COVID-19 Treatment: Hydroxychloroquine, Azithromycin, and More, Guías, Proyectos, Investigaciones de Medicina Interna

¡Descarga COVID-19 Treatment: Hydroxychloroquine, Azithromycin, and More y más Guías, Proyectos, Investigaciones en PDF de Medicina Interna solo en Docsity! Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients with COVID-19 Authors Adarsh Bhimraj,* Rebecca L. Morgan,? Amy Hirsch Shumaker,? Valery Lavergne,* Lindsey Baden,* Vincent Chi-Chung Cheng,f Kathryn M. Edwards,” Rajesh Gandhi? Jason Gallagher,? William J. Muller,*% John C. O'Horo,*! Shmuel Shoham,*? M. Hassan Murad,Y Reem A. Mustafa,** Shahnaz Sultan, Yngve Falck-Ytter? Affiliations ¡Department of Infectious Diseases, Cleveland Clinic, Cleveland, Ohio 2Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Ontario 3VA Northeast Ohio Healthcare System, Case Western Reserve University School of Medicine, Cleveland, Ohio 4Department of Pathology and Laboratory Medicine, Vancouver General Hospital, Vancouver, British Columbia, Canada Brigham and Women's Hospital, Boston, Massachusetts $Queen Mary Hospital, Department of Microbiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China 7Division of Infectious Diseases, Department of Pediatrics, Vanderbilt University Medical Center, Nashville, Tennessee 8Infectious Diseases Division, Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts Department of Pharmacy Practice, Temple University, Philadelphia, Pennsylvania Version 4.2.0 0 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. WDivision of Pediatric Infectious Diseases, Northwestern University, Chicago, Illinois MDivision of Infectious Diseases, Joint Appointment Pulmonary and Critical Care Medicine, Mayo Clinic, Rochester, Minnesota 12Johns Hopkins University School of Medicine, Baltimore, Maryland WDivision of Preventive Medicine, Mayo Clinic, Rochester, Minnesota MDivision of Nephrology and Hypertension, Department of Internal Medicine, University of Kansas Medical Center, Kansas City, Kansas 1SDivision of Gastroenterology, Hepatology, and Nutrition, University of Minnesota, Minneapolis VA Healthcare System, Minneapolis, Minnesota Corresponding Author: Adarsh Bhimraj Panel Members: Adarsh Bhimraj (lead), Lindsey Baden, Vincent Chi-Chung Cheng, Kathryn M. Edwards, Rajesh Gandhi, Jason Gallagher, William J. Muller, John C. O'Horo, Shmuel Shoham, Amy Hirsch Shumaker Methodologists: Yngve Falck-Ytter (lead), Rebecca L. Morgan, Valery Lavergne, M. Hassan Murad, Reem A. Mustafa, Shahnaz Sultan Version 4.2.0 1 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. accordance with the terms and conditions of third-party use, in particular any use of the guidelines in any software product. Executive Summary Coronavirus disease 2019 (COVID-19) is a pandemic with a rapidly increasing incidence of infections and deaths. Many pharmacologic therapies are being used or considered for treatment. Given the rapidity of emerging literature, the Infectious Diseases Society of America (IDSA) identified the need to develop living, frequently updated evidence-based guidelines to support patients, clinicians and other health-care professionals in their decisions about treatment and management of patients with COVID-19. Summarized below are the recommendations with comments related to the clinical practice guideline for the treatment and management of COVID-19. A detailed description of background, methods, evidence summary and rationale that support each recommendation, and research needs can be found online in the full text. In brief, per Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology, recommendations are labeled as “strong” or “conditional”. The word “recommend” indicates strong recommendations and “suggest” indicates conditional recommendations. In situations where promising interventions were judged to have insufficient evidence of benefit to support their use and with potential appreciable harms or costs, the expert panel recommended their use in the context of a clinical trial. These recommendations acknowledge the current “knowledge gap” and aim at avoiding premature favorable recommendations for potentially ineffective or harmful interventions. Recommendation 1: Among patients with COVID-19, the IDSA guideline panel recommends against hydroxychloroquine. (Strong recommendation, Moderate certainty of evidence) + Remark: Chloroquine is considered to be class equivalent to hydroxychloroquine. Version 4.2.0 4 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Recommendation 2: Among hospitalized patients with COVID-19, the IDSA guideline panel recommends against hydroxychloroquine plus azithromycin. (Strong recommendation, Low certainty of evidence) + Remark: Chloroquine is considered to be class equivalent to hydroxychloroquine. Recommendation 3: Among hospitalized patients with COVID-19, the IDSA guideline panel recommends against the use of the combination lopinavir/ritonavir. (Strong recommendation, Moderate certainty of evidence) Recommendation 4: Among hospitalized critically ¡ll patients* with COVID-19, the IDSA guideline panel recommends dexamethasone rather than no dexamethasone. (Strong recommendation, Moderate certainty of evidence) + Remark: If dexamethasone is unavailable, equivalent total daily doses of alternative glucocorticoids may be used. Dexamethasone 6 mg IV or PO for 10 days (or until discharge) or equivalent glucocorticoid dose may be substituted if dexamethasone unavailable. Equivalent total daily doses of alternative glucocorticoids to dexamethasone 6 mg daily are methylprednisolone 32 mg and prednisone 40 mg. *Critical illness is defined as patients on mechanical ventilation and extracorporeal mechanical oxygenation (ECMO). Critical illness includes end organ dysfunction as is seen in sepsis/septic shock. In COVID-19, the most commonly reported form of end organ dysfunction is ARDS Recommendation 5: Among hospitalized patients with severe**, but non-critical, COVID-19, the IDSA guideline panel suggests dexamethasone rather than no dexamethasone. (Conditional recommendation, Moderate certainty of evidence) Version 4.2.0 5 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. + Remark: Dexamethasone 6 mg IV or PO for 10 days (or until discharge) or equivalent glucocorticoid dose may be substituted if dexamethasone unavailable. Equivalent total daily doses of alternative glucocorticoids to dexamethasone 6 mg daily are methylprednisolone 32 mg and prednisone 40 mg. **Severe illness is defined as patients with SpO, <94% on room air, including patients on supplemental oxygen. Recommendation 6: Among hospitalized patients with non-severe*** COVID-19 without hypoxemia requiring supplemental oxygen, the IDSA guideline panel suggests against the use of glucocorticoids. (Conditional recommendation, Low certainty of evidence) ***Non-severe illness is defined as patient with a SpO2 > 