ENDOSCOPY COVID-19 TESTING REQUIREMENTS DISPROPORTIONATELY AFFECT BLACK PATIENTS

Background: The COVID-19 pandemic has led to major disruptions in healthcare and delays in endoscopy. While recent AGA guidelines suggest against routine pre-endoscopy SARSCoV2 testing, many endoscopy practices continue to require testing of all asymptomatic individuals. We hypothesized that SARS-CoV2 testing may disproportionately affect minority groups. The aim of this study was to assess racial differences in endoscopy cancellation rates attributable to SARS-CoV2 testing requirements. Methods: We conducted a retrospective chart review of cancelled endoscopic procedures between 3/1/21 to 9/7/21 from our hospital which mandated a negative SARS-CoV2 RT-PCR test within 72-hours prior to endoscopy. Data was collected on age, race (based on self-report), gender, procedure type/location, primary language, indication, and source of referral. Cancellations related to SARS-CoV2 testing included: test not completed, test not resulted, patient declined to obtain test, or positive test result. Multiple logistic regression was used to control for potential confounders on race and cancellation. Statistics were performed in JMP Pro 14.0.0. This project was IRB approved. Results: Of 847 cancelled procedures, 300 were randomly selected for chart review (for feasibility purposes). After excluding repeat patients and erroneous cancellations, 255 cases remained for analysis (Table 1). SARS-CoV2 testing requirements accounted for 19% of cancellations (Table 2). Of 112 completed tests, only 7 were positive leading to procedure cancellation. Non-Hispanic Black patients were significantly more likely to have procedure cancellations related to SARS-CoV2 testing (OR: 3.4, 95% CI 1.7, 6.7). This group was also less likely to present for their procedure (OR: 2.3, 95% CI 1.2, 4.1);however, there was no difference in provider initiated cancellations. Other factors including age, gender, procedure type, procedure location, primary language, gastroenterology referral, and indication (screening or diagnostic) were not associated with testing-related cancellations. Controlling for potential confounders of age, gender, endoscopy location, and language did not change the association of race and testing-related cancellations. Discussion: Blacks had higher rates of pre-endoscopy cancellations due to SARS-CoV2 testing requirements compared to non-Blacks which may lead to further delays in care. Similar to prior studies, Blacks also had higher no show rates for endoscopy, which is partly explained by socioeconomic factors. However, the additional requirement of pre-procedure testing disproportionately impacted Black patients. Health care systems mandating pre-endoscopy testing should recognize how this requirement may worsen inequities in care and contribute to poor health outcomes among people of color or other marginalized groups. (Table Presented) (Table Presented)

LIVER FELLOW NETWORK: A SURVEY OF THE IMPACT OF A NOVEL ONLINE MEDICAL EDUCATION TOOL

Introduction Free open-access medical education (FOAMed) resources have proliferated inthe past decade, and the COVID-19 pandemic has accelerated this paradigm shift towardsdigital educational platforms. Liver Fellow Network (LFN) was borne out of a perceivedneed for such resources in hepatology. LFN is an educational website established by gastroenterology(GI) and hepatology fellows, containing original content produced by trainees andpeer-reviewed by faculty advisors. LFN’s content spans a variety of topics and formats (e.g.traditional blog posts, infographics, videos and a podcast). We aimed to assess 1) thedemographics of users and 2) the website’s educational impact.Methods A voluntary, anonymous, web-based survey comprised of 22 demographic andLikert-type questions written in English was disseminated via Twitter and to email subscribersvia MailChimp. No incentives were provided. Responses were collected between 11/3-11/25/20. Additional data was collected from the LFN Twitter account and Google Analytics.This project received a formal determination of program evaluation status according toUniversity of Illinois at Chicago institutional policy;this initiative was deemed not humansubjects research and therefore was not reviewed by the IRB.Results A total of 65 individuals completed the survey. Respondents were primarily 31-45years old (61.5%) and male (61.5%), and spanned all continents except Antarctica, with50.8% residing in the United States (U.S.). GI fellows and attendings comprised 33.8% and30.8%, respectively. According to the website’s Analytics data, 53.1% of all sessions originatedin the U.S., but LFN users span a total of 113 countries (Figure 1). The vast majority ofsurvey respondents agreed that LFN was educationally valuable (Table 1). Additionally, the majority agreed that each of the five core educational series were valuable;that they usedLFN to teach;that they felt more comfortable managing liver disease because of LFN;andthat LFN led to changes in their clinical practice. In 6 months, 8,651 unique users visitedthe LFN website and 2,460 individuals followed the LFN Twitter account. The mean numberof tweet impressions was 120,060 (+/- 51,505) per month with a mean of 3,526 (+/- 1750)profile visits per month. A total of 36 tweets advertising content have garnered a medianof 8,328 (5,166-11,897) impressions/tweet and 761 (391-1,275) engagements/tweet.Discussion LFN has a diverse audience with learners that span six continents. Early surveydata suggest that LFN is easy to use, provides valuable educational information, influencesclinical decision-making, and is utilized as a teaching aid. FOAMed harnesses technologyto improve the training of the next generation of medical providers. LFN stands as anexample of the potential reach of such platforms and how content can be sustainably createdand disseminated.(Figure Presented)Figure 1. Geographic distribution of Liver Fellow Network website users (Table Presented)Table 1. Assessment of Liver Fellow Network’s Educational Impact (n=65)

