The paper by Shimazu et al ^{Reference Shimazu, Shimodera, Mino, Nishida, Kamimura and Sawada1} adds robustness to already existing evidence for the role of family psychoeducation in psychiatric disorders. The study has sound methodology (i.e. randomised controlled trial) with adequate masking, in addition to being the first ever study to examine the effect of family psychoeducation for major depressive disorder. The authors describe the possible limitations of the study honestly. A source of funding (Grant-in-Aid for Scientific Research, Ministry of Health, Labour and Welfare, 2004) is also mentioned. However, there are some issues which should be further looked into. The aim was to examine family psychoeducation in the maintenance treatment of depression.

However, the patients included were either on continuation or maintenance treatment. Patients who are in partial remission cannot be considered as being in a continuation/maintenance phase. ² Also it was not mentioned how many patients had single or recurrent episodes (patients with single episodes need not receive maintenance phase treatment). Any other psychiatric comorbidity (substance misuse or personality disorder) in the participants was not mentioned, even though it has treatment implications. The health status and intellectual functioning of the primary family member included in the study was not mentioned, although these might compromise their active participation in psychoeducation sessions. The authors are silent on the ethical clearance of the study. For four caregivers psychoeducation sessions were done in the individual’s home and not in group sessions, which were not included in the final analysis and not part of the methodology mentioned – this could also have had an effect on the efficacy of the study. Remission was defined by the authors as a Hamilton Rating Scale for Depression (HRSD) score <6, but the normal score is mentioned as <7. ^{Reference Blacker, Sadock, Sadock and Ruiz3} The authors have not mentioned the reason for keeping a low score of HRSD in the study. Last, it could be a printing mistake, in the last line of Table 1 it is mentioned that high expressed emotion (as per FMSS) was seen in seven patients in the intervention group and none in the control group, but the results mention that it is seen in six patients in the intervention group and ten in the control group.