Disparate impact of healthcare access restrictions due to COVID-19 pandemic on breast cancer diagnosis and treatment.

Abstract

e13518 Background: The COVID-19 pandemic altered healthcare access avenues for patients. In May 2020, 79% of cancer patients in the US in active treatment experienced some delay in their healthcare due to COVID-19 related restrictions. Hispanic patients are known to be disproportionately affected by their social determinants of health, and therefore are at higher risk of being affected by the pandemic. Our institution, designated as a “Hispanic Serving Institute (HSI)”, is ideally suited to study the impact of the pandemic on this population. Methods: In this retrospective study, the institutional breast cancer database was queried for patients managed between January 1, 2018 and April 18, 2022. Socio-demographic variables, screening, and treatment initiation dates were analyzed to determine the impact of COVID-19 related restrictions. Patients were stratified into groups based on date of diagnosis: Group 1 (pre-pandemic), Group 2 (pandemic), and Group 3 (post pandemic). Statistical analyses were conducted using R statistical software (R version 4.1.3). ANOVA or Kruskal Wallis tests with multiple pairwise comparisons using independent samples t-test or Wilcoxon rank sum test and χ2 tests with multiple pairwise comparisons using Fisher’s exact test were used to compare the groups. Interaction models were created to evaluate the association between 1) Ethnicity and 2) Distance from the treatment center on the time lapses between 1) mammogram and biopsy and 2) biopsy and first treatment separately. Results: Of the 488 patients during study period, 163 (33%) patients were Hispanic. Median (IQR) number of days between biopsy and first treatment was similar across all groups [Group 1=39.5 (102.5) days; Group 2= 50 (121) days; Group 3=57 (54) days - p=0.338]. Patients who lived > 50 miles away from the treatment facility experienced longer time lapses between biopsy date and first treatment compared to those living ≤ 50 miles away (β= 40.60, SE (standard error) =20.87, p = 0.052). Compared to Non-Hispanics, Hispanics in Group 2 experienced a shorter time lapse between mammogram and biopsy (β= -17.34, SE=5.46, p = 0.002). However, these patients experienced a longer time lapse between biopsy and initiation of treatment (β= 43.57, SE=18.25, p = 0.017). There was a significantly higher proportion of non-insured patients in the Hispanic group compared to the non-Hispanic group [Hispanic group=12 (7.4%) patients; Non-Hispanic group=8 (2.5%) patients - p=0.020]. Conclusions: Pandemic related delays in breast cancer treatment were observed mainly in Hispanics and patients living more than 50 miles from the treatment center.