Patterns and predictors of liver injury in liver transplant recipients with coronavirus disease 2019 (COVID-19)

Background: Recent reports have revealed that coronavirus disease 2019 (COVID-19) is associated with liver injury The burden of liver injury in liver transplant (LT) recipients remains unknown We conducted a multi-center study to evaluate the prevalence, pattern and predictors of liver injury in LT recipients with COVID-19 and its impact on clinical outcomes Methods: The was carried out by the consortium of investigators to study COVID-19 in chronic liver disease (COLD) (registered Clinicaltrials gov NCT04439084) Inclusion criteria constituted: age > 18 years, laboratory confirmed diagnosis of COVID-19 and history of LT. We collected de-identified data on patients diagnosed before May 30, 2020 For the analysis on liver injury, only patients who had laboratory values for aspartate aminotransferase (AST) and alanine aminotransferase (ALT) prior to- and during COVID-19 infection were included The primary outcome was the presence of acute liver injury The secondary outcome was all-cause mortality within 90 days of diagnosis Logistic regression analyses were used to determine interdependent risk factors of primary outcome Results: We included 112 adult LT recipients from 21 US medical centers with confirmed diagnosis of COVID-19 The median age of the cohort was 61 years (IQR 20), and 54 5% (n = 61) were male There were 39 3% (n = 44) Hispanic, 27 7% (n = 31) non-Hispanic white, and 25 9% (n = 29) non-Hispanic African-American The all-cause mortality was 22 3% (n = 25);72 3% (n = 81) were hospitalized and 26 8% (n = 30) were admitted to the intensive care unit (ICU) 81 patients had data for analysis of liver injury 34 6% of LT patients had liver injury, Mild to moderate liver injury (ALT 2-5x ULN) in 22 2% (n=18) and severe (ALT > 5x ULN) in 12.3% (n = 10). Younger age (p = 0 009, odds ratio (OR) 2 06 [1 20-3 54]), Hispanic ethnicity (p = 0 011;OR 6 01[1 51-23 9]), metabolic syndrome (p = 0 016;OR 5 87 [1 38-24 99]), receipt of vasopressors (p = 0 018;OR 7 34 [1 39-38 52]) and antibiotic use (p = 0 046;PR 6 93 [1 04-46 26]) were associated with independent risk of liver injury on multivariate logistic regression Immunosuppression was modified in approximately half the patients (49.4%, [n = 40]) Reduction in immunosuppression during COVID-19 was not associated with liver injury (p = 0 156) or risk for mortality (p = 0.084). Presence of liver injury was significantly and independently associated with higher overall mortality (p = 0 007;OR = 6 91 [95% CI: 1 68-28 48]) in LT recipients Conclusion: Mild to moderate liver injury was common in LT patients diagnosed with COVID-19 Immunosuppression was modified during COVID-19 in half the patients but was not associated with liver injury or mortality Younger age, Hispanic ethnicity, metabolic syndrome, vasopressor and antibiotic use were associated with independent risk of liver injury Lastly, liver injury was associated with higher mortality rate in LT recipients with COVID-19.

