Identifying Factors Associated With Postdischarge Mortality After Readmission Following Pancreaticoduodenectomy.

INTRODUCTION

While in-hospital mortality following pancreaticoduodenectomy (PD) is low at high-volume centers, early deaths after discharge remain a concern. This study aimed to identify risk factors for mortality among patients readmitted within 30 d of PD.

METHODS

Data from the American College of Surgeons National Surgical Quality Improvement Program-targeted pancreatectomy files (2014-2021) were analyzed. Adult patients who underwent PD and were readmitted within 30 d were included. Patients were stratified by survival status following readmission, readmitted survivors versus readmitted nonsurvivors. We used descriptive statistics to compare demographics, perioperative factors, causes of readmission, and postoperative complications (categorized by timing [predischarge or postdischarge]). Multivariable logistic regression identified factors independently associated with 30-d mortality among readmitted patients.

RESULTS

Of 22,203 patients who underwent pancreaticoduodenectomy, 96.6% were discharged alive. Among these, 17.9% were readmitted within 30 d (n = 3844). Among the readmitted patient, 1.35% died within 30 d of surgery, representing 50.5% of all postdischargedeaths. Readmitted nonsurvivors were older and more likely to have chronic obstructive pulmonary disease and diabetes mellitus (P < 0.001). Cardiac arrest and septic shock were the most common causes of readmission and were significantly associated with mortality (P < 0.001). Variables independently associated with mortality among readmitted patients included American Society of Anesthesiologists class 4 (adjusted odds ratio: 6.61; P = 0.025), unplanned intubation (29.40; P < 0.001), cardiac arrest (88.58; P < 0.001), myocardial infarction (5.88; P = 0.017), and septic shock (2.97; P = 0.026).

CONCLUSIONS

Early postoperative deaths often occur after discharge, frequently following sepsis or cardiac events; enhanced risk stratification and surveillance could reduce preventable mortality.