Background: Our objective is to identify factors for inpatient death in patients undergoing resection for colorectal cancer (CRC).
Study design: Unmatched 1:3 case-control study of surgically resected CRC at a tertiary care institution between 2004 and 2018. Variables for multivariate analysis were selected using tetrachoric correlation followed by a least absolute shrinkage and selection operator (LASSO) penalized regression model.
Results: A total of 140 patients were included (N = 35 patients who died inpatient, N = 105 patients who did not die). Patients who died were older, had higher Charlson Comorbidity Index (CCI), higher rates of preoperative anemia, hypoalbuminemia, emergency surgeries, blood transfusion, postoperative vasopressor requirement, anastomotic leak, and postoperative ICU admission than patients who underwent surgical resection without inpatient mortality. Anemia (aOR = 8.62, 1.44-91.58), emergency admission (aOR = 5.71, 1.46-24.36), and ICU admission (aOR 45.51, 8.31-448.4) significantly predicted inpatient mortality when controlled for CCI and hypoalbuminemia.
Conclusions: Surprisingly, it appears that pre-existing anemia and perioperative factors are more important in predicting inpatient mortality of patients undergoing CRC surgery than baseline comorbidity or nutritional status.
Keywords: colorectal cancer; inpatient mortality; outcomes.