key: cord-0762950-sjfnhgrk
authors: Gawron, Andrew J.; Sultan, Shahnaz; Glorioso, Thomas J.; Califano, Sophia; Kralovic, Stephen M.; Jones, Makoto; Kirsh, Susan; Dominitz, Jason A.
title: Pre-Endoscopy COVID-19 screening and SARS-CoV-2 nucleic acid amplification testing in the VA Healthcare System: Clinical practice patterns, outcomes, and relationship to procedure volume
date: 2022-04-22
journal: Gastrointest Endosc
DOI: 10.1016/j.gie.2022.04.018
sha: d4c1ebe79da339469afda520f2902cd45418bf46
doc_id: 762950
cord_uid: sjfnhgrk

Background & Aims The COVID-19 pandemic has had profound impacts worldwide, including upon the performance of gastrointestinal endoscopy. We aimed to 1) describe the performance and outcomes of pre-endoscopy COVID-19 symptom and exposure screening and SARS-CoV-2 nucleic acid amplification (NAAT) across the national Veterans Affairs healthcare system and 2) describe the relationship of SARS-CoV-2 NAAT use and resumption of endoscopy services. Methods COVID-19 screening and NAAT results from March 2020-April 2021 were analyzed to determine use, performance characteristics of screening, and association between testing and endoscopic volume trends. Results Of 220,891 completed endoscopies identified, 115,890 (52.5%) had documented pre-procedure COVID-19 symptom and exposure screenings, and 154,127 (69.8%) had pre-procedure NAAT results within 7 days prior to scheduled endoscopy. Of 131,894 total cancelled endoscopies, 26,475 (20.1%) had screening data and 28,505 (21.6%) had SARS-CoV-2 NAAT results. Overall, positive NAAT results were reported in 1.8% of all individuals tested and in 1.3% of those who screened negative. Among completed and cancelled endoscopies, COVID-19 screening had a 34.6% sensitivity (95% CI 32.4%-36.8%) and 96.4% specificity (95% CI 96.2-96.5%) when compared to NAAT. COVID-19 screening had a positive predictive value of 15.0% (95% CI 14.0%-16.1%) and a negative predictive value of 98.7% (95% CI 98.7%-98.8%). There was a very weak correlation between monthly testing and monthly endoscopy volume by site (Spearman rank correlation coefficient = 0.09). Conclusion These findings have important implications for decisions about pre-procedure testing, especially given breakthrough infections among vaccinated individuals during the SARS-CoV-2 delta and omicron variant surge.

The COVID-19 pandemic has had a profound impact on elective procedures, including gastrointestinal (GI) endoscopy. 1, 2 Early in the pandemic, the Veterans Health Administration (VA) acted quickly to cease all elective procedures, prioritizing the safety of Veterans and staff, but this ultimately resulted in significant deferment of care. 2, 3 As the pandemic persisted organizations, including VA, developed detailed guidance for resumption of care, including GI endoscopy. 4 This guidance set minimum standards for infection control precautions (e.g. personal protective equipment (PPE)) based on patient risk stratification, and procedure risk (e.g. aerosol generation)). 5, 6 The protocols and recommendations developed in response to COVID-19 were based on rapidly evolving evidence and expert consensus. Key components of risk assessment for procedures, included screening for presence of symptoms suggestive of COVID -19 7 (based on the Center for Diseases and Control guidance 8 ), risk of potential exposure to contacts with COVID-19, and if available, pre-procedural Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) nucleic acid amplification testing (NAAT).