94% not requiring supplemental oxygen. Recommendation 7: Among hospitalized adults with progressive severe* or critical** COVID-19 who have elevated markers of systemic inflammation, the IDSA guideline panel suggests tocilizumab in addition to standard of care (i.e., steroids) rather than standard of care alone. (Conditional recommendation, Low certainty of evidence) + Remarks: o Patients, particularly those who respond to steroids alone, who put a high value on avoiding possible adverse events of tocilizumab and a low value on the uncertain mortality reduction, would reasonably decline tocilizumab. o In the largest trial on the treatment of tocilizumab, criterion for systemic inflammation was defined as CRP >75 mg/L. Severity definitions: *Severe illness is defined as patients with SpO> <94% on room air, including patients on supplemental oxygen. Version 4.2.0 6 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. + Remark: Baricitinib 4 mg daily dose for 14 days (or until hospital discharge). The benefits of baricitinib plus remdesivir for persons on mechanical ventilation are uncertain. See the remdesivir section for dose and duration. *Severe illness is defined as patients with SpO, <94% on room air, including patients on supplemental oxygen, oxygen through a high-flow device, or non-invasive ventilation. Recommendation 17: Among hospitalized patients with COVID-19, the IDSA guideline panel recommends treatment with baricitinib plus remdesivir plus corticosteroids only in the context of a clinical trial. (Knowledge gap) Recommendation 18: In hospitalized patients with severe COVID-19, the IDSA panel suggests against ivermectin use outside of the context of a clinical trial. (Conditional recommendation, very low certainty of evidence) Recommendation 19: In outpatients with COVID-19, the IDSA panel suggests against ivermectin use outside of the context of a clinical trial. (Conditional recommendation, very low certainty of evidence) Since the inception of its work, the panel has expressed the overarching goal that patients be recruited into ongoing trials, which would provide much needed evidence on the efficacy and safety of various therapies for COVID-19. The panel has determined that when an explicit trade-off between highly uncertain benefits and known putative harms of these therapeutic agents were considered, a net positive benefit was not reached and could possibly be negative (risk of excess harm). The panel acknowledges that enrolling patients in randomized controlled trials (RCTs) might not be feasible for many frontline providers due to limited access and infrastructure. Should lack of access to clinical trials exist, we encourage setting up local or collaborative registries to systematically evaluate the efficacy and safety of Version 4.2.0 9 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. drugs to contribute to the knowledge base. Each clinician can play a role in advancing our understanding of this disease through a local registry or other data collection efforts. Background The first cases of COVID-19 were reported from Wuhan, China in early December 2019 [1], now known to be caused by a novel beta-coronavirus, named as Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Within a span of months, COVID-19 has become pandemic due to its transmissibility, spreading across continents with the number of cases and deaths rising daily [2]. Although most infected individuals exhibit a mild illness (80%+), 14% have serious and 5% have critical illness. Approximately 10% will require hospital admission due to COVID-19 pneumonia, of which approximately 10% will require ICU care, including invasive ventilation due to acute respiratory distress syndrome (ARDS) [3]. While mortality appears to be more common in older individuals and those with comorbidities, such as chronic lung disease, cardiovascular disease, hypertension and diabetes, young people with no comorbidities also appear to be at risk for critical illness including multi-organ failure and death. There has been an expanding number of studies rapidly published online and in academic journals; however, some of these may be of limited quality and are pre-published without sufficient peer-review. Critical appraisal of the existing studies is needed to determine if the existing evidence is sufficient to support currently proposed management strategies. Given the rapid global spread of SARS-CoV-2 and the difficulty for the overburdened front-line providers and policymakers to stay up to date on emerging literature, IDSA has recognized the necessity of developing a rapid guideline for the treatment of COVID-19. The guideline panel is using a methodologically rigorous process for evaluating the best available evidence and providing treatment recommendations. Two additional guidelines on diagnostic testing and infection prevention also have been developed. These guidelines will be frequently updated as substantive literature becomes available and are accessible on an easy to navigate web and device interface at http: //www.idsociety.org/covid19guidelines. Version 4.2.0 10 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. There continue to be several ongoing trials evaluating therapeutic agents for the treatment of COVID-19. As data becomes available from these trials and if there is a preponderance of evidence to suggest the use of a therapeutic agent even in the context of clinical trials is no longer warranted it will be removed from future updates of the guideline (and the removal will be noted in the updated guidelines). If there is emerging evidence on the efficacy or safety of a therapeutic agent not mentioned in the current version of the guideline it will be included in future updates of the guideline. These recommendations are intended to inform patients, clinicians, and other health professionals by providing the latest available evidence. Methods This guideline was developed using the GRADE approach for evidence assessment. In addition, given the need for an urgent response to a major public health crisis, the methodological approach was modified according to the Guidelines International Network/McMaster checklist for the development of rapid recommendations [4]. Panel composition The initial guideline panel assembled in March 2020 was composed of nine members including infectious diseases specialists as well as experts in public health as well as other front- line clinicians, specializing in pharmacology, pediatrics, medical microbiology, preventive care, critical care, hepatology, nephrology and gastroenterology. Organizational representatives were included from the Society for Healthcare Epidemiology of America (SHEA), and the Pediatric Infectious Diseases Society (PIDS). In May 2020, an additional panel member was included as a representative from the Society of Infectious Diseases Pharmacists (SIDP). The Evidence Foundation provided technical support and guideline methodologists for the development of this guideline. Version 4.2.0 11 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Risk of bias and certainty of evidence Risk of bias was assessed using the Cochrane Risk of Bias Tool for RCTs and the Risk of Bias Instrument for Non-randomized Studies — of Interventions (ROBINS-I) [9, 10]. The certainty of evidence was assessed using the GRADE approach [11]. Within GRADE, the body of evidence across each outcome is assessed for domains that may reduce or increase one”s certainty in the evidence. Factors that may reduce one's certainty include risk of bias (study limitations), inconsistency (unexplained heterogeneity across study findings), indirectness (applicability or generalizability to the research question), imprecision (the confidence in the estimate of an effect to support a particular decision) or publication bias (selective publication of studies). One*s certainty in the evidence may be strengthened if the following considerations are present: large or very large magnitude of effect, evidence of a dose-response gradient, or opposing residual confounding. GRADE summary of findings tables were developed in GRADEpro Guideline Development Tool [12]. As higher quality direct evidence for clinical vutcomes becomes available, outcomes previously deemed critical by the panel became less important for decision-making. For example, at the time of the first guideline, clinical improvement outcomes (e.g., need for mechanical ventilation) were not reported, only the results of radiographic findings. However, with the recent publication of RCTs and non-randomized studies reporting on direct measures of clinical improvement, results of radiographic studies were deemed to be less critical for decision making. Evidence to recommendations The panel considered core elements of the GRADE evidence in the decision process, including Certainty of evidence and balance between desirable and undesirable effects. Additional domains were acknowledged where applicable (feasibility, resource use, acceptability). For all recommendations, the expert panelists reached consensus. Voting rules were agreed on prior to the panel meetings for situations when consensus could not be reached. Version 4.2.0 14 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. As per GRADE methodology, recommendations are labeled as “strong” or “conditional”. The words “we recommend” indicate strong recommendations and “we suggest” indicate conditional recommendations. Figure 1 provides the suggested interpretation of strong and weak recommendations for patients, clinicians, and healthcare policymakers. For recommendations where the comparators are not formally stated, the comparison of interest is implicitly referred to as “not using the intervention”. These recommendations acknowledge the current “knowledge gap” and aim at avoiding premature favorable recommendations for their use and to avoid encouraging the rapid diffusion of potentially ineffective or harmful interventions. Detailed suggestions about the specific research questions that should be addressed are found in the table (see Table s2). Figure 1. Approach and implications to rating the quality of evidence and strength of recommendations using the GRADE methodology (unrestricted use of the figure granted by the U.S. GRADE Network) 1 2. 3. Establish initial Consider loweringor raising Final level of 3 level of confidence level of confidence confidence rating 5 5 2 Study design Anitial ¡Reasons for considering lowering Confidence 5 confidence or raising confitence in an estimate of effect 2 in an estimate across those considerations E ofeffect Y Lower Anhigherif E z High Large effect High É Randomized trials P- cn 3 fer Doseresponse EE 2 AI plausible Moderate > confounding 8e bias seso so > woldraduce E A a ON a 3 2 or S + would suggest a. <pumion ofcer fo effect was observed 4 Population: Most people in this situation would want the recommended course of action and only a small proportion would not ++ Health care workers: Most people should receive the recommended course ofaction %* Pollcy makers: The recommendation can be adapted as a Policy in mos situstions Ealance batween benefits, harms 8cburdens Strong 4 Population: The majority of people in thissituation would want the recommended course of action, but many would not + Health care workers: Be orepared to help people to make a decision thatis consistert with their own values/decision aids and shared decision making 4 Policy makers: There is a need for substartial debate and involvement of stakeholders Recommendation 3. Implication of the Strength of Recommendation Resources cani cost 2. Determinants of she Strength of Conditional Version 4.2.0 15 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Review process This guideline has been rapidly reviewed and approved by the IDSA Board of Directors Executive Committee external to the guideline development panel. SHEA and PIDS have reviewed and provided endorsement of its contents. Updating process and terminology Regular, frequent screening of the literature will take place to determine the need for revisions based on the likelihood that any new data will have an impact on the recommendations. When necessary, the entire expert panel is reconvened to discuss potential changes. Changes to these guidelines will fall into one of two categories: update or amendment. An update involves a search for new studies, and if any new studies are found, they will be critically appraisal and the pertinent section will be removed and replaced with the updated section. An amendment involves a change or correction to the document, without any search for new studies and their appraisal. It will also involve changes made to clarify or explain a section based on “living” feedback from the readers. Guideline revisions may result in major, minor, or “patch” version changes, defined as follows: - — Major version (e.g., 1.0.0): Synonymous with a newly published version in the journal. This is usually called a "breaking version", ¡.e., prior recommendations may not be valid anymore. - — Minor version (e.g., 1.1.0): Includes new information, maybe even added PICOs, but not a breaking version, ¡.e., existing recommendations are still valid, although new recommendations may be available. - — Patch version (e.g., 1.0.1): Small changes, i.e., typos, adding words, removing words, but there are no material changes to the document or changes in recommendations. Version 4.2.0 16 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Hydroxychloroquine Five RCTs showed a trend toward mortality among patients with COVID-19 treated with HCQ compared to those who were not (relative risk [RR]: 1.08; 95% confidence interval [Cl]: 0.99, 1.19, Moderate certainty in the evidence) (Table 