30-Day Readmission Following Inpatient Total Shoulder Arthroplasty in Pennsylvania

Background: The growth of value-based purchasing programs has increased interest in understanding risk factors for readmission and preventing unplanned readmissions. The aim of this study was to estimate rates and risk factors for 30-day readmission following inpatient total shoulder arthroplasty (TSA) in Pennsylvania between 2010 and 2018. Methods: We examined inpatient TSA between 2010 and 2018 in Pennsylvania using a statewide administrative discharge data set. Readmissions were for any cause and to any hospital in the state, and not only to the operating hospital. Potential risk factors included patient demographics, comorbidities, and discharge destination. Logistic regression was used to identify significant risk factors. Results: Among 14,333 patients receiving TSA, 469 (3.27%) were readmitted within 30 days. Patients admitted on an urgent or emergent basis had 65% greater odds (odds ratio [OR] = 1.65, p = 0.009) than patients treated on an elective basis. While patients covered by commercial insurance had 24% lower odds of readmission (OR = 0.76, p = 0.04), patients covered by Medicaid had twice (OR = 1.95, p = 0.002) the odds of readmission. Relative to patients with no comorbidities, patients with 1 to 2 comorbidities had 54% greater odds (OR = 1.54, p < 0.0001), and patients with 3+ comorbidities had triple the odds (OR = 3.14, p < 0.0001) of readmission within 30 days. Discharge destination was a significant predictor of readmission, with patients discharged with home health having 50% greater odds (OR = 1.49, p < 0.0001) and patients discharged to a skilled nursing facility having more than twice the odds (OR = 2.19, p < 0.0001) of readmission within 30 days. Conclusions: In this statewide analysis, there were several significant risk factors for 30-day readmission following inpatient admission for TSA, many of which may be useful targets for hospitals to prevent costly orthopaedic surgery readmissions. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.