Review Article Measurement and reporting of the duration of untreated psychosis Matthew Large, 1 Olav Nielssen, 1,2 Tim Slade 3 and Anthony Harris 4 1 Private Practice, Sydney, 2 Clinical Research Unit for Anxiety Disorders, School of Psychiatry, University of New South Wales at St Vincent’s Hospital, Darlinghurst, and 3 National Drug and Alcohol Research Centre, University of New South Wales, Sydney, and 4 Discipline of Psychological Medicine, University of Sydney, and Brain Dynamics Centre, Westmead Hospital, Westmead, New South Wales, Australia Corresponding author: Dr Matthew Large, PO Box 110, Double Bay, 1360 NSW, Australia. Email: mmbl@bigpond.com Received 04 October 2007; accepted 22 May 2008 Abstract Aim: The aim of this study was to investigate the demographic, illness and methodological factors associ- ated with mean and median duration of untreated psychosis (DUP). Methods: A systematic review and meta-analysis of the published studies of DUP and an examination of available DUP distributions. Results: DUP was longer in samples with a higher proportion of patients with schizophrenia and was shorter in samples that included affective psy- chosis. Sex, age, and the methods of measuring the onset and end-point of DUP and the type of service in which the studies were performed did not contribute to the heterogeneity of the mean or median DUP values. Mean DUP is significantly prolonged by a small number of patients, and the median DUP is a poor indicator of the rate at which patients present. Conclusions: The DUP of patients with affective and non-affective psy- chosis should be examined separately in order to make measures of DUP more meaningful and comparable, and DUP should be reported using more comprehensive measures. We suggest a method of reporting DUP based on the rate of presentation of first-episode psychosis patients rather than the length of DUP. Key words: duration of untreated psychosis, first episode psychosis, meta analysis. INTRODUCTION Interest in the duration of untreated psychosis (DUP) has grown from the finding that prolonged DUP is associated with a poorer prognosis after treatment 1,2 and an increased risk of suicide 3 and possibly serious violence 4 during the first episode of psychosis (FEP). In an influential commentary on the concept of DUP, Norman and Malla noted the wide range of mean DUP values in published studies 5 and dis- cussed the possible effect of different definitions of DUP on the reported durations. They also noted that the distribution of DUP values within studies is con- sistently skewed as a result of a small number of subjects with longer DUP, thus raising the possibility that differing numbers of subjects with much longer DUP may contribute to the variation in reported mean DUP. A range of clinical factors have also been associ- ated with the variation in DUP, including whether the patients were recruited by specialist first-onset and early intervention in psychosis services 3,5 and the characteristics of the patients. Younger patients 6 and males 7 have been reported to have a longer DUP, and patients with an affective psychosis have a shorter DUP than patients with schizophrenia and other forms of non-affective psychosis. 8 There have been calls for the adoption of stan- dardized instruments to assess DUP, in the hope that this will allow comparisons between studies of DUP conducted in different settings. 9 In this paper, we examine the extent to which differences in methods and patient details contribute to the het- erogeneity of reported mean and median DUP, and the degree to which mean DUP varies because of differing numbers of patients with a very prolonged DUP. Early Intervention in Psychiatry 2008; 2: 201–211 doi:10.1111/j.1751-7893.2008.00080.x © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Asia Pty Ltd 201