This year Psychiatric Services is focusing on evidence-based practices, encouraging both inspection of the scientific evidence for treatments and introspection about our own practices. In this issue Valenstein and her colleagues report that veterans undergoing treatment for schizophrenia between 1991 and 1995 often received dosages of conventional antipsychotic medications that did not meet recommended guidelines, that African Americans were more likely to have received treatment that did not meet guidelines, and that treatment varied substantially across sites. Unfortunately, these results can no longer be considered as breaking news, not just because the conventional antipsychotic medications are "old," but also because these findings replicate those of several recent studies.
It might be tempting to dismiss this latest study as dated, reporting on treatments that appear to be on their way out given the marketing data on the new antipsychotics. However, the main concern here is not what happened with conventional antipsychotics five to ten years ago but whether scientifically based treatment is incorporated into practice any faster now than was the case in the past with the old agents. It seems highly likely that as new practice guidelines covering the new antipsychotic agents become available, we will learn that dosing remains a problem, that patients from ethnic minorities are treated differently than Caucasians, and that practices vary considerably from provider to provider and from setting to setting. We will also likely learn that patients are kept on medication regimens indefinitely despite inadequate clinical response and significant side effects.
The article by Valenstein and colleagues provides critics of evidence-based practice with convenient diversions from this main plot. We could focus on how difficult it is to keep practice guidelines up to date in the face of rapidly emerging technologies. We could debate the wisdom of recommended dosage ranges—for example, recent analyses of the Schizophrenia Patient Outcomes Research Team suggest that dosages below this range may be less problematic than those above the range. We could argue that clinical wisdom must prevail in individual cases, outweighing group tendencies established through randomized clinical trials.
However, how do we explain continued disparities in care due to race or ethnicity? How do we explain widely divergent practices among practitioners and practice sites? How do we explain the fact that patients are maintained on treatments for which outcomes are poor despite evidence that other treatments might work better? The promotion of evidence-based practices is attended by serious debate, but we cannot lose sight of the fact that this debate exists because significant numbers of patients receive treatment that appears to be substantially underinformed by recent science.
