I read with interest the paper by Forbes et al, ^{Reference Forbes, Cash and Lawrie1} which investigated the impact of a crisis resolution service. I am intrigued by their finding that the introduction of the crisis service was followed by an unexpected increase in the absolute numbers of patients detained under the Mental Health Act. In their discussion a number of possible explanations are explored. However, I believe there is one possible explanation, which is not fully discussed, although it is perhaps hinted at in the clinical implications section of their abstract. This is that the intervention might have a negative impact on some patients.

This is now the third study to find this association, ^{Reference Keown, Tacchi, Niemiec and Hughes2,Reference Tyrer, Gordon, Nourmand, Lawrence, Curran and Southgate3} with only one group failing to replicate it. ^{Reference Johnson, Nolan, Pilling, Sandor, Hoult and McKenzie4} Tyrer et al ^{Reference Tyrer, Gordon, Nourmand, Lawrence, Curran and Southgate3} explicitly and at some length discuss the notion that negative effects on some patients of this type of service are one of the most plausible explanations for the increase in compulsory admissions. Furthermore, they suggest that any benefit from crisis services through reducing informal admissions may be cancelled out by the increase in compulsory admissions.

There does not yet seem to be a consensus around this important issue. Further research to explore the association is therefore warranted. Also of importance is research to clarify any risk factors that predict compulsory admission to hospital following a period of treatment by crisis services. Identification of such factors could potentially be used to improve services to patients.