To the Editor: The article by Ms. Phillips and her colleagues, "Moving Assertive Community Treatment Into Standard Practice" (1), is indeed timely. In the United Kingdom we face analogous issues, albeit within a public mental health service that is less fragmented and that has a basic level of coordination despite chronic underinvestment. Our government is strongly committed to the introduction of assertive community treatment, but government authorities are frustrated that European research rarely demonstrates the major reduction in hospitalization seen in the U.S. studies of assertive community treatment. Controversy remains about whether this phenomenon reflects the content of the programs or the context in which they operate.
Like Ms. Phillips and her coauthors, we are interested in identifying the features of assertive community treatment that are most strongly associated with successful outcomes. In our search for these factors, we conducted a systematic review of all studies of home-based care for people with mental health problems. We deliberately avoided too narrow a focus on assertive community treatment, which Ms. Phillips and colleagues suggested may have limited the usefulness of the Lewin Group's findings in this regard (2), and we included any service that aimed to treat patients outside of the hospital. This approach enabled us to look at a wide range of services studied and to examine how the service components provided to the intervention groups and the control groups were associated with reduction in hospitalization. Our results have been published in detail in a Health Technology Assessment Monograph (3) and are soon to appear in Psychological Medicine (4).
Our analysis identified a group of features that are common to intervention services: regular visits to the client's home, responsibility for both health and social care, lower caseloads (defined as fewer than 15 clients), multidisciplinary teams, and full integration of the psychiatrist into the work of the team. The first two components were found to be significantly associated with reduced hospitalization.
We found it interesting that services with these features overlap with but are not identical to the services identified in the article by Phillips and colleagues as constituting assertive community treatment. Despite admirable attempts to define assertive community treatment as a model, the danger remains of applying the label without first ensuring that practitioners are actually delivering "assertive community treatment." Phillips and colleagues clearly recognize the problems related to definition in their focus on how best to ensure fidelity to the assertive community treatment model. We would add that the interpretation of evidence from past studies is made difficult by the paucity of detail in many reports about the contents of the "black box."
A surprising finding from our study was that many experimental services had ceased to exist or had changed substantially when we followed them up; a few closed before our study was published (3). In many ways, our approach was the reverse of that used by Ms. Phillips and her colleagues. Although we commend them for seeking to determine the most effective means of implementing assertive community treatment, we would argue that consideration should also be given to whether such services can be sustained—an issue that may be overlooked once studies have become part of the literature and the services they tested are forgotten.