Discriminant analysis: a technique for adding value to patient satisfaction surveys.

Abstract

Quality improvement in healthcare organizations requires effective measurement of patient satisfaction. This paper describes a methodology that identifies dimensions of care most closely associated with overall perceptions of quality. A patient satisfaction survey was mailed to 2,055 discharged patients of 13 home health agencies. Patients were asked to evaluate service dimensions of home health relating to scheduling, nursing care, home health aide services, and discharge arrangements. Overall satisfaction with quality of services was used as the dependent variable in two discriminant analysis equations. Eleven dimensions discriminated between "excellent" and "good" quality, and seven dimensions discriminated between "satisfactory" and "unsatisfactory" quality. Using discriminant analysis, items most closely associated with quality indeces can be identified and used in CQI initiatives.