A17 17..17

Early intervention by teams specialising in the treatment of early psychosis has been suggested to enhance outcome in the early phase of psychotic illness. Two papers in the Journal address this topic. The first examines the cost-effectiveness of these services, while the second examines the 5-year outcome of one such service. McCrone and colleagues (pp. 377–382) report that although their specialised early-onset service did not increase costs relative to treatment as usual over an 18-month period, it failed to show any significant benefit in cost-effectiveness. A 5-year follow-up study in the same service found that there were no differences in outcome, as assessed by number of admissions or bed-days, compared with those patients receiving the routine service. Gafoor et al (pp. 372–376) suggest that this result is surprising given the clear benefit observed at the earlier 18-month follow-up, but highlight that the specialised interventions were stopped at 18 months, and that a longer duration of treatment was required for it to maintain its advantage. An accompanying editorial by Friis (pp. 339–340) places both these studies in a wider context and considers the reasons for these patients not retaining the advantageous outcomes observed at the earlier follow-up. He suggests that the critical factor is reduced engagement with key personnel, as the early-intervention services invest a lot of energy into engagement with patients, and this may not be possible in the standard community services. A related reappraisal by Singh (pp. 343–345) provides an erudite background to the development of early-onset services. Commenting on both the articles, Singh concludes that these services make a difference to outcomes in the early phase of the illness, but that the transition to generic teams may undo this advantage. He suggests that the answer may lie in restructuring generic teams to have a more specialised, possibly disorder-specific, role; interestingly, this is the path that some services in the UK are now following.