The Veterans Health Administration National Surgical Quality Improvement Program (NSQIP) has developed reliable comparative measures of 30-day mortality and morbidity for surgical procedures that have led to significant transformations in quality. However, 30-day outcomes provide only a limited picture of true surgical efficiency. True surgical efficiency should be a reflection of both quality and cost. The objective of this proposal is to develop models for predicting post-surgery costs, adjusting for patient's case-mix, and to develop measures of surgical quality that combine both outcomes (e.g. mortality and morbidity), costs, and methods for analyzing those measures. Our approach is based on the hypothesis that costs represent an important independent assessment of efficiency of care, and should combined with clinical quality measures to gain a complete picture of true efficiency of care. Our long-term goal is to identify the most efficient facilities with optimal measures of both surgical outcomes and cost, so that these processes at these facilities can be emulated throughout the VA. We propose a retrospective cohort study of all patients included in the NSQIP, DSS and relevant VA databases with seven common surgical procedures between 2007 and 2009. A comprehensive database that includes all records will be developed, which will include two years of records prior to date of the surgical procedure and at least one year of follow-up. The seven surgical procedures were selected to represent a mix of complexities ranging from hernia repair to CABG to demonstrate the feasibility and reliability of the methods across a range of facilities.
The project will provide more comprehensive comparisons of VA surgical care and add statistical methods that jointly assess surgical quality and cost outcomes across facilities. Our long-term goal is to develop methods that will identify the most efficient facilities with optimal measures of both outcomes and cost, so processes at these facilities can be emulated throughout the VA.
