Study of Risk Factors for Readmission After Pancreatectomy for Cancer

Abstract

Objective: To identify the factors associated with readmission after pancreatectomy for cancer and to assess their impact on the 1-year mortality in a French multicentric population. Background: Pancreatectomy is a complex procedure with high morbidity that increases the length of hospital stay and jeopardizes survival. Hospital readmissions lead to increased health system costs, making this a topic of great interest. Methods: Data collected from patients who underwent pancreatectomy for cancer between 2011 and 2019 were extracted from a French national medico-administrative database. A descriptive analysis was conducted to evaluate the association of baseline variables, including age, sex, liver-related comorbidities, Charlson Comorbidity Index, tumor localization, and use of neoadjuvant therapy, along with hospital type and volume, with readmission status. Centers were divided into low and high volumes according to the cutoff of 26 cases/year. Logistic regression models were developed to determine whether the identified bivariate associations persisted after adjusting for the patient characteristics. The mortality rates during readmission and at 1 year postoperatively were also determined. Results: Of 22,935 patients who underwent pancreatectomy, 9129 (39.3%) were readmitted within 6 months. Readmission rates by year did not vary over the study period, and mean readmissions occurred within 20 days after discharge. Multivariate analysis showed that male sex [odds ratio (OR) = 1.12], age >70 years (OR = 1.16), comorbidities (OR = 1.21), distal pancreatectomy (OR = 1.11), and major postoperative complications (OR = 1.37) were predictors of readmission. Interestingly, readmission and surgery in low-volume centers increased the risk of death at 1 year by a factor of 2.15 [(2.01–2.31), P < 0.001] and 1.31 [(1.17–1.47), P < 0.001], respectively. Conclusions: Readmission after pancreatectomy for cancer is high with an increased rate of 1-year mortality.