Author's reply

First, as presented in the Results of our paper, ¹ patients in standard care were mainly treated at the local community mental health centre (56.5%), at a private psychiatrist (24.7%) or at a local psychiatrist associated with the discharging ward (15.3%). Only three patients (3.5%) were treated at the general practitioner. Further, patients could at any time, if needed, be referred to a higher level of care (e.g. a community mental health centre).

Second, it is an advantage of our study compared with most other studies that we have register-based information available on the number of non-participants and are able to compare participants and non-participants in the trial. As presented in the Results, the 158 patients who participated in the trial did not differ from the 316 other potentially eligible patients regarding gender (54.4% female compared with 48.4%; P = 0.2) but were considerably younger (median 35.6 years (quartiles 28–47) compared with median 44.0 years (quartiles 33–57); P<0.001). We are rather convinced that the large proportion of non-participants in the trial is mainly due to the fact that the vast majority of these patients were simply not asked to participate in the trial (but we have register data on these as well). Clinicians may have been more observant to ask younger patients to participate (as the aim of the study was to investigate effects of interventions early in the course of bipolar disorder) and younger patients may have been more willing to participate in the trial. Nevertheless, as can be seen from the paper, the median age at inclusion in the trial was 35.6 years, very similar to findings in other studies recruiting patients following first admission to hospital ^2,3 (mean age 31.4 years (s.d. = 12.9) and 38.4 years (s.d. = 12.6) respectively). Thus, we believe that our findings can be generalised to patients with bipolar disorder discharged from their first psychiatric hospital admission.

Third, patients treated in the mood disorder clinic more often used a mood stabiliser (lithium or anticonvulsants) or an antipsychotic compared with patients allocated to standard care. This difference is an effect of the interventions and as medication (most likely) is an intermediary factor in relation to hospital admission as the primary outcome, adjustment for medication in the analyses would be incorrect. Finally, differences in costs were mainly due to the decreased in-patient costs for patients randomised to the mood disorder clinic, but costs to medication was included in the calculations.