key: cord-1045392-ys6s9rps
authors: Sultan, Shahnaz; Lim, Joseph K.; Altayar, Osama; Davitkov, Perica; Feuerstein, Joseph D.; Siddique, Shazia M.; Falck-Ytter, Yngve; El-Serag, Hashem B.
title: AGA Institute Rapid Recommendations for Gastrointestinal Procedures During the COVID-19 Pandemic
date: 2020-04-01
journal: Gastroenterology
DOI: 10.1053/j.gastro.2020.03.072
sha: 06c0e35690f8ac9fec0f075cf13ef7b5ac7405fe
doc_id: 1045392
cord_uid: ys6s9rps

nan

In early December 2019, a series of pneumonia cases was reported in Wuhan, China resulting from a novel coronavirus infection designated as SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2) by the International Committee on Taxonomy of Viruses (ICTV) as of January 7, 2020, and named coronavirus-19 disease (COVID- 19) by the World Health Organization (WHO) as of February 11, 2020. 1 SARS-CoV-2 is a novel enveloped RNA betacoronavirus, that represents the seventh member of the coronavirus family, which includes four common human coronaviruses (229E, NL63, OC43, HKU1) and two other strains including SARS-CoV and MERS-CoV. 2,3 SARS-CoV-2 has approximately 79% and 50% phylogenetic similarity to SARS-Co-V and MERS-CoV, respectively. 2 This virus is suspected to have a zoonotic origin and is estimated to have resulted in 591,802 cases in 176 countries with 26,996 deaths as of March 27, 2020. 4 was first reported in the United States (U.S.) on January 20, 2020 and accounted for a total number of 100,717 cases and 1544 deaths as of March 27, 2020. 4 The morbidity and mortality associated with COVID-19 exceeds previous coronavirus infection outbreaks including SARS (8, 098 infections, 774 deaths) and MERS (2,458 infections, 848 deaths). 5, 6 An initial analysis of 72,314 cases from China revealed that an estimated 81% of infections are characterized as mild, 14% are severe, and 5% are critical (defined as respiratory failure, septic shock, and/or multiple organ dysfunction or failure), with an overall fatality rate of 2.3%. 7 In the U.S., an analysis of 4,226 cases from the Center for Disease Control and Prevention (CDC) as of reported estimated rates of hospitalization (20.7-31.4%), Intensive Care Unit (ICU) admission (4.9-11.5%), and case fatality (1.8-3.4%). 8 The WHO declared a global health emergency on January 30, 2020 9 and pandemic status on March 11, 2020, respectively. 10 The most common presenting symptoms for COVID-19 include fever, cough, and shortness of breath, although other frequently observed symptoms include fatigue, headache, and muscle soreness. Extrapulmonary symptoms may occur early in the disease course. Gastrointestinal (GI) symptoms, including anorexia, nausea, vomiting, abdominal pain, and/or diarrhea may occur early, but are rarely the sole presenting feature 11 ; GI symptoms may be associated with poor clinical outcomes including higher risk of mortality. 11 Of note, the first reported case of COVID-19 in the U.S. presented with a 2-day history of dry cough, fatigue, nausea and vomiting, followed by diarrhea on hospital day #2, with subsequent confirmation of SARS-CoV-2 in a stool specimen. 12 Subsequent studies have confirmed positive SARS-CoV-2 cases using real-time reverse transcriptase polymerase chain reaction (rRT-PCR) in stool specimens of patients with COVID-19 infection [13] [14] , with immunofluorescence data demonstrating that ACE2 (angiotensin converting enzyme II) is abundantly expressed in gastric, duodenal, and rectal epithelia, thereby implicating ACE2 as a potential viral receptor for entry to uninfected host cells, and raising the possibility for fecal-oral transmission although it is unclear if the viral concentration in the stool is sufficient for transmission. 14 Furthermore, ACE2 receptors may additionally be expressed in hepatic cholangiocytes, potentially permitting direct infection of hepatic cells, and early cohort studies of COVID-19 have revealed that abnormal liver enzymes are commonly observed. 15 

Multiple questions have been raised regarding the gastrointestinal and liver manifestations of COVID-19 infection, and implications of SARS-CoV-2 infection on gastrointestinal endoscopy. A joint society statement of the American Gastroenterological Association (AGA), the American Association for the Study of Liver Diseases (AASLD), the American College of Gastroenterology (ACG), and the American Society for Gastrointestinal Endoscopy (ASGE) on March 15, 2020 highlighted the potential for SARS-CoV-2 transmission through droplets, an established mode of transmission, and possibly fecal shedding, and the associated risk for transmission to endoscopy personnel during gastrointestinal endoscopy procedures. 16 In this document, we seek to summarize the data and provide evidence-based recommendation and clinical guidance. This rapid recommendation document was commissioned and approved by the AGA Institute Clinical Guidelines Committee (CGC), AGA Institute Clinical Practice Updates Committee (CPUC), and the AGA Governing Board to provide timely, methodologically rigorous guidance on a topic of high clinical importance to the AGA membership and the public.

This rapid guideline was developed by gastroenterologists and guideline methodologists from the AGA CGC and CPUC, who were assembled on March 15, 2020 in collaboration with the AGA Governing Board to define time-urgent clinical questions, perform systematic reviews, develop summary evidence profiles, and formulate rapid recommendations. Additionally, to ensure representation of the public/consumer, this guideline was reviewed by two COVID-19 positive patients. Panel members disclosed all potential conflicts of interest according to the AGA Institute policy.

The target audience of these guidelines includes gastroenterologists, hepatologists, advanced practice providers, nurses, and other healthcare professionals involved in GI endoscopy. Patients, the public, as well as policy makers may also benefit from these guidelines. These guidelines are not intended to impose a standard of care for individual institutions, healthcare systems or countries. They provide the basis for rational informed decisions for patients, parents, clinicians, and other health care professionals in the setting of a pandemic.

