Value of real world economic evaluations of crisis accommodation programmes for patients with severe mental illness

To the Editor

We appreciate the comments provided by Dr Amos (Amos, 2014) regarding our evaluation of a crisis house for people with severe and persistent mental illness (Siskind et al., 2013). The core of Dr Amos’s comments relate to the difficulties in selecting appropriate controls from administrative data extracts for use in retrospective quasi-experimental evaluations of existing mental health services.

While we agree that a limitation of this study was that there were differences between the intervention and control groups, we disagree with the view that no valid conclusions can be drawn from the study. In consultation with our statistical advisor, we believe that the statistical techniques used to adjust for differences between the intervention and control groups permit us to make meaningful economic interpretations of our results.

There are inherent limitations when using retrospective administrative data. As variables cannot be prospectively selected, researchers must make interpretations form available data. Missing data can make certain variables unusable, as was the case for our data related to community clinical contacts. However retrospective administrative datasets allow for the evaluation of real world interventions. Insights into the efficacy of programmes targeted to actual public mental health service patients are inherently valuable as the results of randomised controlled trials with strict inclusion criteria can be difficult to translate to the real world.

We thank Dr Amos for his comments, but believe that the economic evaluation of real world programmes such as ours provides useful information for service planners regarding programmes that can have an immediate benefit for mental health patients in need.