We agree with Marneros (Reference Marneros2006) that the concept of acute psychosis challenges the traditional Kraepelinian dichotomy. Clinical experience suggests that this condition is heterogeneous with the only similarity being the acuteness of onset. On longitudinal follow-up acute psychoses are separated into different categories; many resolve rapidly; some resolve only to recur again in a similar fashion; and others present differently over time to be reclassified as schizophrenia or mood disorders. It is therefore difficult to identify demographic or other characteristics that define this group.

The majority of patients who received this diagnosis in our study were young adults (mean age 29.75 years, s.d.=10.95) and male (52%) (Reference Thangadurai, Gopalakrishnan and KurianThangadurai et al, 2006). This is in contrast to Marneros (Reference Marneros2006) who reported a later age at onset and a predominance of women. A large group of our patients were later given a diagnosis of affective disorder (9.2%), schizophrenia (26.4%) or recurrent episodes of acute psychosis (11.5%); others did not present with psychotic symptoms over the follow-up period. Although these results suggest that it is difficult to predict response to medication, course and outcome, it is well known that acuteness of onset is a good prognostic factor in both schizophrenia and mood disorders.

We argue that the concept of acute psychosis is necessary since many patients may present soon after the onset of illness, when the clinical features may not allow them to be categorised into any of the more classic disorders. Although many patients recover, some relapse with similar acute psychotic presentations; a significant proportion also develop schizophrenia and mood disorders. The difficulty in reaching a diagnosis at the initial presentation arises because it is often difficult to recognise the classic syndromes at this time. However, these can be identified over time. Thus, acute psychoses can be a presentation of the more traditional syndromes but can also be separate clinical entities which may or may not recur. Assuming that those who present with acute psychosis confirm to a homogeneous group does not fit with the clinical reality.