This rapid review and guideline was developed using a process described elsewhere. 17 Briefly, the AGA process for developing clinical practice guidelines uses the GRADE framework and best practices as outlined by the National Academy of Medicine (formerly known as the Institute of Medicine) and Guidelines International Network (GIN). 18

With the help of an information specialist, we electronically searched OVID Medline to identify all relevant English studies from inception to March 23, 2020 

One reviewer (SS) screened titles and abstracts and retrieved relevant articles for each question. A second reviewer (OA, PD, JF, SMS) confirmed the selected studies and, in certain circumstances, conducted additional Google scholar searches to identify relevant articles. The following websites were also reviewed for relevant articles: WHO and CDC. Pairs of reviewers extracted the data from the primary studies identified from existing systematic review documents, reviewed the judgments for risk of bias and conducted specific subgroup analyses using Review Manager. 19 Certainty in the Evidence Evidence profiles were used to display the summary estimates as well as the judgments about the overall certainty of the body of evidence for each clinical question across outcomes. Within the GRADE framework, evidence from randomized controlled trials (RCTs) start as high-certainty evidence and observational studies start out as lowcertainty evidence but can be rated down for several reasons: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Additionally, evidence from well conducted observational studies start as low certainty evidence but can be rated up for large effects or dose-response. Judgments about the certainty were determined via video conference discussion to achieve consensus. The certainty of evidence was categorized into 4 levels ranging from very low to high (see Table 1 ). For each question, an overall judgment of certainty of evidence was made based on critical outcomes.

Evidence to Decision Considerations: During online communications and conference calls, the guideline panel developed several recommendations based on the following elements of the GRADE evidence to decision framework: the certainty of the evidence, the balance of benefits and harms, assumptions about values and preferences, and resource implications. For each guideline statement, the strength of the recommendation and the certainty of evidence to support the recommendation is provided. The words "the AGA recommends" are used for strong recommendations, and "the AGA suggests" for conditional recommendations (see Table 2 ). The panel deliberated over the impact of resource limitations on the feasibility and implementation of these recommendations. Therefore, the panel's main recommendations assume an ideal scenario where there are no resource constraints. However, in settings in which resources require rationing, additional guidance is also provided.

Low confidence in effect estimates may rarely be tied to strong recommendations. Within the GRADE framework, there are 5 paradigmatic situations in which strong recommendations may be warranted despite low or very low certainty of evidence 20 These situations can be conceptualized as ones in which there are clear benefits in the setting of a life-threatening situation, clear catastrophic harms, or equivalence between two interventions with clear harms for one of the alternatives. The panel invoked these paradigmatic situations in developing these recommendations.

Recommendations in this document may not be valid in the near or immediate future. We will conduct periodic reviews of the literature and monitor the evidence to determine if recommendations require modification. Based on the rapidly evolving nature of this pandemic, this guideline will likely need to be updated within the next few months.

Guan et al published the largest cohort study to date which included 1,099 hospitalized patients with confirmed COVID-19 infection from China. They reported that 5.0% of COVID-19 infected patients had nausea or vomiting and 3.8% had diarrhea. 21 30 A recent meta-analysis of 4243 patients from China suggested that approximately 17.6% of patients had any gastrointestinal symptom, including 9.2% with pain, 12.5% with diarrhea, 10.2% with nausea/vomiting. 31 One of the concerns with many of the published studies is the possible duplicate inclusion of the patients across reports, thereby limiting valid performance of pooled estimates in a meta-analysis. 32 There is evidence for the presence of SARS-CoV-2 RNA in stool specimens independent of the presence of diarrhea. Some studies showed that stool continued to be positive for SARS-CoV-2 RNA even after respiratory samples became negative. Chen et al reported a case of COVID-19 based on compatible symptoms and lung imaging in a patient with positive stool real-time RT-PCR for SARS-CoV-2 RNA but negative pharyngeal swabs and sputum samples. Furthermore, Wang et al reported confirmation of SARS-CoV-2 positive fecal samples in 2 patients without diarrhea. 12, 35, 36, 37, 38, 39, 40 What are the liver manifestations of COVID-19? Liver injury is estimated to occur in up to 20-30% of patients at the time of diagnosis with SARS-CoV-2 infection. 14 Severe hepatitis has been reported but liver failure appears to be rare. 39 The pattern of liver injury appears to be predominantly hepatocellular, and the etiology remains uncertain but may represent a secondary effect of the systemic inflammatory response observed with COVID-19 disease, although direct viral infection and drug-induced liver injury cannot be excluded. One study of liver biopsy specimens obtained from a patient with COVID-19 disease revealed microvesicular steatosis and mild lobular and portal activity, suggestive of either SARS-CoV-2 infection or drug-induced liver injury. 41 Abnormal liver enzymes may be observed in both adults and children with COVID-19, 42 and do not appear to be a major predictor of clinical outcomes. 15 Early studies have multiple methodologic limitations, with variable laboratory thresholds, limited longitudinal assessment of liver enzymes, heterogeneous evaluation for alternative etiologies, and limited information regarding baseline liver diseases and confounding variables. Additional studies are needed to further characterize the unique clinical considerations for SARS-CoV-2 infection in patients with chronic liver disease and/or cirrhosis, 43 although preliminary guidance has been provided by the AASLD on March 23, 2020. 44 What are the potential risks to health care workers performing endoscopy? SARS-CoV-2 is presumed to spread primarily via respiratory droplets from talking, coughing, sneezing, and close contact with symptomatic individuals. However humanto-human transmission can occur from unknown infected persons (e.g. asymptomatic carriers or individuals with mild symptoms) as well as individuals with virus shedding during the pre-incubation period before symptoms develop. 45 Data related to the spread of SARS-CoV-2 in the early phase of the pandemic have confirmed that health care professionals are at higher risk of infection than the general population. The WHO and Chinese Center for Disease Control and Prevention (China CDC) reported infection of 2055 health care workers as of February 20, 2020 during the index outbreak in Hubei Province, with health care workers facing a rate of infection approximately three times the general population. 46 This prompted the Chinese Department of Health Reform to deploy more than 40,000 additional health-care workers to the region, preserve personal protective equipment (PPE), and implement surveillance measures and quarantine protocols. 46 Such measures appear to have slowed the spread to health care workers, with recent cases primarily attributable to household contacts rather than occupational exposure. Similar trends have been observed in Europe, with an estimated 20% of COVID-19 infections in Italy occurring in health care workers. 47 Preliminary reports in the US also suggest that health care workers are at risk of nosocomial infections, including infection of 20 health care workers among the first 67 COVID-19 positive individuals in Philadelphia, and additional health care workers cases in WA, NY, and MA. 48, 49, 50 The spread of disease via health care workers is concerning for several reasons: a) appropriate PPE may not be utilized effectively, especially when COVID-19 patients cannot be identified quickly, b) shortage of health care workers due to infection and/or quarantine, and c) the concern of the role of infected health care workers to act as a vector for transmission to patients.

