Deborah S Keller, MS, MD1, Nina Ferraris, MD1, Irlna Tantchou, MD2, Matt Schultzel, DO2, Eric M Haas, MD, FACS, FASCRS3. 1Colorectal Surgical Associates, 2University of Texas Medical Center at Houston, 3Colorectal Surgical Associates; Houston Methodist Hospital; University of Texas Medical Center at Houston
Background: Despite proven safety and efficacy, rates of laparoscopic surgery for colon cancer remain low in the United States. Given the known clinical and financial benefits, investigation of the root causes of underutilization and methods to increase laparoscopy are warranted. Our goal was to develop a predictive model of the factors that impact use of laparoscopic surgery for colon cancer.
Methods: The Premier Perspective national inpatient database was reviewed from 2010-2014 to identify patients with colon cancer who underwent elective colorectal resection. Patients were identified by ICD-9 diagnosis code, then stratified into open or laparoscopic approaches by ICD-9 procedure codes. Patient, hospital, and surgeon demographics were evaluated. An adjusted multivariate logistic regression model was used to identify variables predictive of use of laparoscopy for colon cancer.
Results: 24,245 patients were included in the analysis- 12,523 (52%) laparoscopic and 11,722 (48%) open. General surgeons performed the majority of all procedures (77.99% open, 71.60% laparoscopic). The overall use of laparoscopy significantly increased from 48.94% to 52.03% over the study period (p<0.0001). From the regression model, patients with private insurance were more likely to have a laparoscopic procedure (OR 1.089, 95% CI [1.004, 1.181], p=0.0388). Obesity had no effect on utilization. Higher volume surgeons (OR 3.518, 95% CI [2.796, 4.428], p<0.0001) were more likely to approach colon cancer laparoscopically. Colorectal surgeons were 32% more likely to approach a case laparoscopically than general surgeons (OR 1.315, 95% CI [1.222, 1.415], p<0.0001). Compared to higher volume hospitals (500+ beds), lower volume hospitals were 22% (less than 100 beds), 15% (100-249 beds) and 19% (250-499 beds) less likely to approach colon cancer laparoscopically. Teaching hospitals were also 21% less likely to perform colon cancer cases through a laparoscopic approach (OR 0.787, 95% CI [0.738, 0.839], p<0.0001).
Conclusions: There are patient, provider and hospital characteristics that can be identified preoperatively to predict who will undergo laparoscopic surgery for colon cancer. These results have implications for regionalization and increasing teaching of minimally invasive surgery. Recognizing and addressing these variables could increase use of laparoscopy, with its clinical and financial benefits.