This study will apply recent advances in data envelopment analysis (DEA) to the empirical study of best practices of surgeons performing coronary artery bypass graft procedures in Pennsylvania in 1992 and 1993. Data for this study will be drawn from a proprietary database maintained by MediQual Systems, Inc. and the American Hospital Association's Annual Survey of Hospitals. Patient care in the surgeon's practice will be modeled as a multi- output service production process. Taking into account differences in the quality of care and patient case mix associated with surgeons in the sample, DEA will be used to conduct an empirical study of surgeon efficiency and identify best practices (surgeons who provide equality care to patients using the fewest clinical and technical resources). Tobit regression will be used in the second phase of the analysis to determine whether relatively efficient surgeons are more likely to be associated with hospitals with certain characteristics (e.g., number of beds). Developed in the fields of operations research and management science, data envelopment analysis is an analytic tool of relatively recent origin. The first paper using DEA was published in 1978. Data envelopment analysis has the potential to be a valuable new tool in cost-effectiveness research in health services, and adapting this methodology for health services research is the focus of the proposed study.