1) [28, 29, 33]. Hydroxychloroquine + Azithromycin One RCT could not exclude the risk of in-hospital mortality among patients treated with HCQ+AZ compared to those not receiving HCQ or HCQ+AZ (hazard ratio [HR]: 0.64; 95% Cl: 0.18, 2.21; Low certainty of evidence [CoE]) [28]. Three non-randomized studies failed to identify an association between treatment with HCQ+AZ and mortality: Ip reported an adjusted HR of 0.98 (95% Cl: 0.75, 1.28); Magagnoli reported an adjusted HR in a subset after propensity score adjustment of 0.89 (95% Cl: 0.45, 1.77); Rosenberg 2020 reported an adjusted HR of 1.35 (95% Cl: 0.79, 2.40) [37, 39, 41]. As stated in the HCQ section, one non-randomized study reported a reduction in mortality among patients receiving HCO+AZ (HR: 0.29; 95% Cl: 0.22, 0.40); however, it failed to adjust for the critical confounder of disease severity and imbalances in steroid use [27]. As described in the HCQ section, similar methodologic concerns exist among patients allocated to HCQ+AZ in the Arshad study, leading to several sources of bias in interpreting their favorable results. Harms Hydroxychloroquine One RCT reported that persons treated with HCQ experienced a longer time until hospital discharge (median 16 days compared with 13 days) and lower probability of being discharged alive within the 28-day study period (rate ratio: 0.92; 95% Cl: 0.85, 0.99) [29]. In addition, persons treated with HCQ who were not on mechanical ventilation at baseline were more likely to be placed on mechanical ventilation during follow up (rate ratio: 1.10; 95% Cl: 0.92, 1.31; Low CoE) [29, 32]. Across the body of evidence from four RCTs, treatment with HCQ. may increase the risk of experiencing adverse events (RR: 2.36; 95% Cl: 1.49, 3.75; Low CoE) and severe adverse events (adjusted odds ratio: 1.26; 95% Cl: 0.56, 2.84; Low CoE) [28, 30, 31, Version 4.2.0 19 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. 35]. One RCT and two non-randomized studies suggest increased risk of QT prolongation among patients treated with HCQ compared to those not receiving HCQ (RR: 8.47; 95% Cl: 1.14, 63.03; Low CoE and RR: 2.89; 95% Cl: 1.62, 5.16; Very low CoE, respectively) [28, 38, 39]. In addition, Rosenberg 2020 reported 16% of patients in the HCQ arm experienced arrhythmias compared with 10% in the non-HCQ arm (RR: 1.56; 95% Cl: 0.97, 2.50; Very low CoE). Gastrointestinal side effects occurred in 7% of patients in a prospective cohort study in 224 COVID-19 uninfected patients with systemic lupus erythematosus (SLE) who received either chloroquine or hydroxychloroquine for routine care [46]. While the 4-aminoquinolines, chloroquine and HCQ, have not been demonstrated to cause hemolysis in people with glucose-6-phosphate dehydrogenase (G6PD) deficiency [47, 48], case reports of hemolysis have emerged when these agents have been used for the treatment of COVID-19 [49-51]. It is possible that infection with SARS-CoV-2 may trigger hemolysis in G6PD deficient individuals in the absence of a 4-aminoquinolone. Caution should be exercised in administering these agents to G6PD deficient individuals with COVID-19, particularly if used for extended durations. Renal clearance accounts for 15-25% of total clearance of HCQ; however, dose adjustments are not recommended with kidney dysfunction. Chloroquine and HCQ are metabolized by cytochrome P450 isoenzymes 2C8, 2D6, and 3A4 [52]. Therefore, inhibitors and inducers of these enzymes may result in altered pharmacokinetics of these agents. Hydroxychloroquine + Azithromycin One RCT suggests increased risk of QT prolongation among patients treated with HCQ+AZ compared to those not receiving HCQ (RR: 8.50; 95% Cl: 1.16, 62.31; Low CoE) [28]. Two studies described significant QT prolongation in 10 of 95 patients treated with HCQ+AZ, illustrating the high risk for clinically relevant arrhythmias with this treatment [43, 45]. In addition, several case reports of QT prolongation related to HCQ have also been published [53- 56]. A case-control study of persons with COVID-19 treated with HCQ+AZ compared to healthy, untreated controls reported higher values of minimum (415 vs. 376 ms), mean (453 vs. 407 ms) Version 4.2.0 20 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. and maximum QTc-interval (533 vs. 452 ms) among COVID-19 cases (n=22) compared to controls (n=34) [42]. Additional case reports have cited the risk of a prolonged QT prolongation, torsades de pointes, and ventricular tachycardia in patients without COVID-19 receiving AZ alone. In a large cohort study, patients taking a five-day course of AZ had an increased risk of sudden cardiac death with a HR of 2.71 (1.58-4.64) vs. 0.85 (0.45-1.60), compared to patients receiving either no antibiotic or amoxicillin, respectively [57]. Given the cumulative effect on cardiac conduction seen with HCQ and AZ, if this combination was used, baseline and follow-up electrocardiogram (ECG) monitoring would be indicated, as well as careful surveillance for other concomitant medications known to prolong the QT interval. Azithromycin has a low risk for cytochrome P450 interactions [58]; however, additional pharmacologic adverse events including gastrointestinal effects and QT prolongation need to be carefully considered, particularly in the outpatient setting where frequent ECG monitoring is not feasible. Providers are encouraged to visit resources such as https://www.covid19- druginteractions.org/ to aid in the evaluation and management of drug interactions with current and emerging investigational agents for COVID-19. Other considerations The panel agreed that the overall certainty of evidence against treatment with HCQ was moderate due to concerns with imprecision around the risk for a trend towards harms from increased mortality. When considering the addition of AZ, the overall certainty of the evidence was low; however, the panel recognized even greater concern with the toxicity. In addition, based on the moderate certainty of increased QT prolongation, the panel determined that this demonstrated certain harm with uncertain benefit; therefore, the panel made a strong recommendation against HCQ+AZ. Conclusions and research needs for this recommendation Version 4.2.0 21 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Penes O ne te | n [espe li ET] % Other A Relative | Absolute AEREOS cra | Md IA 42789 | randomized [seriousi| notserious not serious serious * none 94/315 (29.8%) i 18/176 | RR236 | 139 more o 8600 IMPORTANT trials (10.299*| (1.49to | per1,000 Low 3.75) | (from50 more to 281 more) Severe AEs (assessed with: untoward medical event leading to death, a life-threatening experience, prolongation of hospitalization, or persistent or significant disability or incapacity) 11 randomized | not not serious not serious | very serious * none 14/242 (5.8%) 11/237 | 0R 1.26 [11 more per se O O CRITICAL trials