While COVID-19 is spread primarily through droplet transmission, endoscopic procedures can lead to aerosolization and subsequent airborne transmission. Currently there is significant debate about the type of PPE that should be worn by health care workers involved with endoscopy.

This section outlines a series of recommendations addressing PPE recommendations for GI endoscopy personnel in the context of the COVID-19 pandemic. We review the evidence on masks (surgical masks, N95s, or respirator masks), gloves (single versus double), and type of rooms (e.g. negative pressure) that should be utilized when performing endoscopy. All recommendations are included in Table 3 .

Aerosol-generating procedures, procedures that generate small droplet nuclei in high concentrations and permit airborne transmission, include upper GI endoscopic procedures such as esophagogastroduodenoscopy, small bowel enteroscopy, endoscopic ultrasound, endoscopic retrograde cholangiopancreatography (ERCP), breath tests, and esophageal manometry. Aerosolization of viral particles may occur during insertion of the scope into the pharynx during intubation as well as during insertion and removal of instruments through the endoscope channel. 51, 52, 53, 54 The risk of aerosolization of viral particles during lower GI procedures, such as colonoscopy, sigmoidoscopy and anorectal manometry, has been less well studied.

As outlined by the WHO, phases 5 and 6 of a pandemic refer to sustained community outbreaks at a global level with human-to-human transmission. 55 Once community spread has been established in these pandemic phases and there is documentation of spread via asymptomatic individuals, pre-screening checklists have limited utility. Additionally, given the currently limited COVID-19 testing in the US, individuals at-risk of spreading disease cannot be easily identified. 45 Our panel acknowledges that recommendations may change if rapid testing is available, and GI patients can be tested prior to undergoing procedures. However, all patients undergoing endoscopy should be considered potentially infected or capable of infecting others.

Surgical masks (also known as medical masks) are used often for droplet precautions, as they are designed to block large particles, but are less effective in blocking smaller particle aerosols (<5 μm). Unlike surgical masks, respirator masks are designed to block aerosols. Respiratory protection in health care for airborne precautions commonly follows two filtering device paths, N95 mask respirators and powered air-purifying respirators (PAPRs). The N95 masks filter at least 95% of aerosols (<5 μm) and dropletsize (5 μm to 50 μm) particles and are not resistant to oil. Light-weight, no-hose, powered air-purifying respirators (PAPR) are a highly effective alternative to face masks. Air is forced through a large, multi-layer filter housed in the helmet and provide positive pressure within the face-shield compartment. These devices are approved by US National Institute for Occupational Safety and Hazard (NIOSH) and can provide high level protection from common airborne viruses that exceed N95 face masks without the need for "fit-testing" and have been used in a variety of settings. 56 PAPR also has the advantage of providing head and neck protection. See Figures 1 and 2 .

Airborne isolation rooms utilize negative pressure ventilation to create inward directional airflow to prevent generated aerosols from diffusing outside the room. The door of the room should remain closed except when entering and leaving. An anteroom that contains another sink separates the isolation room and the hallways. The anteroom is utilized to transition patients and health care workers in and out of the room, storage of PPE, and donning and doffing of PPE. The negative pressure rooms are designed to maintain a pressure differential and air flow differential between the isolation room and the anteroom in addition to a minimum number of air changes per hour. 57

Recommendation 1: In health care workers performing upper GI procedures, regardless of COVID-19 status*, the AGA recommends use of N95 (or N99, or PAPR) masks instead of surgical masks, as part of appropriate personal protective equipment (Strong recommendation, moderate certainty of evidence) Recommendation 2: In health care workers performing lower GI procedures, regardless of COVID-19 status*, the AGA recommends the use of N95 (or N99 or PAPR) masks instead of surgical masks as part of appropriate personal protective equipment. (Strong recommendation, low certainty of evidence) Recommendation 3: In health care workers performing any GI procedure, in known or presumptive COVID-19 patients, the AGA recommends against the use of surgical masks only, as part of adequate personal protective equipment (Strong recommendation, low certainty of evidence) *These recommendations assume the absence of widespread reliable and accurate rapid testing for the diagnosis of COVID-19 infection or immunity Summary of the Evidence Our systematic literature search did not identify any studies that provided direct evidence to inform our clinical questions for PPE in COVID-19. However, several studies from the SARS outbreak were identified that provide indirect evidence. The SARS outbreak reinforced the vital role of PPE in protecting health care workers from occupationally acquired infection. We used data from two existing systematic reviews by Offeddu 2017 and Tran 2012 to inform our recommendations. 58, 59 First, the systematic reviews by Offeddu et al included a meta-analysis of 3 observational studies that showed a benefit in using N95 respirators over standard masks in protecting health care workers from SARS (OR = 0.86; 95% CI: 0. 22 Table 4B ). Second, a systematic review from Tran et al revealed an increased risk of viral transmission in health care workers performing aerosol-generating procedures (mostly bronchoscopy or tracheal intubation). 59 (Supplemental Figure 3 ). Zamora and colleagues investigated the amount of contamination on the neck and face from individuals using a PAPR mask (in combination with N95) compared with a N95 mask alone 60 ; Individuals who used the PAPR-based strategy experienced a lower risk of face and neck contamination compared to N95 mask alone (RR = 0.08; 95% CI: 0.03-0.19). See Evidence Profile Table 5 , Supplemental Figure 4 . Limitations of these studies include small numbers of health care workers, and data on tracheal intubation or bronchoscopy, not GI endoscopy.

