M. Smith1,2, U. Nuliyalu2, S. P. Shubeck1,2,3, J. B. Dimick1,2, H. Nathan1,2 1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA 2University Of Michigan,Center For Healthcare Outcomes & Policy,Ann Arbor, MI, USA 3University Of Michigan,National Clinician Scholars Program,Ann Arbor, MI, USA
Introduction: Improving the value of healthcare delivery is a major focus of healthcare reform. Previous studies have documented substantial cost savings for surgical care delivered in high quality hospitals, with particularly large cost differences for high-risk patients. Practically, shifting high-risk patients to high-quality hospitals must be done within small geographic areas. We sought to determine the availability of high-quality hospitals, the distribution of high-risk patients, and the potential benefit of referral of high-risk patients for surgery within small geographic areas.
Methods: Using 100% Medicare claims data for 2012-2013, we identified elderly patients undergoing elective colectomy (Col), lung resection (Lung), total hip arthroplasty (THA), and total knee arthroplasty (TKA). Risk- and reliability-adjusted hospital rates of serious complications were assessed using a hierarchical logistic regression model, and hospitals were grouped into quintiles; lowest complication rate = high quality. A similar model was used to stratify patients into quintiles of high and low risk for complications. Price-standardized, risk-adjusted Medicare payments were calculated for the entire “surgical episode” from index admission through 30 days after discharge. The geographic units of analysis were Metropolitan Statistical Areas (MSAs), which consist of a relatively high population density (≥50,000) and include surrounding areas that roughly mirror typical commuting distances.
Results: The proportion of MSAs containing a high quality hospital ranged from 47% (Lung) to 58% (THA). A minority of MSAs contained both a high quality and low quality hospital (n=79, 22% Lung; 118, 30% Col; 120, 31% TKA; 122, 32% THA). In these MSAs, 25% of high-risk patients received care at the lowest quality hospitals (TKA 23%, THA 24%, Lung 26%, Col 27%), and 38% of high-risk patients were treated at high quality hospitals (34% Col, 38% TKA, 39% Lung, 39% THA). There was wide variation in costs between high and low quality hospitals within MSAs, and this difference was particularly large for high-risk patients (Figure). Referral of a high-risk patient from a low to high quality hospital within a MSA would generate an average savings of $13,840 for Lung ($31,659 vs $45,499), $8,981 for Col ($29,230 vs $38,211), $2,583 for THA ($20,954 vs $23,537), and $1,936 for TKA ($19,992 vs $21,928, all P <0.001).
Conclusion: In small geographic areas containing high and low quality hospitals, 25% of high-risk patients received care at the lowest quality hospitals. Triaging of high-risk patients to high quality hospitals within small geographic areas may serve as a template for strategic local referral as a means of reducing costs in Medicare.