Pre-procedure testing guidance has evolved over the course of the pandemic. In July 2020, a rapid systematic review and guideline found that a pre-endoscopy testing strategy in asymptomatic individuals could be useful, depending on the local prevalence of the disease. 9 In addition to helping with infection prevention and control, preendoscopy testing could provide reassurance to patients, staff, and endoscopists, and inform decisions regarding PPE. [10] [11] [12] The guidance also highlighted the downsides of a pre-endoscopy testing strategy, which included additional system, provider, and patient J o u r n a l P r e -p r o o f burden related to testing logistics. 13 Updated guidance in May 2021, which preceded the current SARS-CoV-2 Delta variant surge, included consideration of post-vaccination status and conditional recommendations to cease routine SARS-CoV-2 testing prior to endoscopy as a systematic review found asymptomatic prevalence and transmission to be very low and assumed centers have adequate access to PPE and universal screening protocols. 14 The VA is comprised of 170 medical facilities and 1,074 outpatient sites serving over 6 million Veterans annually, 15 accounting for approximately 400,000 endoscopy visits annually prior to the pandemic. To ensure safe resumption of care during the pandemic, the VA instituted a universal screening protocol for COVID-19 symptoms or exposure prior to permitting entry into medical facilities. Many endoscopy units implemented formal pre-procedure screening via telephone (with documentation in the electronic health record) and SARS-CoV-2 NAAT testing before the endoscopy appointment. As the largest integrated healthcare system in the United States, there is an opportunity to learn from the VA's nationwide experiences as the pandemic continues. The objectives of this evaluation are to 1) describe the real-world results from a strategy of preendoscopy COVID-19 symptom and exposure screening and SARS-CoV-2 NAAT testing across the national VA healthcare system, including prevalence of asymptomatic infection among Veterans undergoing endoscopy and the sensitivity and specificity of COVID-19 screening, and 2) geographically illustrate the use of pre-endoscopy SARS-CoV-2 NAAT testing and explore its association with endoscopy procedure volumes across the VA. J o u r n a l P r e -p r o o f minimize cancellations for other reasons, we only included those endoscopy appointments that were cancelled within 7 days of the scheduled appointment date as this corresponded to the timeframe of pre-procedure screening and testing. While most appointments in the endoscopy clinic are for colonoscopy or EGD, we could not definitively exclude appointments for other uncommon indications (e.g., esophageal pH testing). VA uses 3-digit codes to classify outpatient specialty clinics, 19 including endoscopy. However, some small facilities sites perform endoscopy in the operating room. With these limitations in mind, we used the following 2 approaches to define cancelled endoscopies due to COVID-19: a. Inclusive: Our primary ("inclusive") approach aimed to optimize sensitivity at the risk of lower specificity. The inclusive approach included all clinics that had: (1) at least one completed endoscopy within the clinic during calendar years 2019-2020 and either 2a) clinic code for endoscopy assigned to the clinic or 2b) more than 50% of the completed visits within the clinic resulted in endoscopy. b. Conservative: As part of a sensitivity analysis, a second ("conservative") approach was used that likely captured fewer true endoscopy appointments (lower sensitivity) but reduced the number of false positives (higher specificity). The conservative approach only included clinic locations where: (1) 10% or more of the completed endoscopies at the facility occurred at the clinic and (2) 75% or more of the completed visits at the clinic resulted in endoscopy.

Documentation of pre-procedure COVID-19 symptom and exposure screening was identified using data generated from a nationally standardized COVID-19 screening tool implemented in the VA electronic health record (EHR) (Supplemental Figure 1 ). Text string searches were used to extract structured data correlating to the standard template that were then classified into categories for screen results (positive/negative), exposure (yes/no), travel (yes/no), flu (yes/no), cough (yes/no), and fever (yes/no). The 'screen results' category was used to flag positive screens; however, if this information was not available, any 'yes' response to exposure, travel, flu, cough, or fever was identified as a positive screen. The timeframe of documentation of screening was restricted to 7 days prior to endoscopy. Screening the day of the procedure or before were considered screening prior to endoscopy. If multiple screening data were available during the pre-endoscopy period, any positive result was used first. Otherwise, the result closest to the date of procedure was included in the analysis.

SARS-CoV-2 NAAT results are documented in the EHR and were collected from the testing datasets generated by the BASIC program from the VA CDW. Additionally, documentation of positive tests from the community were collected from clinical notes using natural language processing. 20 Like screening, if multiple test results were available during the 7-day pre-endoscopy period, any positive result was used first, followed by the result closest to the procedure.

Descriptive and analytical statistics using R (Version 4.0.3) were performed to determine the following findings:

1. The overall and individual facility implementation and pre-endoscopy timing of COVID-19 screening and exposure, and SARS-CoV-2 NAAT testing results.

2. COVID-19 screening and SARS-CoV-2 NAAT results within 7 days prior to a completed or cancelled endoscopy. To explore the effect of missing data, a sensitivity analysis was also performed limiting data to sites in the top 50% of lowest missing rates for both screening and NAAT results.

3. Geographic location, endoscopy volume, and proportion of patients with NAAT illustrated with geographic information system maps and correlation explored between volume and testing using a within-cluster resampling approach. 21 Using SARS-CoV-2 NAAT results as the reference standard, we estimated the performance of pre-procedure screening using data from all cancelled and completed endoscopies for which there was both screening and PCR data ( A sensitivity analysis limited to the sites in the top 50% with the least missing data for both screening and NAAT results included 91,494 (25.9% of the primary cohort) completed or cancelled appointments at 30 sites. The sensitivity analysis cohort for completed endoscopies had a missing rate of 17% for screening data and 5% for NAAT results compared to 48% and 30%, respectively, in the primary cohort. When considering completed and cancelled endoscopy appointments combined, screening data was missing in 38% and NAAT results were missing for 29% compared to 60% and 48%, respectively, in the primary cohort. There was not a clinically significant change in the test performance results or proportion of patients with negative/positive screening or NAAT results in the top 50% of sites with the least missing screening and NAAT results (supplementary Tables 1 and 2) .

