All things not being equal: readmission associated with procedure type

BackgroundThere is an accelerated effort to reduce hospital readmissions despite minimal data detailing risk factors associated with this outcome.Materials and methodsWe analyzed National Surgical Quality Improvement Project data from January 1, 2011–December 31, 2011, evaluating all patients undergoing one of 34 targeted operative procedures across all surgical specialties. Multivariate regression models of risk for readmission were developed including targeted procedure codes, demographic variables, preoperative variables, intraoperative variables, and postoperative adverse events. Our main outcome measure was hospital readmission.ResultsA total of 217, 389 patients met study inclusion criteria. Minimal associations existed between patient factors and risk of readmission. Adverse events including unplanned operating room return (odds ratio [OR] 8.5; confidence interval [CI] 8.0–9.0), pulmonary embolism (OR 8.2; CI 7.1–9.6), deep incisional infection (OR 7.5; CI 6.7–8.5), and organ space infection (OR 5.8; CI 5.3–6.3) were associated with increased risk of readmission. Our data suggest the type of procedure performed is significantly associated with risk of readmission. Furthermore, multivariate analysis revealed procedures, involving the pancreas, rectum, bladder, and lower extremity vascular bypass, were associated with the highest risk of readmission.ConclusionsPostoperative complications demonstrated stronger association with readmission than patient factors. Focused analysis of higher risk procedures may provide insight into strategies for risk reduction.