PERFORMANCE-MEASURES AND MEASUREMENT - SHORT-FORM MEASURES OF PHYSICIAN AND EMPLOYEE JUDGMENTS ABOUT HOSPITAL QUALITY

Objective: This article describes the development of short-form measures of physician and employee judgments about hospital quality from reliable and valid long-form measures previously published. Design: The study used a cross-sectional sample of survey results from the fourth quarter of 1989 through the second quarter of 1991. Setting: The measures were tested in a mix of urban and rural hospitals and of voluntary and investor-owned organizations with bed sizes ranging from 75 to 470. Respondents: A sample of 3,435 physicians (from 44 hospitals) and 17,315 employees (from 36 hospitals) was used. Methods: Stepwise regression analytic techniques were used to guide selection of the short-form items (31 physician items, 24 employee items). Hospital-level reliabilities and correlations between the short-form and long-form total process measures were estimated. Results: Despite an estimated reduction in response time of two minutes and five minutes, respectively, for the physician and employee measures, hospital-level reliability estimates were virtually identical for corresponding short-form and long-form measures. In addition, the short-form total process measures correlated 0.99 (physician) and 0.98 (employee) with the long-form measures. Conclusions: The short-form measures are an improvement over the original full-length instruments in terms of operational efficiency and, therefore, are preferable for many applications.