M. E. Smith1,2, U. Nuliyalu2, J. B. Dimick1,2, H. Nathan1,2 1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA 2University Of Michigan,Center For Healthcare Outcomes And Policy,Ann Arbor, MI, USA
Introduction: Surgery at low-quality hospitals is associated with increased morbidity and mortality. High-risk patients have disproportionately worse outcomes and achieve the greatest benefit when referred to high-quality hospitals. However, referral to high-quality hospitals may impose unreasonable travel burdens and lead to discontinuity of care. A more selective referral strategy focused on diverting high-risk patients from the lowest-quality hospitals may be practical and beneficial. We sought to quantify the impact and feasibility of a local referral strategy focused on high-risk patients.
Methods: We identified patients age >65 years undergoing elective lung resection (Lung), proctectomy (Rectal), coronary artery bypass graft (CABG), esophagectomy (Esoph), and pancreatectomy (Panc) in 2012-2014 Medicare claims data. Hospitals were stratified into 5 grades of quality, A-F, by risk- and reliability- adjusted serious complication rates. Patients were risk-stratified by clinical factors. Travel burden was calculated by comparing the distance between a patient’s home zip code and Grade F hospital’s zip code versus the distance between a patient’s zip code and that of the nearest higher-quality hospital (Grade A or Grade A-C).
Results: One quarter of high-risk patients were treated at Grade F hospitals (24% Lung, 28% Rectal, 18% CABG, 23% Esoph, 19% Panc). Shifting these patients to Grade A hospitals would decrease serious complications from 54% to 13% (Absolute Difference: Lung 31%, Rectal 47%, CABG 27%, Esoph 58%, Panc 43%) and mortality from 13% to 7% (Lung 5%, Rectal 3%, CABG 4%, Esoph 11%, Panc 7%) (Table). High-risk patients at Grade F hospitals could travel a median of 4 additional miles to reach a Grade A hospital. Expanding the definition of high-quality to include Grade A, B, and C hospitals results in comparable benefits. Diverting high-risk patients from Grade F to Grade A-C hospitals would decrease serious complications from 54% to 16% (Lung 28%, Rectal 43%, CABG 23%, Esoph 55%, Panc 41%) and mortality from 13% to 8% (Lung 5%, Rectal 2%, CABG 3%, Esoph 9%, Panc 7%) (Table). Notably, patients could travel shorter distances to reach the nearest A, B, or C hospital than the F hospital they were treated at.
Conclusion: Shifting the highest-risk patients out of the worst hospitals for complex surgical procedures would require selectively referring only 5% of patients. Triaging patients from Grade F to the nearest A-C hospital would optimize surgical value while requiring the majority of patients to travel acceptable distances. This suggests that local referral of high-risk patients out of the lowest quality hospitals is a necessary and practical strategy for improving the value of surgical care.