serious (46%) | (0.56 to 1,000 284)! | (from20 Low fewer to 75 more) QT prolongation (RCTs) 12 randomized | not not serious notserious | very serious ? none 13/89 (14.6%) 1/58 RR 8.47 | 129 more se O O IMPORTANT trials serious (17%) | (1.14to | per1,000 63.03) |(from 2 more Low to 1,000 more) QT prolongation (NRS) 2810 | observational | very not serious not serious serious h none 46/355 (13.0%) 13/311 | RR 2.89 |79 more per o O OO IMPORTANT studies serious (42% | (1.62 to 1,000 VERY LOW gm 5.16) (from 26 more to 174 more) GRADE Working Group grades of evidence High certainty: We are very confident that the true effectlies close to that of the estimate of the effect Moderate certainty: 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 itis substantialy different Low certainty: Our confidence in the effect estimate is limited: The rue effect may be substantially diferent from the estimate ofthe effect Very low certainty: We have very lítle confidence in the effect estimate: The true effectis likely to be substantiall diferent from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplaed heterogeneity across study findings Indirectness: Applicabilty or generalizabiity to the research question Imprecision: The confidence in the estimate of an effectto support a particular decision Publication bias: Selective publication of studies Version 4.2.0 24 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Cl: Confidence interval, RR: Risk ratio; OR: Odds ratio Explanations TIE Co-interventions were provided to patients in both studies but balanced across arms. Cavalcanti 2020 excludes persons receiving supplemental oxygen at a rate of more than 4 liters per minute. The 95%C1 cannot exclude the potential for no benefit or harm. Cavalcanti was an open-label trial. The 95%C! includes the potential for both benefit and harm. Few events suggest the potential for fragility in the estimate. Few events suggest the potential for fragility in the estimate. Concems with unmeasured and residual confounding. Multiple co-interventions received across arms. Few events reported do not meet the optimal information size and suggest fragility in the estimate. Did not report on blinding (including outcome adjudication committee), sequence generation or allocation concealment; Chen J 2020: all patients received nebulized alpha- interferon, 80%ys. 67.7% of subjects received Abidiol in the hydroxychloroquine vs. placebo arm, respectively. Two subjects in the control arm received lopinavit/ritonavir. Chen J 2020: 4 AEs include diarrhea, fatigue and transient AST elevation. Chen Z 2020: 1 rash, 1 headache. Tang 2020: 21 AEs include disease progression (1%), URI (1%, diarrhea (10%), vomiting (3%. K. 3AEs reported in 2 patients include: AST elevation, creatinine elevation and anemia L — 20R: age, sex, baseline COVID Outcome Scale category, baseline Sequential Organ Failure Assessment score, and duration of acute respiratory infection symptoms prior to randomization m. Mahevas 2020 does not report on AEs in the comparator arm. References 1. 2 3, 4 RECOVERY Collaborative Group, Horby P, Mafham M, et al. Effect of Hydroxychloroquine in Hospitalized Patients with Covid-19. N Engl J Med 2020; 383(21): 2030-40. . Cavalcanti AB, Zampieri FG, Rosa RG, et al. Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19. N Engl J Med 2020; 383: 2041-52. WHO Solidarity Trial Consortium, Pan H, Peto R, et al. Repurposed Antiviral Drugs for Covid-19 - Interim WHO Solidarity Trial Results. N Engl J Med 2021; 384: 497-511. Self WH, Semler MW, Leither L, et al. Effect of hydroxychloroquine on clinical status at 14 days in hospitalized patients with COVID-19: A randomized clinical trial. JAMA 2020; 324(21): 2165-76. Ulrich Ry, Troxel AB, Carmody E, et al. Treating COVID-19 With Hydroxychloroquine (TEACH): A Multicenter, Double-Blind Randomized Controlled Trial in Hospitalized Patients Open Forum Infect Dis 2020; 7(10): ofaa446. Rosenberg ES, Dufort EM, Udo T, et al. Association of treatment with hydroxychloroquine or azithromycin with in-hospital mortality in patients with COVID-19 in New York state. JAMA 2020; 323(4): 2493:502. Chen J, Liu D, Liu L, et al. A pilot study of hydroxychloroquine in treatment of patients with moderate COVID-19. Journal of Zhejiang University (Medical Sciences) 2020; 49(2): 215- 9. Chen Z, Huy, Zhang, et al. Efficacy of hydroxychloroquine in patients with COVID-19: results of a randomized clinical trial. medRxiv 2020; Available at: https://doi.org/10.1101/2020.03.22.20040758 [Preprint 10 April 2020]. Tang W, Cao Z, Han M, et al. Hydroxychloroquine in patients with mainly mild to moderate coronavirus disease 2019: open label, randomised controlled trial. bmj 2020; 369: m1849. |. Mahevas M, Tran V-T, Roumier M, et al. No evidence of clinical efficacy of hydroxychloroquine in patients hospitalized for COVID-19 infection with oxygen requirement results of a study using routinely collected data to emulate a target trial. medRxiv 2020; Available at: https://doi ory/10.1101/2020.04.10.20060699 [Preprint 14 April 2020]. Version 4.2.0 25 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 2. GRADE evidence profile, Recommendation 2 Question: Hydroxychloroquine and azithromycin compared to no hydroxychloroquine/azithromycin for hospitalized patients with COVID-19 Last updated 8/20/2020; lastreviewed 12/23/2020 Ds EAS ME A Ne of (ela 07 E DEIEES LOA 1) OE VENAS 8 e A y considerations | and azithrom A USAN MUA Mortality (RCTs) (follow up: range 22 days to 49 days) 41 randomized | not not serious not serious? | very serious 64 none 5/172 (2.9% 6/173 (3.5%) HR0.64 | 12 fewer 00 O O CRITICAL trials serious (0.18 to | per 1,000 LOW a 2.21) | (from 28 fewer to 40 more) Mortality (NRS) 3234 | observational | very not serious not serious serious Y none Three non-randomized studies failed to identify an association between o O O O CRITICAL studies serious persons treated with HCQ +AZ and mortality: Ip reported an adjusted e HR of0.98 (95%CI: 0.75, 1.28); Magagnoli reported an adjusted HR in a] VERY LOW subset after propensity score adjustment of 0.89 (95%-CI: 0.45, 1.77); Rosenberg 2020 reported an adjusted hazard ratio (HR) of 1.35 (95% CI: 10.79, 2.40)(Ip, Magagnoli 2020, Rosenberg 2020). Clinical status (assessed with: 7-point scale, higher values represent worse clinical outcomes) 4 randomized |serious| notserious not serious? | serious de none 172 173 - MD 0.99 CRITICAL trials r higher ee00 (0.57 Low higher to 1.73 higher) Version 4.2.0 26 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Lopinavir/Ritonavir Section last reviewed and updated 11/22/2020 Last literature search conducted 11/18/2020 Recommendation 3: Among hospitalized patients with COVID-19, the IDSA guideline panel recommends against the use of the combination lopinavir/ritonavir. (Strong recommendation, Moderate certainty of evidence) Why