To estimate the risk of viral transmission in endoscopic procedures, we examined data evaluating non-GI aerosolizing-generating procedures such as bronchoscopy and tracheal intubation. Our search strategy did not yield comparative studies on the degree of aerosolization with upper or lower GI endoscopy compared with bronchoscopy or tracheal intubation. However, we assume that insertion of the endoscope into the pharynx and esophagus is likely to be associated with a similar risk of aerosolization of respiratory droplets to that of bronchoscopy.

To inform our estimate of the risk of infection for individuals performing endoscopy, we used evidence from the review by Tran et al which examined the risk of respiratory infections among health care workers from aerosol generating procedures. 59 We conducted an original meta-analysis of retrospective cohort studies identified in this review. The data revealed a higher risk of viral transmission to health care workers exposed to aerosol generating procedures compared to unexposed health care workers (RR = 4.66; 95% CI: 3.13-6.94). Therefore, we recommend utilizing N95s (or masks that are equivalent or better), for all patients regardless of COVID-19 status, given higher risk of transmission during aerosol-generating procedures.

Finally, the panel's decision to extend this recommendation to all patients, regardless of COVID-19 status, is specifically in the context of documented community spread during a pandemic. It also assumes a small proportion of persons who are negative or have recovered from COVID-19; this may change with the availability of wider testing and the ability to test for past infection or immunity. Recent data from China, by Chang et al, revealed the greatest risk of COVID-19 exposure to health care workers during early stages of the pandemic when testing was not yet widely available. 61 In a JAMA report published from Zhongnan Hospital in Wuhan, 29.3% (40 of 138) of COVID-19 infected patients were health care workers who presumably had hospital-acquired infections. 27 Among 493 health care workers caring for hospitalized patients, 10/493 health care workers became infected with COVID-19; all 10 were unprotected health care workers (no mask) caring for patients on medical wards with a low risk of exposure (no known or suspected COVID-19 patients). In contrast, none of the 278 protected health care workers (N95 mask) caring for high risk patients (known or suspected COVID-19) became infected (aOR 464.82; 95% CI: 97.73 to infinite). 62 One study, evaluating health care worker exposure in the care of one COVID-19 positive patient, revealed that none of 41 health care workers (surgical masks only) developed infection despite absence of N95 mask, although studies evaluating health care workers in context of larger cohorts of COVID-19 positive patients are not yet available. 63 The decision to extend the recommendation to lower GI procedures is based on evidence of possible aerosolization during colonoscopy especially during the insertion and removal of instruments through the biopsy channel. 53 and the uncertain risks associated with evidence of the presence of SARS-CoV-2 RNA in fecal samples. These data provided indirect evidence to extend the recommendation to lower GI procedures pending more definitive evidence. 33 Limited resource settings Recommendation 4: In extreme resource-constrained settings involving health care workers performing any GI procedures, regardless of COVID-19 status, the AGA suggests extended use/re-use of N95 masks over surgical masks, as part of appropriate personal protective equipment. (Conditional recommendation, very low certainty evidence).

No direct evidence on the prolonged use or reuse of N95, N99, or PAPR masks in a COVID-19 pandemic was identified. We also did not find indirect comparative evidence on any mask reuse strategies that would impact infection rates and subsequent morbidity and mortality of health care workers. Furthermore, there were no studies on aerosol-generating procedures in context of SARS or MERS. The available evidence was limited to low quality reports evaluating N95 protection in combination with face shield or surgical mask, mathematical models, experimental studies examining decontamination strategies for PPE preservation during pandemics, and laboratory tests evaluating durability and fit endurance of respirator masks.

CDC recommendations during H1N1 pandemic included guidance to use a cleanable face shield or surgical mask over the N95 respirator to reduce contamination and extend respirator use. 64 These strategies were utilized during the SARS outbreak, but the effects of prolonged use of a combination of a face shield or surgical mask over an N95 mask have not been reported. 65 During the H1N1 pandemic, an estimated 40% or more of health care workers reported reuse of their N95 respirator but no data are available to estimate the impact on influenza infections. 66, 67 A mathematical model to calculate the potential influenza contamination of facemasks from aerosol sources in various exposure scenarios revealed that the amount of exposure in a single cough (≈19 viruses) is much lower than that transmitted from aerosols (4,473 viruses on N95 masks, 3,476 viruses on surgical masks). 68 Finally, in laboratory testing, an estimated 5 consecutive donnings of PPE can be performed before fit factors consistently drop to unsafe levels. 46 In addition, in experiments examining decontamination of N95 with hydrogen peroxide and mechanical testing, up to 50 cycles of exposure to hydrogen peroxide did not lead to any degradation of the filtration media but the elastic straps were stiffer after exposure to up to 20 cycles and this could impair proper fit. 69 See Evidence Profile Table 6A and Table 6B ). The data on PAPR re-use after cleaning and disinfection were also limited with select institutions reporting on their experience with established PAPR programs and instructions for cleaning. 70 Discussion and Rationale There is insufficient evidence to comment on the safety of re-use (up to 5 consecutive donnings) and extended use (over 8 hours) of masks and other PPE. Limited indirect evidence suggests loss of durability and fit of N95 masks under these conditions. With regards to PAPRs with disposable protective shields, the protective shields may be disinfected with standard biocidal containing wipes and reused. However, no evidence of safety of such an approach was identified.

