First-episode psychosis and vocational outcomes: A predictive model.

Most studies on predictors of vocational outcomes are cross-sectional and results are varied. This study aimed to examine the vocational rates of patients with first-episode psychosis (FEP), identify factors predicting a lack of engagement in age-appropriate roles, and evaluate the predictive ability of a model with baseline sociodemographic information and 2-year symptom and functioning trajectories on vocational outcomes. The Singapore Early Psychosis Intervention Program (EPIP) has maintained a standing database on patient clinico-demographic information. The primary outcome, vocational status, was operationalized as "meaningfully employed", that is, being gainfully employed or engaged in an age-appropriate role, and "unemployed". Using logistic regression, the predictive ability of the proposed model was evaluated. Vocational data was available for 1177 patients accepted into EPIP between 2001 and 2012. At the end of two years in the service, 829 (70.4%) patients were meaningfully employed and 348 (29.6%) patients were unemployed. The binary logistic regression model on the prediction of 2-year vocational outcomes yielded an AUC of 0.759 (SE = 0.016, p-value < 0.001). Clinico-demographic risk factors for being unemployed at the end of two years included being Malay, single, and unemployed at baseline; having a longer duration of untreated psychosis (DUP); a diagnosis of schizophrenia, schizophreniform, or delusional disorder at baseline; and belonging to the 'delayed response' or 'slower response and no response' general psychopathology trajectories. We have proposed a model that allows vocational outcomes to be predicted with high specificity. The results of this study will be relevant in developing future intervention models to improve outcomes among FEP patients with different illness trajectories.