The average 30-day endoscopy volume was determined in relation to the proportion of patients with pre-procedure SARS-CoV-2 testing in completed endoscopies. Figure 4 shows geographic information system (GIS) trend maps to visualize endoscopy volume J o u r n a l P r e -p r o o f and the proportion of patients tested over time. By December 2020, most endoscopy sites utilized SARS-CoV-2 NAAT prior to endoscopy. There was a very weak correlation between monthly testing and monthly endoscopy volume by site across the VA using a within-cluster-resampling approach (Spearman rank correlation coefficient = 0.09, Supplemental Figure) . 21 Using data on completed procedures through April 2021, a total of 71 sites (60.2%) competed testing in 80-100% of procedures, accounting for 120,952 total procedures (54.8% of 220,891 total procedures performed). During the same timeframe, 17 sites (14.4%) were testing 0-20% of individuals pre-endoscopy, accounting for 32,002 total procedures (14.5% of total procedures performed) across the VA system.

The COVID-19 pandemic required swift adaptations in healthcare delivery in the US and globally. 2, 22 While rapid growth in telemedicine mitigated negative effects for many specialties, procedure-based specialties, such as GI endoscopy, were significantly impacted. 23, 24 This report describes the use of pre-procedure COVID-19 symptom screening and viral testing prior to endoscopy, the outcomes of a pretesting strategy, and the temporal relationship with endoscopy volumes within the largest integrated healthcare system in the United States. Our data from over 300,000 procedure appointments show the uptake of pre-procedure testing and documented screening as the pandemic spread across the US. This likely helps explain our NAAT positivity rate of 15.0% in screen-positive patients, much higher than previously reported early in the pandemic. Two cost-effectiveness studies on preendoscopy testing and PPE use found that routine pre-endoscopy testing is more-cost effective as the prevalence of COVID-19 increases. 25, 26 Overall, 1.3% of endoscopy patients with negative screening tested positive for SARS-CoV-2, similar to an observational study recently reported from the Netherlands evaluating pre-endoscopy screening and testing prior to elective surgery. 27 Prior data early in the pandemic on pre-endoscopy SARS-CoV-2 testing and screening showed similarly high negative predictive value for symptom screening but low positive predictive value. 28 These VA results provide important information about the discriminative value of preprocedure COVID-19 screening and SARS-CoV-2 NAAT that can be used to optimize pre-procedure protocols for endoscopy as the pandemic continues, driven by SARS-CoV-2 variants and variable vaccination status of patients across the country. The VA guidance provided detailed guidance on screening, testing, and PPE recommendations.

For GI endoscopy, the guidance recommended N95 masks with eye protection or a PAPR irrespective of COVID test status. If COVID status was unknown, then the procedure room cannot be used for another patient until enough air exchanges occur to meet infection prevention rules for downtime to clear the virus. This VA guidance likely helped encourage sites to adopt pre-procedure testing. The variability in reported J o u r n a l P r e -p r o o f testing between sites could be due to a variety of reasons, such as test availability or local testing policy. There is also variability in the proportion of patients with documented screening. However, even if a pre-procedural phone call was not performed or documented, any patient reporting symptoms or exposure would not be permitted entry to a VA facility except to receive care specific to that symptom or exposure. It is possible that VA requirements for symptom and exposure screening for entry into its facilities may have decreased the perceived importance of screening prior to arrival.

We hypothesized that facilities with greater access to testing with NAAT may have been able to offer more endoscopic services. However, our results did not reveal evidence to support a meaningful correlation between endoscopy volume and use of pre-endoscopy NAAT. There were likely multiple factors that played a role in resuming endoscopy volume, including pre-endoscopy testing. Other factors that we are unable to measure include staff availability, PPE availability, local leadership decisions regarding endoscopy services, and of course, COVID burden at individual sites. Pre-procedural SARS-CoV-2 testing, though recommended, is not mandatory in the VA. Variation in practice patterns may have been influenced by local factors including local COVID epidemiology, NAAT availability, turnaround time of test results, patient burden and acceptability, access to PPE and staff and provider comfort level with no testing. 9 While sites across VA have some access to SARS-CoV-2 testing with rapid turnaround, the volume of rapid testing that can be accomplished varies from site to site and some facilities may have opted to limit testing to high risk aerosol generating procedures (e.g.

