We read the article by Cantwell et al (Reference Cantwell, Brewin and Glazebrook1999) with interest. The study relates to an important aspect on the changing pattern of substance misuse in patients with first-episode psychosis. The authors have found that substance users were more likely to be males and to have a younger age at onset of psychosis. However, it would be more informative if the authors gave the prevalence figures in either gender and in different age groups to enable identification of high-risk groups for health promotion measures.

We would also like to highlight some discrepancies in the paper. The magnitudes of comorbid substance misuse among affective and delusional disorders have been miscalculated in Table 2. Considering the fact that affective disorders include manic psychosis and depressive psychosis in the study, the calculated prevalence of substance misuse among affective disorders is found to be 18.9% instead of 11.9%, and for delusional disorder it is 15.4% instead of 7.7% as reported by the authors. Similarly, the total number of stimulant misusers is four, instead of three given by the authors in Table 2. Based on this newly calculated substance misuse rate, there is no significant difference in the substance misuse between people with schizophrenia (23.5%) and those with affective disorders (18.9%) (χ²=0.27, P=0.603). Therefore, the authors' observation that subjects with affective disorder were less likely to be substance misusers needs to be modified.