is lopinavir plus ritonavir considered for treatment? Lopinavir/ritonavir is a protease inhibitor that was U.S. Food and Drug Administration (FDA)- approved for the treatment of HIV in September 2000. Ritonavir is added to the combination as a pharmacokinetic enhancer due to its strong inhibition of cytochrome P450 3A4, a metabolic pathway for lopinavir metabolism. Lopinavir/ritonavir demonstrated in vitro inhibition of SARS CoV-1 and MERS- CoV replication [59-61]. A trial of lopinavir/ritonavir and ribavirin vs historical controls in SARS CoV-1 patients, showed a reduced rate of ARDS and mortality in those receiving lopinavir/ritonavir. This study had limitations including a control group from early in the outbreak when management strategies likely differed significantly [62]. During the MERS outbreak, case reports cited efficacy of lopinavir/ritonavir with interferon in the management of MERS patients [63, 64]. During the early phase of COVID-19, triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin shortened the duration of viral shedding and hospital stay in patients with mild to moderate COVID-19 in an open-label, randomized, phase 1! trial [65]. Summary of the evidence Three RCTs reported on treatment with combination lopinavir/ritonavir or placebo for hospitalized patients with COVID-19 [32, 66, 67] (Table 3). The trials reported on the following outcomes: mortality, failure of clinical improvement (measured using a 7-point scale or hospital discharge), need for mechanical ventilation, and adverse events leading to treatment discontinuation. Version 4.2.0 29 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Benefits Among hospitalized patients with COVID-19, treatment with lopinavir/ritonavir failed to show or exclude a beneficial effect on mortality or need for invasive mechanical ventilation (RR: 1.00; 95% Cl: 0.89, 1.13; moderate certainty of evidence and RR: 1.12; 95% Cl: 0.93, 1.34; low CoE). Similarly, lopinavir/ritonavir may reduce failure of clinical improvement at 14 days, but it is uncertain (RR: 0.78; 95% Cl: 0.63, 0.97; very low CoE). Harms RECOVERY reported 1/1588 serious adverse event due to treatment with lopinavir-ritonavir [67]; however, nearly 14% of lopinavir/ritonavir recipients in Cao 2020 were unable to complete the full 14-day course of administration. This was due primarily to gastrointestinal adverse events, including anorexia, nausea, abdominal discomfort, or diarrhea, as well as two serious adverse events, both acute gastritis. Two recipients had self-limited skin eruptions. Such side effects, including the risks of hepatic injury, pancreatitis, more severe cutaneous eruptions, and QT prolongation, and the potential for multiple drug interactions due to CYP3A inhibition, are well documented with this drug combination. The side-effect profile observed in these trials raise concerns about the use of higher or more prolonged lopinavir—ritonavir dose regimens in efforts to improve outcomes. Other considerations The panel determined the certainty of evidence to be moderate due to concerns with imprecision. The guideline panel made a strong recommendation against treatment with the combination of lopinavir/ritonavir for hospitalized patients with COVID-19. Conclusions and research needs for this recommendation The guideline panel recommends against treatment with lopinavir/ritonavir in hospitalized patients with COVID-19. Version 4.2.0 30 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 3. GRADE evidence profile, Recommendation 3 Question: Lopinavir-ritonavir compared to no Lopinavir-ritonavir for hospitalized patients with severe COVID-19 Lastreviewed and updated 11/22/2020 Pes ene] (eva ET ñ 5 ñ LoS Ri COVES di A mpreci considerations | -%P' ch | (9: 1) Mortality (follow up: 28 days) 3123 lrandomized| not not serious | notserious | serious? none 538/3111 (17.3% * | 938/4896 | RR 1.00 | Ofewer 009 O CRITICAL trials — | serious * (19.2%) | (0.89to | Per 1,000 MODERATE 1.13) (from 21 fewer to 25 more) Invasive mechanical ventilation (follow up: 28 days) 213 [randomized|serious *4| notserious | notserious | serious? none 166/1655 (10.0%) | 297/3380 | RR 1.12 | 11 more 6900 CRITICAL trials (8.8%) | (0.93to |per1,000 LOW 1.34) (from 6 fewer to 30 more) AEs leading to treatment discontinuation 11 — [randomized| serious 2 | not serious not serious | very serious * none Nearly 14% of lopinavirritonavir recipients were unable to o 0O0O IMPORTANT trials complete the full 14-day course of administration. This was VERY LOW due primarily to gastrointestinal adverse events, including anorexia, nausea, abdominal discomfort, or diarrhea, as well as two serious adverse events, both acute gastritis. Two recipients had self-limited skin eruptions. Such side effects, including the risks of hepatic injury, pancreatitis, more severe cutaneous eruptions, and QT prolongation, and the potential for multiple drug interactions due to CYP3A inhibition, are well documented with this drug combination. The side-effect profile observed in the current trial arouses concern about the use of higher or more prolonged lopinavir—ritonavir dose regimens in efforts to improve outcomes. Failure of clinical improvement at 14 days (follow up: 14 days) 11 [randomized| serious? | — not serious not serious | very serious Y none 54/99 (54.5% 70/100 RR 0.78 [154 fewer 6000 CRITICAL trials (70.0%) | (0.62to |per1,000 VERY LOW 0.97) | (from 266 fewer to 21 fewer) Version 4.2.0 31 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. *Critical illness is defined as patients on mechanical ventilation and ECMO. Critical illness includes end organ dysfunction as is seen in sepsis/septic shock. In COVID-19, the most commonly reported form of end organ dysfunction is ARDS **Severe illness is defined as patients with SpO, <94% on room air, including patients on supplemental oxygen. ***Non-severe illness is defined as patient with a SpO» > 94% not requiring supplemental oxygen. The last literature search was conducted on September 4, 2020 and we identified eight RCTs and seven comparative non-randomized studies. Why are corticosteroids considered for treatment? In the early days of the SARS-CoV-2 pandemic, based on experience in both SARS and MERS, recommendations [68] cautioned against the use of systemic corticosteroids due to risk of worsening clinical status, delayed viral clearance, and adverse events [69-71]. Given the hyper-inflammatory state in COVID-19, immunomodulatory approaches, including steroids, continue to be evaluated to address both ARDS and systemic inflammation. ARDS stemming from dysregulated systemic inflammation may translate into prolonged ventilatory requirements and in-hospital mortality. In non-viral ARDS settings there is increasing support for the role of steroids in the management of ARD [72]. A recent multicenter RCT in patients with moderate to severe ARDS