Major Perioperative Bleeding in Patients on Dialysis Undergoing Nonelective Abdominal Surgeries

Patients with end-stage renal disease (ESRD) are at increased risk for bleeding complications following surgery. However, the approach to the preoperative risk assessment and risk reduction, if feasible, in ESRD patients undergoing nonelective abdominal surgery has not been comprehensively studied. We aim to determine the prevalence and risk factors for perioperative bleeding in patients on dialysis undergoing nonelective abdominal surgery.

Using the American College of Surgeons National Surgical Quality Improvement Program 2005-2017 database, we identified patients on dialysis who underwent a variety of nonelective abdominal surgeries by Current Procedural Terminology code. Rates of major perioperative bleeding, defined as bleeding requiring red blood cell transfusion within 72 h after surgery, were calculated and stratified by procedure type. Multivariate logistic regression was used to identify risk factors for major perioperative bleeding. Thirty-day mortality rates were compared between those who had a major perioperative bleed and those who did not.

Of 9102 patients on dialysis undergoing nonelective abdominal surgery, 2793 (30.7%) experienced major perioperative bleeding requiring transfusion and 2002 (22.0%) died within 30 d of surgery. By multivariable logistic regression, patients who were female, independent or partially dependent in activities of daily living, ventilator dependent, had disseminated cancer, or had chronic steroid use at baseline were found to be at elevated risk for major perioperative bleeding. Elevated partial thromboplastin time, blood urea nitrogen, anemia, and hypoalbuminemia were also associated with higher odds of major bleeding. Compared to patients undergoing herniorrhaphy (lowest risk), the odds of major perioperative bleeding were highest for patients undergoing hepatic surgery (odds ratio [OR] = 18.09), splenic surgery (OR = 10.86), and pancreatic surgery (OR = 9.59). Major perioperative bleeding was associated with increased 30-d mortality (34.0% versus 16.7%, P < 0.001).

Patients with ESRD experience high rates of bleeding requiring transfusion following emergent abdominal surgery. Derangements in preoperative laboratories and baseline patient characteristics may be useful in assessing bleeding risk in this patient population.