Characterization of Readmission by Day of Rehospitalization After Colorectal Surgery.

Abstract

Factors associated with readmission stratified by the day of postdischarge rehospitalization after colorectal surgery have not been characterized previously. The purpose of this study was to identify factors leading to readmission on a day-to-day basis after discharge from colorectal surgery. This was a retrospective analysis of patients readmitted within 30-days after colorectal surgery. Reasons and factors associated with readmission each day after discharge were evaluated. Early readmitted patients (day 0-5 postdischarge) were compared with those readmitted later (day 6-29 postdischarge). The study was conducted at a tertiary center. Patients included those who had undergone primary colorectal resection from the American College of Surgeons National Surgical Quality Improvement Program (2012-2013). The study intended to identify factors associated with any early versus late hospital readmission and to evaluate diagnoses for unplanned readmissions on a day-to-day basis after discharge. For 69,222 elective colorectal procedures, 7476 patients (10.8%) were readmitted to the hospital within 30 days. Early (median, 3 days) and late (median, 11 days) readmissions were 3278 (43.8%) and 4198 (56.2%). Except for sex, patient demographics were similar between groups. Neurologic comorbidity; wound disruption; sepsis or septic shock; unplanned reintubation and reoperation; anastomotic leak and ileus; and neurological, cardiovascular, and pulmonary complications were significantly higher in the early readmission, whereas disseminated malignancy, stoma creation, and renal/urological complications were significantly higher in the late readmission group. On multivariable analysis, early readmission was significantly associated with male patients, wound disruption, sepsis or septic shock, reoperation, reintubation, and postoperative neurological complications. Disseminated malignancy, ostomy creation, and postoperative renal dysfunction/urological infection were associated with delayed readmission. Thirty-day readmissions and reasons for unplanned rehospitalizations were evaluated. Differing factors are associated with early versus late readmission after colorectal resection. These data suggest that early readmission is intricately related to patient and operative complexity and hence may be inevitable, whereas delayed hospital presentation is associated with identifiable perioperative predictors at the time of discharge and hence more likely to be targetable.