For the past 3 years, the nursing case management team at North Shore University Hospital in Manhasset, NY, has been involved in a project to implement more than 50 clinical pathways, which provide a written "time line" for clinical events that should occur during a patient's hospital stay. A major objective of this project was to evaluate the efficacy of these pathways with respect to a number of important outcomes, such as length of stay, hospital costs, quality of patient care, and nursing and patient satisfaction. This article discusses several statistics-related issues in the design and evaluation of such case management studies. In particular, the role of a research approach in implementing and evaluating hospital programs, the choice of a comparison (control) group, the exclusion of selected patients from analysis, and the problems of equating pathways with diagnosis-related groups are addressed.