Predicting the unpredictable: comparing readmitted versus non-readmitted colorectal surgery patients

Abstract Background To evaluate readmissions to determine predictors and patterns of readmission. Methods Prospective database review identified readmitted and non-readmitted patients after colorectal surgery. Variables for the index and readmission episode were examined. Results A total of 212 readmissions and 3,292 nonreadmissions were analyzed. The majority was elective. Readmitted patients were older ( P = .003), had more comorbidities ( P < .0001), longer operative times ( P < .0001), length of stay ( P < .0001), and higher costs ( P = .002). At the time of discharge, more readmitted patients required temporary nursing ( P < .0001). Independent readmission predictors were higher American Society of Anesthesiologists score, previous abdominal operation, intensive care unit stay, and dysmotility/constipation surgery. At the time of readmission, 29.2% required reoperation. More than half had an open procedure initially (55.2%). After initial open procedures, reoperative time ( P = .05) and LOS were longer ( P = .028), and more patients required temporary nursing care at the time of discharge ( P = .046). Readmissions caused an additional mean hospital cost of $12,670.89. Conclusions Readmitted patients have distinct demographic and outcomes variables. As most were elective cases, stratifying patients preoperatively may enable perioperative planning for this higher risk group.