Persistent racial disparities in refusal of resection in non-small cell lung cancer patients at high-volume and Black-serving institutions

Abstract

Background: Surgical resection is the standard of care for early-stage non-small cell lung cancer. Black patients have higher surgical refusal rates than White patients. We evaluated factors associated with the refusal of resection and subsequent non-small cell lung cancer outcomes.

Methods: We identified patients with non-small cell lung cancer stages IA to IIIA eligible for surgical resection (lobectomy or pneumonectomy) listed between 2004 and 2017 in the National Cancer Database. We stratified hospitals by the proportion of Black patients served and lung cancer resection volume. We used multivariable regression models to identify factors associated with refusal of resection and assessed 5-year mortality using Kaplan-Meier analysis and Cox proportional hazard modeling.

Results: Of 221,396 patients identified, 7,753 (3.5%) refused surgery. Black race was associated with increased refusal (adjusted odds ratio 2.06, 95% confidence interval 1.90-2.22). Compared to White race, Black race was associated with increased refusal across the highest (adjusted odds ratio 2.29, 95% confidence interval 1.94-2.54), intermediate (adjusted odds ratio 2.05, 95% confidence interval 1.78-2.37), and lowest (adjusted odds ratio 1.77, 95% confidence interval 1.58-1.99) volume tertiles. Similarly, Black race was associated with increased refusal across the highest (adjusted odds ratio 1.97, 95% confidence interval 1.78-2.17), intermediate (adjusted odds ratio 2.08, 95% confidence interval 1.80-2.40), and lowest (adjusted odds ratio 1.53, 95% confidence interval 1.13-2.06) Black-serving tertiles. However, surgical resection yielded similar 5-year survival for Black and White patients.

Conclusion: Racial disparities in non-small cell lung cancer surgery refusal persist regardless of hospital volume or proportion of Black patients served. These findings suggest that a better understanding of patient and patient-provider level interventions could facilitate a better understanding of treatment decision-making.