Recommendation 5: In health care workers performing any GI procedure, regardless of COVID-19 status, the AGA recommends the use of double gloves compared with single gloves as part of appropriate personal protective equipment (Strong recommendation, moderate quality evidence)

The evidence to support this recommendation is largely derived from observations of health care workers during the SARS epidemic in 2003. Transfer of organisms from contaminated PPE to hands or clothing may contribute to infection of health care workers and associated contacts. Casanova and colleagues performed a human challenge study using the bacteriophage MS2 for simulated droplet contamination. 71 One group of participants donned a full set of PPE with one pair of gloves. The second group donned identical PPE with 2 pairs of latex gloves. The first (inner) pair of gloves was applied so that the wrist of the glove was under the elastic cuff at the wrist of the gown sleeve. The second (outer) pair, one size larger, was worn over the first pair so that the wrist of the glove was positioned over the gown sleeve. During the doffing phase, the inner pair of gloves was removed last. The double-glove strategy was associated with less contamination than the single-glove strategy (RR 0.36; 95% CI 0.16 to 0.78) See Evidence Profile Table 7 , Supplemental Figure 4 .

The Casanova et al study highlights the importance of double gloving as part of the doffing process for PPE with either N95 mask or PAPR to minimize contamination and reduce the risk of viral transmission.

Recommendation 6: In health care workers performing any GI procedure, with known or presumptive COVID-19, the AGA suggests the use of negative pressure rooms over regular endoscopy rooms, when available (Conditional recommendation, very low certainty of evidence).

We did not find any direct evidence to inform this recommendation but indirect evidence was identified to confirm the viability of coronaviruses as an aerosol. In an experimental model, Van Doremalen et al demonstrated that SARS-CoV-2 could remain viable in aerosol form for up to 3 hours, similar to what has been previously reported for the SARS-CoV-1 virus. 72 Epidemiologic and airflow dynamics modeling studies from the SARS 2003 and MERS-CoV outbreaks additionally support airborne spread. 73, 74, 75 As GI procedures may generate aerosols, indirect evidence to support the viability of the SARS-CoV-2 virus in aerosols and airborne transmission support a recommendation in favor of preferential use of negative pressure rooms pending further evidence.

The experimental study by van Doremalen et al further demonstrated that SARS-CoV-2 may stay viable on copper surfaces up to 4 hours, on cardboard surfaces up to 24 hours, and on plastic and stainless steel surfaces up to 72 hours. 72 These data combined with the available epidemiologic and airflow dynamics studies of related coronavirus infections, suggest that GI procedures may contribute to nosocomial transmission of COVID-19. Thus, the use of negative pressure rooms with anterooms may mitigate the spread of the infection within health care facilities. The panel acknowledges that the use of a negative pressure room may impact efficiency and procedural workflow but anticipate that GI procedures performed during the initial pandemic phase will be predominantly limited to time-sensitive procedures performed in hospitalized settings.

In limited-resource settings where negative pressure rooms are unavailable, portable industrial-grade high-efficiency particulate air (HEPA) filters may be a reasonable alternative. Industrial-grade HEPA filters are alternatives suggested by the CDC to enhance filtration when air supply systems are not optimal, when anterooms are not available for patients in airborne isolation rooms, and during intubation and extubation of patients with active tuberculosis patients. 76, 77 

Recommendation 7: For endoscopes utilized on patients regardless of COVIDstatus, the AGA recommends continuing standard cleaning endoscopic disinfection and reprocessing protocols (Good practice statement).

Summary of the Evidence: Current guidelines for infection control during GI endoscopy include mechanical and detergent cleaning, followed by high-level disinfection (HLD), rinsing and drying through sterilization, using FDA-approved liquid chemical germicide solutions. 78 Cleaning must precede HLD to remove any organic debris (e.g., blood, feces, and respiratory secretions) from the external surface, lumens, and channels of flexible endoscopes. Studies examining the natural bioburden levels detected on flexible GI endoscopes show ranges from 10 5 CFU/ml to 10 10 CFU/ml after clinical use; appropriate cleaning followed by HLD (a process that eliminates or kills all vegetative bacteria, mycobacteria, fungi, and viruses, except for small numbers of bacterial spores) reduces the number of microorganisms and organic debris by 4 logs, or 99.99%. 79 Studies examining the risk of viral transmission of hepatitis B, C or HIV among patients have demonstrated a very low risk of transmission. 80 Several cases of patient-to-patient HCV transmission have been reported but these were related to inadequate cleaning and disinfection of GI endoscopes and accessories and/or the use of contaminated anesthetic vials or syringes. A recent review by Kampf et al shows effective inactivation of coronaviruses, including SARS-CoV, by standard biocidal agents, which are active ingredients in current endoscopic disinfecting solutions ( Table 8) . 81 Discussion and Rationale Decontamination of coronavirus species has been confirmed with commonly used biocidal agents for decontamination, such as hydrogen peroxide, alcohols, sodium hypochlorite or benzalkonium chloride. 81, 82 There are ample data to support continuation of current endoscope decontamination practices in the context of known COVID-19. 79 Similar biocidal agents are additionally present in hospital-grade disinfecting wipes commonly used to decontaminate surfaces for endoscopy room cleaning. 81 Guidance on how to implement a triage system See accompanying Flowchart Figure 5 

Data on the urgency of when to perform GI procedures and complications related to delays on patient important outcomes are sparse. Studies in lower GI bleeding suggest little difference in outcomes such as blood transfusions or surgery when comparing urgent colonoscopy (< 24 hours) vs delayed colonoscopy (up to 72 hours after presentation) 86, 87 In a pandemic setting, one might consider opting to delay the procedure (especially while awaiting COVID-19 testing). In contrast, a patient presenting with an upper GI bleed likely should have an EGD performed within 24 hours. 88, 89 The impact of delays in diagnosis may also have significant ramifications on immediate management (e.g. in question of inflammatory bowel disease diagnosis or treatment) and on cancer treatment decisions (e.g. colon cancer, pancreatic cancer etc). Additionally, tests related to treatment of precancerous lesions may also lead to anxiety among patients and providers (e.g. treatment of high-grade dysplasia in Barrett's or an endoscopic mucosal resection for a larger colon polyp). Indirect evidence supports that delays of weeks to a few months in some cancer diagnoses may not lead to progression of stage or worse clinical outcomes even when symptoms are present in some GI cancers. 90, 91, 92 Non-time sensitive procedures are most routine screening and surveillance colonoscopy. There is evidence to suggest that following a positive FIT test, a colonoscopy can be delayed up to six months without negatively impacting patient outcomes. Corley et al. reported on 70,124 patients with a positive FIT test and found no difference in outcome of colorectal cancer diagnosis and advanced stage disease when the colonoscopy was performed in 8-30 days following the test vs waiting up to six months. However, when delaying 7-9 months there was a non-significant increase in risk and a more profound increase risk when delayed > 12 months. Using data from this study, one could suggest that in patients undergoing colorectal cancer screening, even when a test suggests a possible polyp or cancer, delaying the procedure for some period of time may not be harmful on the population level. 93 Discussion/Rationale:

