University of South Florida Morsani College of Medicine Tampa, Florida
Introduction: Colorectal cancer screening via colonoscopy decreased significantly due to the COVID-19 pandemic, with mail-out fecal immunochemical testing (FIT) initiated to maintain screening. Due to concerns surrounding +FIT follow-up we added FIT navigation (FITNav) via a nurse practitioner who followed +FIT to colonoscopy in August 2020. After implementation we noted little improvement in colonoscopy < 180 days compliance. This prompted a quality improvement (QI) project which resulted in a centralized database. Here we report a subgroup analysis to answer the question: were there racial disparities in +FIT follow-up prior to FITNav implementation?
Methods: We queried +FIT from patients 45-85 y/o from 3/1/2019/20-9/3/2019/20, defined as the pre-pandemic and pandemic cohorts respectively. Patients with dementia & >65 y/o, diagnostic/inpatient FIT, or provider-initiated cancellation of colonoscopy due to comorbidities were excluded. Chart review retrieved FIT indications, patient/navigator notification time, GI consult placement time, and colonoscopy. We added Area deprivation index (ADI) to evaluate neighborhood-level disparities. An adjusted and unadjusted cox regression model was used to evaluate colonoscopy < 180 days between pandemic/pre-pandemic, summarizing via hazard ratios (HR) and 95% confidence intervals (CI).
Results: There were 121 & 103 +FIT meeting criteria in the pandemic & pre-pandemic respectively. Demographics (age, marital status, race, ADI, and sex) between periods showed no statistically significant differences. Proportion receiving colonoscopy < 180 days in the pre-pandemic and pandemic periods was 53.7% and 60.2% (unadjusted HR 1.08, 95% CI 0.76-1.54, p=0.676). This remained insignificant when adjusted for race/ethnicity, marital status, priority group, ADI, time to notification, and age (adjusted HR 1.03, 95% CI 0.71-1.50, p=0.872). While Black, non-Hispanic individuals had a univariate HR of 2.09 (95% CI 1.33-3.29 p=0.001), multivariate HR was 1.59 (95% CI 0.92-2.74, p=0.093). ADI did not show a statistically significant difference upon univariate or multivariate analysis.
Discussion: No findings were present which suggested new or exacerbated racial disparities. Additionally, neighborhood-level disparities did not modify these findings; however, this evaluation is limited by sample size.
