BACKGROUND: Most studies focus on readmission within 30 d of surgery and may therefore underestimate the true burden of readmission after complex procedures. We therefore sought to explore factors associated with readmission within 90 d of discharge after pancreaticoduodenectomy (PD). METHODS: Patients discharged after PD between 2010 and 2012 were identified from the Truven Health MarketScan database. Determinants of early (≤30 d) and late (31-90 d) readmission were identified and analyzed. RESULTS: A total of 2209 patients met inclusion criteria with 615 patients being readmitted within 90 d; 20.37% (n = 450) had an early readmission, whereas 7.47% (n = 165) had a late readmission. Patients readmitted early had a longer length-of-stay (LOS) for the readmission compared with patients readmitted late (median LOS = 5 d, interquartile range, 3-8 versus median LOS = 3 d, interquartile range, 2-5, P < 0.001). Common causes for readmission differed between these two patient groups. On multivariable analysis, postoperative complications (odds ratio [OR] 1.46, 95% confidence interval [CI] 1.16-1.84, P < 0.001), an extended LOS (OR 1.34, CI 1.05-1.71, P = 0.019), and the presence of preexisting coronary heart disease (OR 1.82, CI 1.09-3.01, P = 0.020) or renal disease (OR 2.15, CI 1.01-4.61, P = 0.048) were associated with an early readmission. No patient- or procedure-related factors were associated with late readmission. CONCLUSIONS: Readmission within 30 d after PD was associated with patient- and procedure-related factors. Readmission in the 31-90-d window was not associated with patient- or procedure-related factors and may be influenced by the underlying pathology or subsequent medical management for that disease.
BACKGROUND: Most studies focus on readmission within 30 d of surgery and may therefore underestimate the true burden of readmission after complex procedures. We therefore sought to explore factors associated with readmission within 90 d of discharge after pancreaticoduodenectomy (PD). METHODS:Patients discharged after PD between 2010 and 2012 were identified from the Truven Health MarketScan database. Determinants of early (≤30 d) and late (31-90 d) readmission were identified and analyzed. RESULTS: A total of 2209 patients met inclusion criteria with 615 patients being readmitted within 90 d; 20.37% (n = 450) had an early readmission, whereas 7.47% (n = 165) had a late readmission. Patients readmitted early had a longer length-of-stay (LOS) for the readmission compared with patients readmitted late (median LOS = 5 d, interquartile range, 3-8 versus median LOS = 3 d, interquartile range, 2-5, P < 0.001). Common causes for readmission differed between these two patient groups. On multivariable analysis, postoperative complications (odds ratio [OR] 1.46, 95% confidence interval [CI] 1.16-1.84, P < 0.001), an extended LOS (OR 1.34, CI 1.05-1.71, P = 0.019), and the presence of preexisting coronary heart disease (OR 1.82, CI 1.09-3.01, P = 0.020) or renal disease (OR 2.15, CI 1.01-4.61, P = 0.048) were associated with an early readmission. No patient- or procedure-related factors were associated with late readmission. CONCLUSIONS: Readmission within 30 d after PD was associated with patient- and procedure-related factors. Readmission in the 31-90-d window was not associated with patient- or procedure-related factors and may be influenced by the underlying pathology or subsequent medical management for that disease.
Authors: Gregory A Williams; Jingxia Liu; William C Chapman; William G Hawkins; Ryan C Fields; Dominic E Sanford; Majella B Doyle; Chet W Hammill; Adeel S Khan; Steven M Strasberg Journal: J Gastrointest Surg Date: 2019-12-16 Impact factor: 3.452