INDEX SCREENING COLONOSCOPY QUALITY METRICS FOR TRAINEES HAVE IMPROVED DURING THE COVID-19 PANDEMIC

Introduction: The coronavirus-2019 (COVID-19) pandemic has led to the suspension of elective outpatient procedures. Clinical educators have raised concerns that gastroenterology fellows have gotten reduced endoscopy exposure with possible decrease in quality metrics. We investigated whether quality parameters such as adenoma detection rate (ADR) and withdrawal times in index screening colonoscopies where fellows were involved have been affected now that elective procedures have resumed in greater numbers. Methods: This is a single-center observational study comparing patients who underwent index screening colonoscopies since our outpatient facilities reopened (7/1/2021 through 9/30/2021) with a pre-pandemic cohort (7/1/2018 through 9/30/2018). We collected demographic information and we tabulated the number of polyps detected, the number of tubular adenomas, and the withdrawal times. We used Chi-square testing to determine significance among our categorical variables and used t-tests to compare means for our numerical variables, particularly between procedures with and without fellow involvement, and between cohorts. All statistical analysis was done in R (R Core Team, 2020). Results: We identified 121 patients in the 2018 cohort without trainee involvement and 38 patients in the same cohort wherein fellows did the procedure. We also identified 186 patients in the 2021 cohort without fellow involvement and 58 patients in the same cohort wherein a trainee did the procedure (Table 1). The 2018 group had an overall ADR of 32.2% (25.0% for women, 39.3% for men) without fellows, and an overall ADR of 21.1% (5.6% for women, 35.0% for men) with trainees. The ADR for women was significantly lower for trainees (p=0.01), but otherwise ADR and withdrawal times were not significantly different. The 2021 cohort, however, showed improvement in all ADRs for fellows compared to 2018, and the ADR for women was equivalent to that of the attendings. Moreover, the withdrawal times were significantly longer for trainees in 2021 (18.9 vs. 15.3 minutes;p=0.01). When comparing fellows' metrics from 2018 and 2021, there was no statistically significant difference in overall ADR, ADR for men, or withdrawal times, though ADR for women had improved (p=0.02). Conclusion: When compared to a similar pre-pandemic cohort, the quality metrics for index colonoscopies have improved for our fellows, particularly with regard to female patients. We are undertaking further data collection to ensure that our findings are stable over a longer time period.(Table Presented) (Table Presented)

INDEX SCREENING COLONOSCOPY QUALITY METRICS HAVE NOT DIMINISHED DURING THE PANDEMIC

Introduction: The ongoing coronavirus-2019 (COVID-19) pandemic has seen a decrease in the overall number of screening colonoscopies performed with the suspension of elective outpatient procedures. We sought to evaluate whether screening colonoscopy quality parameters such as adenoma detection rate (ADR) and withdrawal times have been affected now that such elective procedures have resumed in greater numbers. Methods: This is a singlecenter observational study comparing patients who underwent index screening colonoscopies since our outpatient facilities reopened (7/1/2021 through 9/30/2021) with a pre-pandemic cohort (7/1/2018-9/30/2018). We collected demographic information including patient age, sex, comorbidities, among others, and we tabulated the number of polyps detected, the number of tubular adenomas, and the withdrawal times. We used Chi-square testing to determine significance among our categorical variables and used t-tests to compare means for our numerical variables. All statistical analysis was done in R (R Core Team, 2020). Results: We identified and included 159 patients in the 2018 cohort and another 244 patients in the 2021 cohort (Table 1). The pre-pandemic group included 78 women (49.1%) and had an overall ADR of 29.6% (20.5% for women, 38.3% for men). The 2021 group was larger and included 133 women (54.5%) and had an overall ADR of 32.0% (27.1% for women, 37.8% for men). We found no statistically significant differences in patient characteristics, number of polyps detected, or ADR between the two groups (Tables 1 and 2). There was no significant difference in the odds of detecting an adenoma based on day of the week, morning versus afternoon session, or involvement of a fellow. We found an increase in mean withdrawal time in the 2021 cohort (16.2 minutes vs. 13.6 minutes, p= 0.005) compared to pre-pandemic. Conclusion: The pandemic has not adversely affected the quality of screening colonoscopies performed at our institution, regardless of when the procedure was done or whether a trainee was involved. We are undertaking further data collection to ensure that our findings are stable over a longer time period. (Table Presented) (Table Presented)