EGDs), while others may have tested prior to all GI procedures. The ability to prioritize pre-procedural testing for turnaround that is fast enough to allow for clinical decision making (i.e., postponement of the procedure) may be negatively impacted in areas where prevalence of SARS-CoV-2 is high and testing demands are also high. At smaller facilities, there may be less capacity for in-house testing and turnaround time for nonurgent testing may be longer than the recommended 72-hour window. 5 Thus, any factor affecting time-to-result of SARS-CoV-2 testing limits utility in pre-procedural risk assessment and may lead to a decision not to implement a pre-procedure testing strategy. Furthermore, prevalence of COVID-19 in the community may influence test performance and, in areas of low prevalence, the false positive rate may be an additional limitation to testing.

Recently updated guidance to stop COVID-19 testing prior to endoscopy was based on a systematic review finding very low rates of asymptomatic prevalence and transmission (0-0.5%). 14 Between March and December 2020, 93,558 Veterans with a negative COVID screen underwent NAAT prior to endoscopy appointments and 1.3% were positive. In the early months of the pandemic, there was much uncertainty on best practices as knowledge of the virus evolved, vaccine was not available, and there were significant shortages of PPE. Pre-endoscopy screening and testing provided benefits to inform PPE distribution for lower risk (non-aerosol generating) procedures and allowed safe resumption of elective endoscopy by protecting and reassuring staff and patients.

Our data shows that symptom screening has a PPV of 15.0% with a negative predictive value of 98.7% and supports continuation of this practice across VA prior to endoscopy, J o u r n a l P r e -p r o o f particularly given the relative ease of implementation compared to pre-procedural testing.

Our report does have several limitations. First, we likely underestimate pre-procedure screening as we were unable to capture episodes of screening that were not documented within the standard screening tool. However, use of a validated text processing workflow developed by VA for COVID-19 surveillance was used to improve sensitivity. 20 Second, we did not explore screening and testing prior to advanced procedures, such as EUS and ERCP. Third, appointment data, used to identify cancelled endoscopy, do not provide a clear intended purpose for the appointment, though misclassification would be unlikely to significantly alter our findings. Historically, VA has a cancellation rate of approximately 30%. At the onset of the pandemic, this rate peaked at over 70% before returning to near baseline levels in late 2020 (Supplemental Figure) . Though we are unable to determine an exact reason for cancellation in most cases, the 4% cancellation rate attributed to COVID represents a floor as there may be others who cancelled due to COVID concerns that were not reported. Next, we may have underestimated the use of pre-endoscopy SARS-CoV-2 NAAT if testing occurred outside of the VA. We only examined COVID tests within 7 days of the procedure and did not measure or include NAAT tests done more than 7 days pre-endoscopy. NAAT tests performed prior to this time frame may have impacted clinical decision making and added to the reasons for procedure cancellation. However, the overall US population rate for COVID-19 disease remained at less than 5% of the population during this time and would likely have had little overall effect on cancellation rates in our study. 32 J o u r n a l P r e -p r o o f Finally, missing data on screening and NAAT could be non-random, leading to biased conclusions. It is reassuring that a sensitivity analysis that was restricted to 30 sites with the least missing data yielded very similar results.

This report has several strengths including a very large sample size spanning endoscopy centers that are geographically distributed throughout the U.S., accounting for variability in COVID prevalence across the nation and throughout the first 13.5months of the pandemic. Furthermore, the inclusion of screening and testing data from cancelled appointments is critical to estimating the impact upon endoscopy services.

The presence of varying practices regarding SARS-CoV-2 screening and testing, where testing policies are determined at the local level based upon local risk analysis, illustrates generalizability and real-world applicability. Although practice settings differ within and outside VA, most healthcare systems have ability to incorporate intake or pre-procedure COVID screening analogous to intake protocols that assess medication use and NPO status. SARS-CoV-2 testing availability does depend on local resources, including widespread availability of rapid antigen tests as an option for consideration.

