T. Seykora1,2, A. Zeymo2, M. Bayasi3, T. DeLeire4, N. Shara5,6, W. Al-Refaie2,3, R. Essig3, K. Chan2 4Georgetown University,McCourt School Of Public Policy,Washington, DC, USA 5MedStar Health Research Institute,Washington, DC, USA 6Georgetown-Howard Universities Center for Clinical and Translational Science,Washington, DC, USA 1Georgetown University Medical Center,Georgetown University School Of Medicine,Washington, DC, USA 2MedStar Health Research Institute,MedStar Georgetown Surgical Outcomes Research Center,Washington, DC, USA 3Georgetown University Medical Center,MedStar Georgetown University Hospital,Washington, DC, USA
Background: Despite growing evidence supporting clinical benefits of minimally invasive surgery (MIS) for colon cancer, this approach is less likely utilized for patients with lower income and Medicaid or no health insurance. It is unclear whether the Affordable Care Act (ACA) affected MIS utilization for colon cancer across various patient socio-demographic characteristics.
Methods: Data on surgical approach and patient characteristics for 193,474 colorectal cancer cases were queried from National Cancer Database for 2011-15. Separate multivariable logistic regression models were used to determine odds of receiving MIS, controlling for patient and clinical characteristics. An interaction term was added to each model to examine the relationship between the post-ACA period and each patient variable (race, insurance, patient zipcode income, education, rurality).
Results: Laparoscopic (LS) and robotic-assisted (RAS) surgery for colon cancer increased over time (Fig. 1). Odds significantly increased (all p<0.001) post-ACA for LS (OR range=1.37-1.46) and RAS (OR range=2.77-3.03) across all models. For LS, Blacks had lower odds (0.90, 0.87-0.94) than Whites, while Asians had higher odds (1.08, 1.01-1.15). The odds of RAS did not differ by race. Uninsured and Medicaid patients had lower odds for LS (0.55, 0.51-0.59; 0.70, 0.65-0.74) and RAS (0.33, 0.25-0.42; 0.63, 0.52-0.76). Patients from higher income zipcodes had greater odds of receiving LS (OR range=1.09-1.34 for 2nd– 4th vs 1st income quartile, all p<0.001) and RAS (1.16, 1.03-1.30, 3rd vs 1st quartile). Similarly, patients from higher education zipcodes had greater odds of LS (1.17, 1.12-1.23, 4th vs 1st quartile) and RAS (3rd: 1.14, 1.02-1.29; 4th: 1.40, 1.23-1.59; vs 1st quartile). More rural areas had lower odds of LS (OR range=0.84-0.89, all p<0.001) with even smaller odds for RAS (OR range=0.47-0.71, all p<0.001). Examining the same variables with a post-ACA interaction term revealed increased odds for Hispanics (1.19, 1.10-1.30), while Blacks had lower odds post-ACA (0.83, 0.72-0.95). There was a further increase in LS odds for higher income (3rd: 1.07, 1.01-1.13; 4th quartile: 1.06, 1.00-1.13; vs 1st) although there were no significant effects for RAS. Post-ACA, non-metro areas had a further reduction in odds for LS (0.90, 0.84-0.96) while small metro increased in odds for RAS (1.13, 1.02-1.25). There were no significant post-ACA effects for insurance or education (LS and RAS).
Conclusion: Utilization of MIS increased substantially 2011-15. However, this reflects underlying trends in overall utilization rather than ACA implementation. Post-ACA, race and insurance disparities generally persisted while pre-ACA differences increased for lower income and non-metro patients.