In the setting of a pandemic, the limited availability of resources (such as critical shortages of PPE) combined with the risk of potential exposure and spread of infection to patients and the availability of appropriate health care workers, often become the main drivers for provision of health care services. The proposed framework of separating procedures into time-sensitive and non-time sensitive cases may be useful in determining which procedures if delayed may negatively impact on patient-important outcomes. The panel intentionally chose to focus on patient-important outcomes as a driver for decision-making acknowledging the difficulties with using specific indications to categorize procedures as elective versus non-elective. The panel also acknowledged the limitations of the body of evidence in assessing the time-sensitive nature of endoscopic procedures. While there were data to support a delay of up to 3-6 months for patients undergoing colonoscopy for +FIT and this was likely generalizable to patients undergoing colonoscopy for polyp surveillance, the data to support delays for procedures such EMR for large polyps, are lacking. Moreover, there may be added issues around patient anxiety or worry and concerns about medico-legal risks that may influence decisions about deferring procedures; therefore, the panel suggests the use of a multidisciplinary team approach to facilitate decision-making for complicated patients.

Telemedicine also provides an opportunity to communicate with patients and provide continued patient care while reducing risk of exposure to COVID-19 to patients and health care workers. The AGA and a number of other professional medical organizations have been working to lift restrictions on reimbursement for telehealth visits. 94 The panel chose the time period of 8 weeks based on consensus from the group that some procedures require endoscopy within 24 hours, but others are not as timesensitive and can be delayed in the short-term for a few weeks without affecting important patient outcomes related to the disease state. As there is uncertainty about the duration of the pandemic, a pre-defined time period should be used for reassessment of all deferred procedures especially if resources become available and the time-sensitive nature of the procedure changes.

Moreover, as innovations in testing (rapid tests, serologic tests of immunity) and treatment or vaccines allow for better risk stratification, one may be able to consider restarting non-time sensitive procedures.

The panel also sought feedback from two patients affected by COVID-19 to ensure that we captured the consumer/patient perspective. They understood and agreed with the importance and process of triaging procedures. One patient additionally expressed concerns about the focus on limiting PPE for health care workers when "they are the ones who need the protection the most" and the lack of clear evidence on the variability of GI symptoms.

Clinical guidelines should be informed by a systematic review of evidence and an assessment of the desirable and undesirable consequences of alternative care options.

Rapid guidelines, typically completed within 1-3 months, are needed to provide guidance in response to a time-sensitive need such as during a public health emergency. 95,96 Using a rapid guideline process, the AGA aims to provide timely guidance on appropriate PPE and triage of GI endoscopy in context of the COVID-19 pandemic in the U.S. Due to the paucity of evidence specific to SARS-CoV-2 infection, many questions regarding clinical management remain unanswered, including implications and clinical considerations for vulnerable populations, such as individuals with IBD or other autoimmune GI or liver conditions on immunosuppression, patients with cirrhosis or end-stage liver disease, and individuals with GI malignancies requiring systemic chemotherapy. International registries such as the Surveillance Epidemiology of Coronavirus (COVID-19) Under Research and Exclusion, or SECURE-IBD, (https://covidibd.org), may serve as a valuable data source in the future as clinicians engage in information sharing to inform stronger evidence-based guidance. Ongoing clinical trials for COVID-19 treatment may be associated with GI adverse effects and increase the demands for GI consultative care. Furthermore, the severity and duration of resource limitations for SARS-CoV-2 testing and PPE may further challenge clinical management decisions. Importantly, due to the rapidly evolving nature of the COVID-19 pandemic, these recommendations will likely need to be updated within a short timeframe. We are very confident that the true effect lies close to that of the estimate of the effect.

Moderate We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

Our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect.

We have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect 

The recommendation can be adapted as policy or performance measure in most situations

Policy-making will require substantial debate and involvement of various stakeholders. Performance measures should assess whether decision making is appropriate.

* Strong recommendations are indicated by statements that lead with " we recommend", while conditional recommendations are indicated by statements that lead with "we suggest" All procedures should be reviewed by trained medical personnel and categorized as time-sensitive or not timesensitive as a framework for triaging procedures.

In an open access endoscopy system where the listed indication alone may provide insufficient information to make a determination about the time-sensitive nature of the procedure, consideration should be given for the following options (i) a telephone consultation with the referring provider or (ii) a telehealth visit with the patient or (iii) a multidisciplinary team approach to facilitate decision-making for complicated patients.

Good practice statement *These recommendations assume the absence of widespread reliable rapid testing for the diagnosis of COVID-19 infection or immunity No direct evidence was found in regards to the safety of reuse of masks (surgical masks (SM) and N95) during a COVID-19 pandemic. Furthermore, indirect evidence from other pandemic outbreaks did not reveal empiric data on infection rates, but rather reports of anecdotal experience or experiments under laboratory conditions or mathematical models. Anecdotal reports on using SMs over N95 as a barrier to pathogens and extend the useful life of the N95 respirator has been published 65 . This was sparingly utilized during the SARS outbreak, but the effects of prolonged use of this combination on HCWs and the infection rate have not been reported. Similarly, reports exists that more than 40% of HCWs reused their N95 during the H1N1 pandemic 66, 67 . Furthermore, a mathematical model to calculate the potential influenza contamination of facemasks from aerosol sources in various exposure scenarios, showed that single cough (≈19 viruses) were much less than likely levels from aerosols (4,473 viruses on FFRs and 3,476 viruses on SMs) 68 . In laboratory testing has been reported that 5 consecutive donning's can be performed before fit factors consistently drop to unsafe levels 69 . In addition, decontamination of N95 with hydrogen peroxide has showed that exposure up to 50 cycles does not degrade the filtration media and mechanical testing but has demonstrated that the elastic straps were stiffer after exposure to up to 20 HPV cycles. Thus, more than 20 cycles may impair proper fit 70 . There have been narrative reports, news conference reports and the CDC recommendation 98 during H1N1 pandemic suggesting use of a cleanable face shield or surgical mask to reduce N95 respirator contamination 64 .