Impact of reported level of exposure to COVID-19 patients on physicians in residency and fellowship training programs

Background: During the novel coronavirus-2019 (COVID-19) pandemic, physicians in residency and fellowship training programs are serving as essential healthcare workers while also attempting to continue their preparation for eventual independent practice in their field. We aimed to determine how level of exposure patients with COVID-19 affected the experience of graduate medical trainees in terms of their safety, professional development, and well-being during March and April 2020 Methods: We administered an anonymous, voluntary, web-based survey to physicians enrolled in residency or fellowship training programs in any specialty worldwide A convenience sampling of trainees was obtained through distribution of the survey by email and social media posts from April 20th to May 11th, 2020 To investigate the impact of burden of exposure to COVID-19 the trainee experience, we categorized respondents according to their self-reported estimate of the number of patients with COVID-19 that they provided care for in March and April 2020 (0, 1-30, 31-60, >60). Descriptive statistics were performed and the chi square test was used to evaluate for statistical significance. A multivariable logistic regression analysis was conducted to determine independent predictors of physician burnout Results: Surveys were completed by 1420 trainees, of whom 1031 (73%) were residents Most of the fellows who responded to the survey were training in gastroenterology/ hepatology (27%, 85/280) Trainees who cared for a greater number of COVID-19 patients were more likely to report limited access to PPE and COVID-19 testing and more likely to report testing positive for COVID-19 (Figure 1A) Compared to trainees who did not take care of COVID-19 patients, those who took care of 1-30 patients (adjusted odds ratio [AOR] 1 80, 95% CI 1 29-2 51), 31-60 patients (AOR 3.30, 95% CI 1.86-5.88) and >60 patients (AOR 4.03, 95% CI 2 12-7 63) were increasingly more likely to report burnout More than half (835, 58%) of trainees reported concern about their future preparedness for independent practice Trainees who cared for >60 COVID-19 patients compared to those who did not care for any COVID-19 patients reported similar levels of concern about their preparedness for independent practice (56%, 372/636 vs 58%, 71/125 respectively, p-value 0 57, Figure 1B) Conclusion: Physician trainees who were involved in the care of patients with COVID-19 were more likely to report unsafe working conditions and suffered from higher rates of physician burnout Trainees were concerned about the effects of lost training opportunities on their professional development irrespective of the number of COVID-19 patients they cared for.

Single center assessment of liver injury and mortality based on race and ethnicity in individuals with COVID-19

Background: A significant number of those infected with COVID-19 present with gastrointestinal-related manifestations, particularly acute liver injury Most data originate from China, and it is unclear that similar patterns hold in different ethno-social contexts Methods: We performed a retrospective assessment of all individuals diagnosed with COVID-19 in an ethnically diverse county hospital in Minnesota Individuals with a positive COVID-19 test between March 1 and May 25, 2020 were included We evaluated the role of race, ethnicity, and co-morbidities on the pattern of liver injury, hospital admission, and mortality Logistic regressions were performed using SAS 9.4. Statistical significance was reported at a 05 level Results: 2164 individuals diagnosed with COVID-19 were identified. Median age was 39 years (IQR 28-51) and 52.4% were males. 12.5% were classified as white, 38 6% as African American, 34 6% as Hispanic, and 14 2% as other/unknown Of those admitted to the hospital (N=323), median values for liver markers were as follow: ALT 26 IU/L (IQR 18-45), AST 40 IU/L (IQR 27-63), total bilirubin 0 5mg/dl (IQR 0 3-0 6) and alkaline phosphatase 74 IU/L (IQR 59-100, Table) Only 2 6% and 3 7% of individuals without chronic liver disease presented with elevated ALT or AST, respectively, whereas 9 9% and 14 1% of those with chronic liver disease presented with elevated ALT or AST, respectively Men were more likely to have abnormal ALT and AST (p=0 002 and p=0 005 respectively) Hispanic ethnicity was associated with an elevated ALT on admission, OR 2 5 (95% CI 1 2-5 1) An elevated AST, but not ALT, on admission was associated with an increased OR of mortality, OR 4 2 (95% CI 1 4-12 1), this effect was present after adjusting for race Conclusion: Our findings suggest that elevated AST on admission, but not ALT, was associated with increased risk of mortality related to COVID-19 Overall, minimal hepatic injury was inflicted by COVID-19 in those without concurrent liver disease, regardless of race or ethnicity.