Letter

In Reply: We thank Dr. Roskes for his thought-provoking letter. Coercion is a complex construct that has been poorly defined in the literature. Our pragmatic goal in this study was to define a heuristic set of simple, face-valid, objective indicators hypothesized to function as mechanisms for social control. These indicators do not measure coercion directly but, instead, assess opportunities for coercion. We intend them to complement, rather than replace, existing self-report measures of coercion.

By way of illustration, administrations of depot injections and daily monitoring of oral ingestion of medications provide opportunities for coercive practices, such as threatening to withhold money or using more subtle interpersonal pressures. Our statement that depot injections of antipsychotic medications "virtually eliminate choice" was imprecise; we were specifically referring to the daily choice many consumers face about whether to take their medications as prescribed. We, of course, also did not mean that all consumer choice was thereby eliminated.

In fact, as with all of the control practices we examined, opportunities for choice can be traced both upstream and downstream from the practice itself. Assuming control or exercising coercion in the present may indeed provide greater opportunities for more meaningful choice down the road. As Dr. Roskes suggests, outcome studies examining use of coercive practices in community mental health could shed light on this often contentious topic—clarifying the role of coercion in both short-term and long-term recovery.

As Dr. Roskes notes, our study did not address how the treatment of justice-involved consumers may influence the use of various control practices. Although it is clearly germane to the issue of coercion, we omitted any discussion of criminal justice involvement because of the low three-month prevalence rates of incarceration in our sample—on average, less than 5% of the caseloads. Also, when our study was conducted, there were no forensic assertive community treatment teams in Indiana.

We too were not surprised that a schizophrenia diagnosis and substance use were positively correlated with control, for the reasons Dr. Roskes describes. As explicitly stated in our article, use of agency control ideally should be determined on an individual basis. As a practical matter, the psychiatric field does not have adequate psychometric tools for identifying consumers who lack decisional capacity, so we wonder how realistic Dr. Roskes' observation is that "many coercive interventions require assessment of decision-making capacity." Moreover, many clinicians probably overestimate the prevalence of decisional incapacity in determining the need for intensive supervision, as suggested by a recent major medication study in which very few patients were excluded on this basis ( 1 ).

Lack of evidence-based guidelines prescribing when, how, and for how long restrictive practices should be used creates the potential for misuse. The assignment (and removal) of representative payeeship is an excellent example of a control that requires no formal assessment of financial decision-making capacity—instead it relies on a clinical opinion ( 2 ). Its use may be determined primarily by provider attitudes and agency policy, as our study sought to understand.