Explanations a. Risk of bias: There is no comparator with optimal PPE to understand the risk of the acceptable protection from COVID 19 b. There are multiple layers of indirectness. The population is different -studies were done on Influenza virus or simulation studies on healthy volunteers, and there are no studies on AGP. Outcome is indirect as well; most of these studies have tolerability of the mask or laboratory testing as outcomes. c. Unable to assess for imprecision since outcome cannot be measured. No direct evidence was found in regards to the safety of extended use of masks (surgical masks (SM) and N95) during a COVID-19 pandemic. Furthermore, indirect evidence from other pandemic outbreaks did not reveal empiric data on infection rates, but rather reports of anecdotal experience or experiments under laboratory conditions or mathematical models. Experiment on tolerability of the N95 with prolonged use on HCW showed that HCWs were able to tolerate the N95 for 89 of 215 (41%) total shifts of 8 hr. Other 59% mask was discarded before 8 hr because it became contaminated or intolerance 99 . Furthermore, a mathematical model to calculate the potential influenza contamination of facemasks from aerosol sources in various exposure scenarios, showed that single cough (≈19 viruses) were much less than likely levels from aerosols (4,473 viruses on FFRs and 3,476 viruses on SMs) 68 . Additionally, there was a survey on HCWs during H1N1 pandemic and more than 40 % of the HCWs were reusing or had a prolong use on their N95 66, 67 .

Explanations a. Risk of bias: There is no comparator with optimal PPE to understand the risk of the acceptable protection from COVID 19 b. There are multiple layers of indirectness. The population is different -studies were done on Influenza virus or simulation studies on healthy volunteers, and there are no studies on AGP. Outcome is indirect as well; most of these studies have tolerability of the mask or laboratory testing as outcomes. c. Unable to assess for imprecision since outcome cannot be measured. 

a. Study was done with the bacteriophage MS2, but the drops size was similar to SARS and COVID 19 to simulate droplet contamination, so we decided not to rate down. We recognize that there is some indirectness but we also took into account the large effect size. b. Low event rate 

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Genomic characterization and epidemiology of 2019 novel coronavirus : implication for virus origins and receptor binding

A novel coronavirus from patients with pneumonia in China

Accessed: 3/21/20 (to be updated to date of submission)

SARS -beginning to understand a new virus

Middle east respiratory syndrome coronavirus: risk factors and determinants of primary, household, and nosocomial transmission

Characteristics and important lessons from the Coronavirus Disease 2019 (COVID-19) outbreak in China: summary of a report of 72,314 cases from the Chinese Center for Disease Control and Prevention

Severe outcomes among patients with coronavirus disease 2019 (COVID-19) -United States

China coronavirus: WHO declares international emergency as death toll exceeds 200

COVID-19: towards controlling of a pandemic

Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a descriptive, cross-sectional multicenter study

First case of 2019 novel coronavirus in the United States

COVID-19: gastrointestinal manifestations and potential fecal-oral transmission

Evidence for gastrointestinal infection of SARS-CoV-2

Liver injury in COVID-19: management and challenges

Joint GI society message: COVID-19 clinical insights for our community of gastroenterologists and gastroenterology care providers

AGA Institute clinical practice guideline development process

Institute of Medicine (US) Committee on standards for developing trustworthy clinical practice guidelines; editors. Clinical Practice Guidelines We Can Trust

Review Manager (RevMan) [Computer program]. Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration

GRADE handbook for grading quality of evidence and strength of recommendations

Clinical characteristics of coronavirus 2019 in China

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Analysis of clinical features of 29 patients with 2019 novel coronavirus pneumonia

Epidemiologic and clinical characteristics of novel coronavirus infections involving 13 patients outside Wuhan, China

Clinical characteristics of novel coronavirus cases in tertiary hospitals in Hubei Province

clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China

Epidemiologic features and clinical course of patients infected with SARS-CoV-2 in Singapore

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a descriptive, cross-sectional multicenter study

Gastrointestinal Manifestations of SARS-CoV-2 Infection and Virus Load in Fecal Samples from the Hong Kong Cohort and Systematic Review and Meta-analysis

Editorial concern: possible reporting of the same patients with COVID-19 in different reports

Detection of SARS-CoV-2 in different types of clinical specimens

Evidence for gastrointestinal infection of SARS-CoV-2. Gastroenterology 2020

Viral load of SARS-CoV-2 in clinical samples

First case of 2019 novel coronavirus in the United States

COVID-19 disease with positive fecal and negative pharyngeal and sputum viral tests

Molecular and serological investigation of 2019-nCoV infected patients: implication of multiple shedding routes, Emerging Microbes & Infections

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Liver injury in COVID-19: management and challenges

Pathological finding of COVID-19 associated with respiratory distress syndrome

Clinical and CT features in pediatric patients with COVID-19 infection: different points from adults

Prevention of SARS-Co-V-2 infection in patients with decompensated cirrhosis

Indirect virus transmission in cluster of COVID-19 cases

Report of the WHO-China joint mission on coronavirus disease 2019 (COVID-19)

COVID-19 and Italy: what next? Lancet 2020

COVID-19 testing site opens in South Philadelphia, but with restrictions

With 160 employees in quarantine, Berkshire Medical Center taps 54 temporary nurses

Risk of bacterial exposure to the endoscopist's face during endoscopy

Mucocutaneous exposure to body fluids during digestive endoscopy: the need for universal precautions

Air suctioning during colon biopsy forceps removal reduces bacterial air contamination in the endoscopy suite

Bacterial aerosols during colonoscopy: something to be worried about?

