There is a lack of scientific evidence evaluating the efficacy of OPC ^{Reference Kisely and Campbell1} and doubt around whether randomised controlled trials are realistically the best method of evaluating the impact of OPC. ^{Reference Swanson and Swartz2} Our work is a retrospective case–control study, with all its limitations, and we welcome any contribution that could offer us some improvement for further studies.

Responding to the issues raised by Dr Mustafa, I would like to comment as follows.

First, we had considered in our study that there were differences in the motives for index admissions between the groups. In the OPC group the main reasons were clinical decompensation because of non-adherence to treatment (78%) and aggressive behaviour (22%). In the control group, admission occurred mainly because of clinical decompensation without a clear non-adherence to treatment (47%) and the reasons were inconsistent use of medication, changes in the pharmacological pattern or substance misuse. This could undermine the similarity of the two study groups and, therefore, the suitability for comparison.

Second, owing to the nature of the study, patients eligible for the control group were automatically excluded if within the subsequent 2 years of the study they were placed on OPC. Third, we agree there is a trend of reduced hospitalisation over the 4-year study period in both groups. This may have been driven by factors such as the improvement of community services or home services that could potentially confound the results of this study.