SURVEILLANCE COLONOSCOPY METRICS HAVE IMPROVED FROM BEFORE THE PANDEMIC

Introduction: Since the start of the coronavirus-2019 (COVID-19) pandemic, there has been a decrease in the overall number of elective gastroenterology procedures. We sought to evaluate whether surveillance colonoscopy quality parameters such as polyp detection rate (PDR) and withdrawal times have been affected now that such elective procedures have resumed in greater numbers. Methods: This is a single-center observational study comparing patients who underwent surveillance colonoscopies since our outpatient facilities reopened (7/1/2021 through 9/30/2021) with a pre-pandemic cohort (7/1/2018-9/30/2018). We collected demographic information including patient age, sex, comorbidities, among others. We then tabulated the number of polyps detected, the number of tubular adenomas (TA), and the withdrawal times. We used Chi-square testing to determine significance among our categorical variables and used t-tests to compare means for our numerical variables. All statistical analysis was done in R (R Core Team, 2020). Results: Our pre-pandemic cohort included 195 patients (86 women;44.1%), of whom 81 (41.5%) had a TA on their surveillance colonoscopy. Our 2021 cohort consisted of 308 patients (140 women;45.5%), of whom 149 (48.4%) had a TA on surveillance (Tables 1 and 2). The mean PDR was significantly higher in the 2021 cohort (2.2 vs. 1.5 polyps per colonoscopy;p=0.001), and the mean withdrawal time was likewise longer compared to the pre-pandemic cohort (18.9 minutes vs. 15.8 minutes;p=0.002). While the rate of tubular adenoma detection was similar overall in both groups (41.5% vs. 48.4%;p=0.13), it was higher for male patients in the 2021 group compared to 2018 (55.4% vs. 43.1%;p=0.04). We found no difference in polyp or TA detection based on patient comorbidities, age, sex, race, day of the week, morning versus afternoon endoscopy session, or whether a fellow was involved in the procedure. Conclusion: Despite the 1.5-year hiatus in elective procedures, our quality metrics for surveillance colonoscopies have improved from before the pandemic, and this may be related to longer withdrawal times. We have been able to carry out more surveillance procedures compared to a similar pre-pandemic period irrespective of patient demographic factors, procedure timing, and trainee involvement. We are undertaking further data collection to ensure that our findings are stable over a longer time period. (Table Presented) (Table Presented)

FACT OR MYTH: BLACK PATIENTS DO NOT WANT TO PARTICIPATE IN CLINICAL TRIALS

Introduction Underrepresentation of minority groups, particularly Black patients, has been a major issue for most clinical trials. A commonly cited reason is mistrust amongst Black patients due to historical abuse. In a Historically Black College and University (HBCU) at a major metropolitan area with predominant Black patient population, we examined the patient participation rate in a clinical trial compared to other study sites with primarily White patient population. Methods In April 2021, a large prospective, multi-center clinical trial designed to validate a multiomics blood test for early detection of CRC (PREEMPTCRC) was initiated at a HBCU. To optimize study recruitment, culturally-sensitive methods were employed, including racially congruent recruitment staff, and synchronized timing of consent/study procedures with pre-endoscopy COVID testing and clinic visits. Information for all eligible participants screened for the study were recorded and evaluated for a 7- month period (April 1 - October 31, 2021). The enrollment numbers (defined as those consented to the study and had blood samples drawn) for the HBCU and across all other study sites were compared. Demographic and socio-economic data for patients who enrolled and not enrolled at the HBCU were collected to identify potential factors that affect participation. Results During the study period, the number of patients enrolled at the HBCU site (N=229) was significantly higher than the average number enrolled across the other 168 sites (N=90, p<0.0001). In fact, the HBCU site ranked at the top 11th percentile for patient enrollment across all study sites. The main difference between the HBCU site and other study site was race: participants at HBCU were 88.2% Black and 5.2% White, while at the other sites, the participants were 12.0% Black and 71.5% White (p< 0.0001). Comparison of demographic characteristics and socio-demographic data of the enrolled and not-enrolled subjects at the HBCU were similar (Table 1) and did not identify factors that affect participation in clinical trials. Discussion The enrollment of Black patients at a HBCU site was comparable to other study sites in a large prospective, multi-center study of a multiomics blood test for average-risk CRC screening. The findings of our study highlight the importance of providing access to Black patients to clinical trials to ensure adequate representation in research studies. (Table Presented) Table 1. Baseline Patient Demographic and Sociodemographic Information

Enrollment rate of African Americans in a colon cancer screening trial at a historically black college and university is similar to other patient populations

Introduction: Clinical trials often have low enrollment of minorities, particularly African-Americans (AAs), which may limit the generalizability of research findings. Previously identified barriers to AAs recruitment include historical abuses leading to mistrust, communication issues with providers, socio-economic factors, and a lack of access to clinical trials. In a Historically Black College and University (HBCU) serving a primarily AA population at a large safe-net hospital, we evaluated the enrollment of eligible AA patients for a colorectal cancer (CRC) screening clinical trial. This was compared to the enrollment rates across other study sites. Methods: A large, prospective, multi-centered clinical trial to validate a blood-based test for early detection of CRC (PREEMPT-CRC) was initiated at a HBCU, where 84% of patients are AAs. To maximize study recruitment, culturally sensitive methods were employed including racially congruent recruitment staff as well as synchronized timing of consent/study procedures with preendoscopy COVID testing/clinic visits. Detailed information for all eligible subjects was recorded. Demographic and socio-economic data including census information for enrolled and not enrolled subjects were compared. The enrollment rate (defined as enrolled/eligible patients) over the first 6 weeks at the HBCU and that of the other study sites providing screening logs was analyzed. Results: The enrollment rate at the HBCU was 55% (44 out of 80 eligible patients;95% CI 43.5- 66.2%), compared to 49.8% (258 out of 518 eligible patients;95% CI 45.4- 54.2%) at the other 26 study sites. While age and gender of enrolled patients at the HBCU were comparable to other sites, the main difference was race: at the HBCU the study participants were 79.5% AAs and 9.1% whites, while at the other sites the participants were 11.5% AAs and 82.8% whites (p< 0.001). At the HBCU, the demographic characteristics and socio-demographic data including income, marital status insurance status/type, and census tract median household income of the 44 enrolled and 36 notenrolled subjects were similar (Table 1). Conclusion: Contrary to conventional belief that AAs do not want to be involved in clinical trials, we find their enrollment is similar to a predominant white study population when offered the opportunity in a culturally sensitive setting. Future trials should consider including HBCU sites in order to attain adequate AA enrollment to improve the generalizability of research findings. (Table Presented).