Presentation (CTI)

OBJECTIVES: Minimally invasive surgery (MIS) is associated with lower mortality, length of stay, and hospital re-admission compared to open surgery. Disparities in access to MIS are recognized across the United States. Our aim was to evaluate the community need evaluating cancer incidence and access to MIS, by examining whether community characteristics and cancer incidence are significant predictors of open surgical modality.
METHODS: Hospital catchments were defined by groups of ZCTAs that shared a minimum distance to a surgical hospital. Using data from IQVIA, we identified hospitals and procedure modality within the catchment that performed at least one of the following: hysterectomy, prostatectomy, pulmonary lobectomy, colorectal procedures, and partial nephrectomy. Cancer incidence, comorbidities, Social Vulnerability Index (SVI) and Rural-Urban Commuting Area (RUCA) scores were obtained from publicly available datasets and the Harvard Dataverse. We evaluated whether cancer incidence within a catchment predicts open surgical rates. A spline regression determined the inflection point of cancer incidence and open surgery rates to define an open surgical rate threshold. Using binomial logistic regression, we identified significant predictors of the open threshold. We stratified the model by urban (RUCA<4) and rural (RUCA>=4) designations.
RESULTS: The average cancer incidence within a catchment was 557/100,000. MIS deserts were defined by an open surgical threshold >=49.3%. Cancer quartiles significantly associated with the threshold when adjusting for RUCA and SVI interaction [OR: High=5.0 (2.2, 11.8), Mid-High: 5.7(2.5, 13.6), Mid-Low: 5.6(2.3, 14)]. A significant linear association exists between cancer quartile and open threshold. When stratifying by rurality, the SVI interaction persists in urban areas only.
CONCLUSIONS: Communities with the highest cancer incidence had significantly higher open surgical rates, RUCA, comorbidities, and age compared to lower incidence. Future studies are needed to further characterize MIS deserts using the physician density and travel distance.