Should we operate for an intra-abdominal emergency in the setting of disseminated cancer?

ABSTRACT

BACKGROUND: Patients with advanced cancer and an abdominal surgical emergency pose a dilemma, because rescue surgery may be futile. This study defines morbidity and mortality rates and identifies preoperative risk factors that may predict outcome. METHODS: The National Surgical Quality Improvement Program database was queried for patients with disseminated cancer undergoing emergent abdominal surgery (2005-2012). Preoperative variables were used for prediction models for 30-day major morbidity and mortality. A tree model and logistic regression were used to find factors associated with outcomes. A training dataset was analyzed and then model performance was evaluated on a validation dataset. RESULTS: Study patients had an overall 30-day major morbidity and mortality rate of 48.8% and 26%, respectively. The classification tree model for prediction for a morbidity involved the following variables sepsis, albumin, functional status, and transfusion (misclassification rate, 36%). The tree model for mortality showed that an American Society of Anesthesiologists (ASA) score of 4 or 5 with a dependent functional status to be predictive of mortality (misclassification rate, 24%). There was agreement between models for predictive variables. CONCLUSION: The decision to operate for an abdominal emergency in the setting of disseminated cancer is difficult. Our study confirms the high risk for morbidity and mortality in this population. Preoperative factors including sepsis, increased ASA class, low serum albumin level, and patient functional dependence all predict poor outcomes.