Pandemic Influenza Preparedness and Response: A WHO Guidance Document. Geneva: World Health Organization; 2009. 3, Roles and Responsibilities in Preparedness and Response

The use and effectiveness of powered air purifying respirators in health care: Workshop summary

Guidelines for environmental infection control in health-care facilities. Recommendations from CDC and the Healthcare Infection Control Practices Advisory Committee (HICPAC)

Effectiveness of masks and respirators against respiratory infections in healthcare workers: a systematic review and meta-analysis

Aerosol generating procedures and risk of transmission of acute respiratory infections to healthcare workers: a systematic review

Contamination: a comparison of 2 personal protective systems

Protecting health-care workers from subclinical coronavirus infection. The Lancet Respiratory Medicine

Association between 2019-nCoV transmission and N95 respirator use

COVID-19 and the Risk to Health Care Workers: A Case Report

Reusability of facemasks during an influenza pandemic. Institute of Medicine of the National Academies

Effect of Surgical Masks Worn Concurrently Over N95 Filtering Facepiece Respirators: Extended Service Life Versus Increased User Burden

The Distribution of Fruit and Seed Toxicity during Development for Eleven Neotropical Trees and Vines in Central Panama

Respiratory protection policies and practices among the health care workforce exposed to influenza in New York State: Evaluating emergency preparedness for the next pandemic

Validation and application of models to predict facemask influenza contamination in healthcare settings

The shared leadership process in decisionmaking teams

Final report for the bioquell hydrogen peroxide vapor (HPV) decontamination for 134 reuse of N95 respirators. Prepared for the FDA

Virus transfer from personal protective equipment to healthcare employees' skin and clothing

Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1

A study of the probable transmission routes of MERS-CoV during the first hospital outbreak in the Republic of Korea

Evidence of airborne transmission of the severe acute respiratory syndrome virus

Cluster of SARS among Medical Students Exposed to Single Patient, Hong Kong

Efficacy of portable filtration units in reducing aerosolized particles in the size range of Mycobacterium tuberculosis

Guidelines for environmental infection control in health-care facilities. Recommendations from CDC and the Healthcare Infection Control Practices Advisory Committee

ASGE guideline for infection control during GI endoscopy

Antonoplos PA Natural bioburden levels detected on flexible gastrointestinal endoscopes after clinical use and manual cleaning

Transmission of infection by flexible gastrointestinal endoscopy and bronchoscopy

Persistence of coronavirus on inanimate surfaces and their inactivation with biocidal agents

Antiseptic stewardship: biocide resistance and clinical implications

CDC COVID-19 Response Team. Interim recommendations for us households with suspected/confirmed coronavirus disease

COVID-19: recommendations for management of elective surgical procedures. American College of Surgeons

Surgeon General advises hospitals to cancel elective surgeries

Early colonoscopy does not improve outcomes of patients with lower gastrointestinal bleeding: systematic review of randomized trials

Early versus delayed colonoscopy in hospitalized patients with lower gastrointestinal bleeding

Management of nonvariceal upper gastrointestinal bleeding: guideline recommendations from the international consensus group

Management of patients with ulcer bleeding

Is increased time to diagnosis and treatment in symptomatic cancer associated with poorer outcomes

Does delay in diagnosing colorectal cancer in symptomatic patients affect tumor stage and survival? A population-based observational study

Effect of diagnostic delay on survival in patients with colorectal cancer: a retrospective cohort study

Association between time to colonoscopy after a positive fecal test result and risk of colorectal cancer and cancer stage at diagnosis

WHO: Rapid reviews to strengthen health policy & systems

Introducing the rapid review: how AGA is working to get trustworthy clinical guidelines to practitioners in less time

CMS adult elective surgery and procedures recommendations: limit all non-essential planned surgeries and procedures, including dental, until further notice

ResPECT investigators. N95 respirators vs medical masks for preventing influenza among health care personnel: a randomized clinical trial

Efficacy of various disinfectants against SARS coronavirus

Stability and inactivation of SARS coronavirus

Virucidal Activity of World Health Organization-Recommended Formulations Against Enveloped Viruses, Includ-ing Zika, Ebola, and Emerging Coronaviruses

Databases searched: Ovid MEDLINE: Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE® 1946-Present, Embase Classic+Embase

SARS-CoV-2 or covid19 or covid-19 or covid 19 or (novel adj2 coronavirus) or

Steriliz* or disinfect* or sanitize).ti,ab

31 (personal protective equipment or respirator or respirators or mask*).ti,ab,kw

meta-analysis/ or systematic review/ or meta-analysis as topic/ or "meta analysis (topic)"/ or "systematic review (topic)"/ or exp technology assessment

37 (meta analy* or metaanaly* or health technolog* assess*).ti,ab,kw. 402723 38 (meta-analy* or metaanaly* or systematic review* or biomedical technology assessment* or bio-medical technology assessment*).mp,hw. 768936 39 (((systematic* or methodologic*) adj3 (review* or overview*)) or pooled analysis or published studies or published literature or hand search* or handsearch* or medline or pub med or pubmed or embase or cochrane or cinahl or data synthes* or data extraction* or HTA or HTAs or (technolog* adj

Unexposed HCWs to tracheal intubation as a Risk Factor for SARS Transmission

The authors sincerely thank Kellee Kaulback, Medical Information Officer,

Total events Heterogeneity: Not applicable Test for overall effect: Z = 4.95 (P < 0.00001)

Total events Heterogeneity: Not applicable Test for overall effect: Z = 2.12 (P = 0.03)

Total events Heterogeneity: Chi² = 0.65, df = 2 (P = 0.72); I² = 0% Test for overall effect: Z = 5.56 (P < 0.00001) Test for subgroup differences: Chi² = 0.65, df = 2 (P = 0.72), I² = 0%