Paraphrase procedures for assessing comprehension of health outcome measures: an illustration from schizophrenia research.

Valid measurement of health outcomes require instruments that are meaningful to those who use them. Clinicians, patients, and others involved in health care have different perspectives on health conditions that are likely to lead to different interpretations of outcome measures and reductions in the accuracy and validity of outcome measurements. This study used paraphrase procedures, a systematic approach for evaluating respondent comprehension, to evaluate descriptions designed for use in studies of schizophrenia treatment outcomes. In four stakeholder groups directly and indirectly involved in schizophrenia treatment (clinicians, patients, patients' families, and the general public), the paraphrase procedures identified comprehension problems that could compromise measurement quality and yielded specific information to guide improvement of the outcome descriptions. The paraphrase procedures detected comprehension problems that were not identified by respondents. This study suggests that systematic assessment of comprehension can improve the quality of health outcome measurement.