Letter

Satisfaction as an Outcome Measure

To the Editor: We would like to take issue with Dr. Norquist's ( 1 ) July Taking Issue commentary, "Contented but Not Better: Problems With Satisfaction." His comments about the study by McCrone and colleagues ( 2 ), which used satisfaction as a primary outcome measure in a cost analysis of assertive community treatment teams in London, raised two important issues: the precision of satisfaction as a concept and its utility for policy decisions. His comments ignore an extensive literature that has refined the concept of satisfaction and has demonstrated the role of satisfaction as an indicator of health care quality ( 3 ). Recently, satisfaction has been successfully used as a primary outcome in clinical antipsychotic trials (Canuso C, Grinspan A, Merriman C, et al., unpublished manuscript, 2009). Satisfaction as a multidimensional construct captures the essence of what patients, clinicians, policy makers, and taxpayers hope to achieve ( 4 ).

Thus we disagree with the thrust of Dr. Norquist's comments, which for some readers will raise doubts about such subjective constructs. We do agree that it would have been more informative if McCrone and colleagues had taken a multidimensional approach in addition to using a global measure of satisfaction. Several studies in the oncology field, as well as our studies of quality of life, have demonstrated the reliability of global measures. Although global measures capture patients' preferences, we agree that they do not provide information about factors that contribute to the genesis of the construct of satisfaction. Measures assessing satisfaction must be specifically focused.

We have demonstrated that most psychiatric patients are able to focus on questions about satisfaction with the care they receive and are reliably and consistently able to express their inner feelings and their level of satisfaction ( 4 ). Uncritical acceptance of the notion that psychiatric patients are unreliable in expressing their inner feelings can set the field back 30 years. The paradox is that when psychiatric patients report their hallucinations and delusional experiences, which are subjective in nature, they are believed and their reports are taken to be valid in making a diagnosis. When we submitted our first paper in the mid-1970s, which presented data about negative subjective responses to antipsychotics, the editor returned the manuscript with a polite comment that the reviewers considered the subject as "soft science." It took several years for the concept of measuring subjective responses to antipsychotics to become well-established, mainstream clinical practice, and the practice has been further supported by dopamine neuroimaging studies.

We agree that satisfaction as a construct requires continued refinement and more understanding of its components. However, administrators and policy makers who pay no attention to patients' dissatisfaction with therapeutic interventions should not be surprised when these interventions fail. We are afraid that Dr. Norquist's comments could discourage research and once more relegate subjective experiences such as satisfaction to the realm of "soft science." We do not doubt that many of our patients will never be contented unless they feel better.