The overall purpose of this study is to evaluate the utility of Medicare and Manitoba Health Commission claims data for the non-experimental evaluation of outcomes associated with selected surgical procedures and medical admissions. Previous work has demonstrated the feasibility of using claims data for describing outcomes and evaluating quality of care. The current study proposes to extend this earlier research to a broader range of hospitalizations, a more diverse set of outcomes, and a large population base. The project has three principal aims: (1) to generate descriptive statistics of mortality and morbidity for the most frequent medical admissions and surgical procedures, and to examine the relationship between outcomes and hospital characteristics; (2) to test specific hypotheses concerning the associations between alternate therapeutic approaches and patient outcomes for a selected subset of the frequent medical and surgical hospitalizations; and (3) to validate outcomes for one specific procedure -- prostatectomy -- through a review of the patient records. Data for the study are the enrollment, provider, and claims (for hospital and physician services) files which have been collected by the Medicare program and the Manitoba Health Services Commission. Patient covariables (age, sex, place of residence, prior medical history, comorbid conditions, severity of illness), hospital characteristics (number of beds, volume of selected types of admissions, teaching status, ownership), and outcomes (hospital readmissions, complications of treatment, death) will be defined for individual persons at least 65 years of age by linking information across these files. Specific hypotheses regarding the relationship between outcomes and patient/institutional risk factors will be initially tested on a subset of the data utilizing a proportional hazards model, then evaluated with the remainder of the data. The final report will include (1) a detailed description of the methods and software for defining, evaluating, and validating outcomes of hospital care using claims data; (2) a description of the mortality and morbidity following specific medical and surgical admissions and their relationship with hospital characteristics; and (3) an evaluation of alternative operative techniques. The implications of the study findings for clinical practice, hospital management, and public policy will also be discussed.
