Comparison of hospital system characteristics at high-volume and low-volume hospitals for pancreatectomy

Abstract

Objective: While hospital volume has been associated with improved mortality rates after pancreatectomy, it is often non-modifiable. We sought to determine hospital system characteristics beyond surgical volume that can inform quality improvement initiatives at hospitals of any volume. Methods: Using the Nationwide Inpatient Sample database, we identified hospitals that performed elective pancreatectomies from 2002 to 2011. The American Hospital Association database was linked to obtain hospital system characteristics. High-volume hospitals (HVH) were defined as hospitals performing ≥20 cases/year. Results: We identified 1529 hospital-years performing pancreatectomy throughout the study period, of which 181 (11.8%) were HVHs. HVHs were associated with a lower adjusted mortality (−2.3%, p=0.05), morbidity rate (−4.4%, p=0.07) and shorter length of hospital stay (−1.6 days, p=0.01) when compared to low-volume hospitals (LVH). When assessing hospital system characteristics, HVHs had more hospital beds (322±166 vs 176±110 in LVHs, p<0.001), intensive care unit beds (44±33 vs 19±14, p<0.001) and lower nurses-to-bed ratios (4.8±2.4 vs 3.1±1.7 in LVHs, p<0.001). Additionally, more HVHs had advanced imaging (84.4% vs 50.8%, p<0.001), cardiology services with catheterization lab (94.8% vs 73.1%, p<0.001), comprehensive gastrointestinal endoscopy services (48.6% vs 19.4%, p<0.001) and health research programs (50.9% vs 25.6%, p<0.001). Conclusion: Compared to pancreatectomy-providing LVHs, HVHs were larger and had more intensive care and ancillary services support. We identified specific hospital system characteristics that could inform both LVHs' and HVHs' efforts to improve pancreatectomy outcomes.