Procedure-Specific Trends in Surgical Outcomes

Abstract

Quality improvement efforts have generally focused on hospital benchmarking, and processes and outcomes shared among all operations. However, quality improvement could be inconsistent across different types of operations. The objective of this study was to identify operations needing additional concerted quality improvement efforts by examining their outcomes trends. Ten procedures (colectomy, esophagectomy, hepatectomy, hysterectomy, pancreatectomy, proctectomy, total hip arthroplasty, total knee arthroplasty, thyroidectomy, and ventral hernia repair) commonly accrued into the American College of Surgeons NSQIP between 2008 and 2015 were included. Trends in risk-adjusted, standardized, smoothed rates were constructed for each procedure across 6 outcomes (mortality, pneumonia, renal failure, surgical site infection, unplanned intubation, and urinary tract infection [UTI]). Of 1,255,575 operations analyzed, the overall unadjusted rate for mortality across all 10 procedures was 1.08%, for pneumonia 1.44%, for renal failure 0.67%, for surgical site infection 5.28%, for unplanned intubation 1.11%, and for UTI 1.86%. Hepatectomy demonstrated the greatest improvement across outcomes (4 of 6 outcomes; 362 adverse events avoided out of 10,000 procedures), and UTI demonstrated the greatest improvement across procedures (8 of 10 procedures; 989 adverse events avoided out of 10,000). For pancreatectomy, rates of mortality, unplanned intubation, and UTI improved, but surgical site infection rates were detected to have significantly increased (p < 0.05). Hepatectomy was detected to have improved across the greatest number of outcomes, and UTI rates improved significantly across the greatest number of procedures. Surgical site infection rates after pancreatectomy, however, were detected to have increased, identifying an urgent need for additional concerted quality improvement efforts.