Obesity, diabetes, and smoking are important determinants of resource utilization in liver resection: A multicenter analysis of 1029 patients

Abstract

OBJECTIVE: To investigate independent contributions of obesity, diabetes, and smoking to resource utilization in patients following liver resection. SUMMARY BACKGROUND DATA: Despite being highly resource-intensive, liver resections are performed with increasing frequency. This study evaluates how potentially modifiable factors affect measures of resource utilization after hepatectomy. METHODS: The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) public-use database was queried for patients undergoing liver resection. Resource variables were operative time (OT), intraoperative transfusion, length of stay (LOS), ventilator support at 48 hours, and reoperation. Bivariable and multivariable linear and logistic regressions were performed. RESULTS: There were 1029 patients identified. Most resections involved less than a hemiliver (599 patients, 58.2%). Mean BMI was 28.0 ± 6.0. Mean OT was 253 ± 122 minutes (range, 27 to 794) but varied by procedure (P < 0.001). Mean LOS was 8.7 ± 10.7 days (range, 0 to 202). Morbid obesity added 48 minutes to OT (P ≤ 0.018), 1.1 units to transfusions (P ≤ 0.049), 2.2 days to LOS (P < 0.001), and accounted for delayed ventilator weaning (odds ratio, 4.5; P ≤ 0.022). Underweight patients had shorter OT, but stayed 3.3 days longer than normal weight patients (P < 0.001). Insulin-treated patients with diabetes had longer OT (P < 0.001), increased transfusions (P < 0.001), and delayed ventilator weaning (odds ratio, 6.7; P < 0.001), while orally-treated patients with diabetes showed opposite trends. Smokers stayed 1.9 days longer (P < 0.001), with increased risk of prolonged ventilation (odds ratio, 3.3; P ≤ 0.002) and reoperation (odds ratio, 2.3; P ≤ 0.015). CONCLUSION: Obesity, diabetes, and smoking are each associated with important components of healthcare expenditure. Education and prevention programs are needed to limit their impact on overall resource utilization. © 2009 by Lippincott Williams & Wilkins.