Does treatment delay in first-episode psychosis really matter ? 1 S. M. HARRIGAN, 2 P. D. M C GORRY AND H. KRSTEV From the University of Melbourne, Department of Psychiatry, Early Psychosis Prevention and Intervention Centre (EPPIC), Melbourne, Victoria, Australia ABSTRACT Background. Relatively few predictors of outcome in first-episode psychosis are potentially mal- leable and duration of untreated psychosis (DUP) is one. However, the degree to which DUP is mediated by other predictors of outcome is unclear. This study examines the specific effects of DUP on 12-month outcome after adjusting for effects of potential confounders and moderating variables. Method. The sample comprised 354 first-episode psychosis patients followed up 12-months after remission/stabilization of their psychotic symptoms. Outcome measures included functional out- come, severity of positive symptoms and negative symptoms. Hierarchical multiple regression as- sessed whether DUP significantly predicted 12-month outcome after adjusting for other predictors. Contrast analysis further clarified the differential effects of DUP on 12-month outcome. Results. DUP remained a significant predictor of outcome after adjusting for the effects of other variables. This finding remained robust for the subset of patients with schizophrenia or schizo- phreniform disorder. Functional outcome appeared to decline substantially even after very short treatment delays (>7 days), with more gradual deterioration in functioning until very long DUP (>1 year). Good outcome was variably associated with good pre-morbid adjustment, female gender, diagnosis of affective disorder, short duration of prodromal symptoms, and treatment within the Early Psychosis Prevention and Intervention model in contrast to other models of care. Conclusions. DUP consistently predicts outcome independently of other variables, and is not simply a proxy for other factors. As one of the few potentially malleable factors influencing outcome, DUP could prove to be a target for secondary preventive efforts in early psychosis. INTRODUCTION It is highly likely that an association exists be- tween duration of untreated psychosis (DUP) and outcome in first-episode psychosis (Helga- son, 1990 ; Wyatt, 1991 ; Loebel et al. 1992 ; Beiser et al. 1993 ; Larsen et al. 1996 ; Scully et al. 1997 ; Haas et al. 1998; McGlashan, 1999; Bottlender et al. 2000 ; Drake et al. 2000 ; Larsen et al. 2000 ; Quinn et al. 2000 ; Meagher et al. 2001 ; Norman & Malla, 2001), though this is not beyond dispute (Barnes et al. 2000; Craig et al. 2000; Ho et al. 2000; Ho & Andreasen, 2001). Moreover, DUP is one of the few risk factors for poor outcome that has some potential for malleability through early detection strategies. Timing of treatment is not the only malleable factor influencing out- come ; quality of treatment has also been associ- ated with improved outcomes (McGorry et al. 1996) although evidence suggests that its ca- pacity to influence outcome may be limited where treatment delays are very short or very long (Carbone et al. 1999). Identification of these potentially malleable risk factors for poor out- come contrasts with the relatively fixed nature of most predictors such as gender, age of onset of 1 Presented at the Second International Conference on Early Psychosis, International Early Psychosis Association (IEPA), in New York, NY, USA on 30 March 2000. 2 Address for correspondence : Susy M. Harrigan, Department of Psychiatry, University of Melbourne, Early Psychosis Prevention and Intervention Centre, 35 Poplar Road, Parkville, Victoria 3052, Australia. Psychological Medicine, 2003, 33, 97–110. f 2003 Cambridge University Press DOI : 10.1017/S003329170200675X Printed in the United Kingdom 97