall lower PPVs than that reported by the NAPLS consortium.

CONCLUSIONS

Three out of five of the identified clinical predictors for transition to psychosis from the NAPLS study were replicated in an independent sample. Using a combination of clinical variables the predictive validity of determining whether a UHR individual develops a psychotic disorder was improved above UHR criteria alone. Although psychosis prediction is improved using this model, the probability of a person not developing psychotic disorder is still quite high at 35%.