In summary, we describe the results of pre-endoscopy COVID-19 screening and SARS-CoV-2 NAAT use in the VA healthcare system during the first 13.5 months of the pandemic. Numerous questions and areas of future study remain, including those surrounding unintended consequences related to pretesting requirements prior to receiving endoscopy care. For example, postponement of endoscopy may lead to poor patient outcomes, such as delayed cancer diagnoses. 3, 33 Emergence of variant strains J o u r n a l P r e -p r o o f of the virus, such as delta and omicron, with decreased vaccine efficacy and asymptomatic transmission, has prolonged the course of the pandemic, necessitating careful evaluation of the role of screening and testing practices. Our findings can help inform decisions regarding the role of pre-endoscopy screening and testing. In fact, newly updated guidance from the American Society of Anesthesiologists recommend that all patients undergoing an aerosolizing procedure (such as EGD) should continue to have preoperative PCR testing for SAR-CoV-2, ideally withing 3 days of the procedure, irrespective of vaccination status. 34 Any perceived decrease in safe practices in endoscopy centers may lead to heightened anxiety among patients and staff which could also negatively impact patients' willingness to undergo endoscopy and/or staff morale and performance. 12 In summary, this report illustrates one of many adaptations made within the VA as a learning healthcare system that seeks to continuously generate and apply evidence and innovation to provide high quality care during an unprecedented time as we continue into the second, and hopefully final, year of the COVID-19 pandemic. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f 

Impact of the COVID-19 pandemic on the workflow of an ambulatory endoscopy center: an assessment by discrete event simulation

The Impact of the Coronavirus Disease-19 Pandemic on Access to Endoscopy Procedures in the VA Healthcare System

Impact of COVID-19 on colorectal cancer disparities and the way forward

Gastrointestinal endoscopy during the COVID-19 pandemic: an updated review of guidelines and statements from international and national societies

Veterans Health Administration Moving Forward Guidebook: Safe Care is our Mission. Veterans Health Adminstration

COVID-19: Guidance for elective gastroenterology and hepatology procedures. Veterans Health Administration, 2020. Only available on the VA intranet site

Preventing the spread of COVID-19 in digestive endoscopy during the resuming period: meticulous execution of screening procedures

Guidance for COVID-19

AGA Institute Rapid Review and Recommendations on the Role of Pre-Procedure SARS-CoV-2 Testing and Endoscopy

Gastrointestinal endoscopy during the coronavirus pandemic in the New York area: results from a multi-institutional survey

Most Patients Are Willing to Undergo Elective Endoscopic Procedures During the Reopening Period of the Coronavirus

Endoscopy Staff Are Concerned About Acquiring Coronavirus Disease 2019 Infection When Resuming Elective Endoscopy

Implementation and Impact of Universal Preprocedure Testing of Patients for COVID-19 Before Endoscopy

Department of Veterans Affairs, Office of Research & Development Program Guide: 1200.21, 2019. 17. National Surveillance Tool assesses readiness across VA's health system. VAntage Point

VHA Directive 1158: Clinical Assessment Reporting and Tracking Program for Invasive Cardiac Procedures. Department of Veterans Affairs

Are VHA administrative location codes valid indicators of specialty substance use disorder treatment?

A Natural Language Processing System for National COVID-19 Surveillance in the US Department of Veterans Affairs

Pearson's chi-square test and rank correlation inferences for clustered data

The global impact of COVID-19 on gastrointestinal endoscopy units: An international survey of endoscopists

Impacts of the Coronavirus 2019 Pandemic on Gastrointestinal Endoscopy Volume and Diagnosis of Gastric and Colorectal Cancers: A Population-Based Study

Estimating the Backlog of Colonoscopy due to Coronavirus Disease 2019 and Comparing Strategies to Recover in Ontario

Cost-effectiveness analysis of SARS-CoV-2 infection prevention strategies including pre-endoscopic virus testing and use of high risk personal protective equipment

Effectiveness and cost-effectiveness of four different strategies for SARS-CoV-2 surveillance in the general population (CoV-Surv Study): a structured summary of a study protocol for a clusterrandomised, two-factorial controlled trial

Yield of Screening for COVID-19 in Asymptomatic Patients Before Elective or Emergency Surgery Using Chest CT and RT-PCR (SCOUT): Multicenter Study

Outcomes of symptom screening and universal COVID-19 reverse transcriptase polymerase chain reaction testing before endoscopy in a community-based ambulatory surgery center

Outcomes of Universal Preprocedure Coronavirus Disease 2019 Testing Before Endoscopy in a Tertiary Care Center

Yield and Implications of Pre-Procedural COVID-19 Polymerase Chain Reaction Testing on Routine Endoscopic Practice

Preprocedural SARS-CoV-2 Testing to Sustain Medically Needed Health Care Delivery During the COVID-19 Pandemic: A Prospective Observational Study

Screening and Surveillance Colonoscopy and COVID-19: Avoiding More Casualties

American Society of Anesthesiologists. American Society of Anesthesiologists and the Anesthesia Patient Safety Foundation Statement on Perioperative Testing for the COVID-19 Virus

The contents of this work do not represent the views of the Department of Veterans Affairs or the United States Government