A Surgical Clostridium-Associated Risk of Death Score Predicts Mortality After Colectomy for Clostridium difficile

BACKGROUND: A Clostridium difficile-associated risk of death score was recently developed and validated by using a national cohort of both nonsurgical and surgical patients admitted with C difficile infection. However, risk scores specifically derived from surgical cohorts and designed for patients with C difficile infection are currently unavailable. OBJECTIVE: The aim of this study was to develop a risk of death score for patients with C difficile infection who are being considered for total abdominal colectomy because of the failure of medical therapy. DESIGN: This is a retrospective cohort study. SETTINGS: This study was conducted with the use of a national database. PATIENTS: All patients undergoing total colectomy for C difficile infection were identified in the National Surgical Quality Improvement Program database from 2005 to 2014. MAIN OUTCOME MEASURES: Variables similar to the original scoring system were used in multivariable analyses to determine the risk of 30-day mortality for patients, and a model was constructed to estimate the predicted probability of mortality after surgery. RESULTS: Of 532 patients who underwent surgery, 32.7% experienced 30-day postoperative mortality. Patient covariates associated with significantly increased mortality included age greater than 80 years (OR 5.5, p = 0.003), need for preoperative mechanical ventilation (OR 3.1, p < 0.001), chronic steroid use (OR 2.9, p < 0.001), underlying cardiopulmonary disease (OR 2.0, p = 0.001), and acute renal failure (OR=1.7, p = 0.03). These and other comorbidities, including hepatic disease, a cancer diagnosis, and both insulin-and noninsulin-dependent diabetes mellitus, were used to construct a model to estimate the predicted probability of mortality, which ranged from 8.0% to 96.1% based on individual comorbidity profiles. These estimates differed substantially when compared with those obtained using the National Surgical Quality Improvement Program risk calculator, which estimated the risk of mortality among surgical patients as being consistently lower. LIMITATIONS: This study was limited by its retrospective design. CONCLUSIONS: Our surgical scoring system allows preoperative risk stratification for patients being evaluated for colectomy for C difficile infection, potentially helping to avoid futile surgery. See Video Abstract at http://links.lww.com/DCR/A434.