demonstrated a reduced number of ventilatory days and reduction in mortality with use of a 10-day regimen of dexamethasone [73]. Summary of the evidence Critical illness Our search identified one systematic review that analyzed eight RCTs reporting on treatment with glucocorticoids among 1,844 critically ¡ll patients with COVID-19 [74]. Three RCTs reported on patients treated with low- and high-dose dexamethasone [73, 75, 76]; three RCTs reported on patients treated with low-dose hydrocortisone [77-79]; and two RCTs reported on patients treated with high- Version 4.2.0 34 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. dose methylprednisolone [74, 30]. The definition of critically ¡ll varied across trials; however, the majority of patients had ARDS. Severe and non-severe illness Our search identified one RCT, one “partially” randomized trial, one prospective cohort, and five retrospective cohort studies [75, 81-87]. The RCT provided the best available evidence on treatment with corticosteroids for persons with COVID-19 [75] (Tables 4-6). Corral-Gudino et al. reported on a study that randomized patients to receive methylprednisolone or standard of care; however, patients expressing a preference for methylprednisolone were assigned to the same treatment arm [81]. Corral-Gudino et al. did not report the disaggregated results from the randomized trial; therefore, succumbing to the same potential for bias as reported subsequently for the non- randomized studies. The non-randomized studies had significant limitations with controlling for multiple co-interventions and disease severity at baseline [82-87]. All non-randomized studies had concerns with risk of bias due to lack of adjustment for critical confounders or potential for residual confounding. Timing of receipt, dose and duration of corticosteroids varied across studies. The RECOVERY trial is a randomized trial among hospitalized patients in the United Kingdom [75]. In that study, 2,104 participants were randomized to receive dexamethasone (6 mg daily for up to 10 days) and 4,321 were randomized to usual care. The RECOVERY trial reported on the outcomes of mortality and hospital discharge. Participants and study staff were not blinded to the treatment arms. Benefits Critical illness Among hospitalized, critically ill patients, the odds of mortality at 28 days was 34% less among patients treated with glucocorticoids than among patients not treated with glucocorticoids (OR: 0.66; 95% Cl: 0.54; 0.82; High CoE). In addition, at 28 days, patients receiving dexamethasone were more likely to be discharged from the hospital (RR: 1.11; 95% Cl: 1.04, 1.19; Moderate CoE). Severe illness Version 4.2.0 35 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Among hospitalized patients, 28-day mortality was 17% lower in the group that received dexamethasone than in the group that did not receive dexamethasone (RR 0.83; 0.74-0.92; Moderate CoE). In addition, at 28 days, patients receiving dexamethasone were more likely to be discharged from the hospital (RR: 1.11; 95% Cl: 1.04, 1.19; Moderate CoE). Non-severe illness In a sub-group analyses of patients without hypoxia not receiving supplemental oxygen, there was no evidence for benefit and a trend toward harm with dexamethasone in participants who were not on supplemental oxygen (RR 1.22; 0.86, 1.75; Low CoE). Harms A systematic review of six studies did not report a difference in the events of serious adverse events experienced by patients randomized to receive treatment with glucocorticoids or no treatment with glucocorticoids (64/354 among those receiving glucocorticoids vs. 80/342 among those not receiving glucocorticoids). Patients receiving a short course of steroids may experience hyperglycemia, neurological side effects (e.g., agitation/confusion), adrenal suppression, and risk of bacterial and fungal infection [82, 88, 89]. Other considerations Critical illness The panel agreed that the overall certainty of the evidence for treatment with glucocorticoids for patients with critical COVID-19 was moderate due to concerns with indirectness and imprecision. Severe illness The panel agreed the overall certainty of evidence for treatment with glucocorticoids for patients with severe COVID-19 as moderate due to concerns with indirectness since the evidence was from dexamethasone. Non-severe illness Version 4.2.0 36 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 5. GRADE evidence profile, Recommendation 5 Question: Glucocorticoids compared to no glucocorticoids for hospitalized patients with severe but not critical COVID-19 Lastreviewed and updated 9/25/2020 IE O Ped LETS TEO Cn LE d y P glucocorto SJMEVA] cl) Mortality (follow up: 28 days) 11 [randomized| — not not serious serious > not serious none 454/2104 (21.6%)| 1065/4321 RR0.83 | 42fewer 009 O CRITICAL trials serious 2 (24.6%) (0.74to | per 1,000 MODERATE 0.92) | (from64 fewer to 20 fewer) Hospital discharge (follow up: 28 days) 11 [randomized| not not serious serious? | notserious none 1360/2104 2639/4321 RR1.11 | 67 more 469900 IMPORTANT trials — | serious? (64.6%) (61.1%) (1.04to | per1,000 MODERATE 1.19) | (from24 more to 116 more) Adverse events Patients receiving a short course of steroids may experience - CRITICAL hyperglycemia, neurological side effects (e.g., agitation/confusion), adrenal suppression, and risk of infection (Salton 2020; Henzen 2000; Siemieniuk 2015). GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effectis likely to be close to the estimate of the effect, but there is a possibility thatit is substantally ifferent Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantlly different from the estimate of the effect Very low certainty: We have very litle confidence in the effect estimate: The true effect is likely to be substantilly different from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplaied heterogeneity across study findings Indirectness: Applicabilty or generalizability to the research question Imprecision: The confidence in the estimate of an effect to support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval, RR: Risk ratio Explanations a. Analysis adjusted for baseline age. b. Indirectness due to different health care system (allocation of intensive care resources in an unblinded study). Indirectness to other corticosteroids. Reference 1. RECOVERY Collaborative Group, Horby P, Lim WS, et al. Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med 2021; 384: 693-704. Version 4.2.0 39 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 6. GRADE evidence profile, Recommendation 6 Question: Glucocorticoids compared to no glucocorticoids for hospitalized patients with COVID-19 not receiving supplemental oxygen Lastreviewed and updated 9/25/2020 PT es es] (eva ET 1) E IVA CEBA Ea | LES Mortality (follow up: 28 days) 11 [randomized| serious? | notserious not serious serious > none 85/501 (17.0%) |137/1034 (13.29| RR1.22 | 29more 6900 CRITICAL trials (0.93to | per 1,000 LOW 1.61) (from 9 fewer to 81 more) Hospital discharge (follow up: 28 days) 11 — [randomized| serious 2 | — notserious not serious serious none 366/501 (73.1%) [791/1034 (76.5%| RR0.99 | 8 fewer ese00 IMPORTANT trials (0.87 to | per 1,000 1.12) | (from99 Low fewer to 92 more) Adverse events Patients receiving a short course of steroids may experience: - CRITICAL hyperglycemia, neurological side effects (e.g., agitation/confusion), adrenal suppression, and risk of infection (Salton 2020; Henzen 2000; Siemieniuk 2015). GRADE Working Group grades of evidence High certainty: We are very confdent that the true effect lies close to that of Ihe estimate of the efect Moderate certainty: 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 thatitis substantally different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially diferent from the estimate of the effect Very low certainty: We have very litle corfidence in the effect estimate: The true effect is ikely to be substantially different from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplained heterogeneity across study findings Indirectness: Applicabilty or generalizabiity to the research question Imprecision: The confidence in the estimate of an effect to support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval, RR: Risk ratio Explanations a. RoB due to post-hoc subgroup effect among persons not receiving supplemental oxygen. b. The 95%C1 includes the potential for appreciable harm and cannot exclude the potential for benefit. Few events reported do not meet the optimal information size and suggest fragility in the estimate. c. The 95%CI cannot exclude the potential for either appreciable harm or benefit Version 4.2.0 40 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Reference 1. RECOVERY Collaborative Group, Horby P, Lim WS, et al. Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med 2021; 384: 693-704. Version 4.2.0 41 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Both RECOVERY and REMAP CAP (the two tocilizumab trials that reported a benefit) initiated treatment early (randomization at median of two days of hospitalization in RECOVERY; <24 hours in the ICU for REMAP-CAP), suggesting tocilizumab may be more beneficial in people with early rapidly progressive disease. Benefits Among hospitalized patients, tocilizumab showed a trend toward reduced mortality at 28 days compared to no tocilizumab treatment (RR: 0.91; 95% Cl: 0.79, 1.04); moderate CoE). Tocilizumab demonstrated a lower relative risk of clinical deterioration, defined as death, need for mechanical ventilation, ECMO, or ICU admission, compared to placebo/usual care, RR: 0.83 (0.77, 0.89; moderate CoE). Four studies were not blinded, while healthcare personnel and outcome assessors were blinded in the remaining three trials. The panel noted that tocilizumab causes reduction in CRP levels, which would reveal the treatment arm designations of the patients, therefore introducing bias for the more subjectively measured outcomes of clinical deterioration and serious adverse events. Harms Serious adverse events among patients receiving tocilizumab did not differ from those not receiving tocilizumab (RR: 0.89; 95% Cl: 0.74, 1.07; low CoE). An additional trial attributed treatment with tocilizumab to three serious adverse events; however, did not report events among patients not receiving tocilizumab [99]. Previously, tocilizumab has been associated with gastrointestinal perforations in non-COVID-19 settings, and case reports of bowel perforations have recently emerged with the use of tocilizumab for COVID-19 [105-108]. Increased infection risks have been noted in uncontrolled studies, and it is possible that this risk may be compounded by the combination of glucocorticoids and tocilizumab. [109, 110]. Other considerations While the overall certainty of evidence for the trend toward a reduction in mortality was moderate, the panel believes that differences in mortality rates across the trials may be the result of the differences in baseline severity of study participants and timing of tocilizumab receipt in the Version 4.2.0 44 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. disease course. In REMAP-CAP, tocilizumab was administered within 24 hours of participants” initiating organ support in an intensive care unit, raising the possibility that this may be the optimal time to administer the drug. In RECOVERY, tocilizumab was administered to participants with oxygen saturation <92% on room air or receiving oxygen therapy, and CRP >75 mg/L. Given the reduction in clinical deterioration and trend toward mortality reduction, the guideline panel made a conditional recommendation for treatment of adults with tocilizumab. The use of tocilizumab, as with other therapeutic agents that can suppress the immune system, presents additional considerations and potential concerns when used in immunocompromised hosts. The panel did not conduct an analysis of available data to assess differences in efficacy and/or adverse effects of tocilizumab among oncology or other immunocompromised patients at this time. Conclusions and research needs for this recommendation The guideline panel suggests tocilizumab for hospitalized adults with COVID-19. Additional research is needed to understand the efficacy of tocilizumab when taken at different times during the course of disease. For example, there are no data to guide recommendations in patient <18 years old at this time. In addition, future studies are needed to inform the generalizability of tocilizumab with different IL-6 receptor inhibitors for patients with COVID-19 (Table s2). At the time of update, preliminary data from a trial of treatment with sarilumab has been shared as a pre-print [97]; however, number of patients who received sarilumab is limited (n=45) and the published manuscript was not available for analysis or inclusion to inform this recommendation. Other studies of sarilumab have not been made available. Version 4.2.0 45 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 7. GRADE evidence profile, Recommendation 7 Question: Tocilizumab compared to no tocilizumab for hospitalized patients with COVID-19 Lastreviewed and updated 2/17/2021 es ñ A ñ no Relative | Absolute Cuil OSO Ad tocilizamab | (95% cp | (95% cp Mortality (follow up: range 28 days to 30 days) 8 [randomized| not serious | notserious not serious serious > none 810/3280 | 893/3054 RR0.91 | 26fewer a) CRITICAL 1305078| — hials a (47% | (292% | (0.79to |per1,000 1.04) — | (fromó1 | MODERATE fewer to 12 more) Clinical deterioration (follow up: range 14 days to 30 days) 7 — [randomized| serious“ | notserious | notserious Y | notserious none 799/2712 | 939/2503 RR0.83 | 64fewer a) CRITICAL 1234508 | — trials (295% | (37.5% | (0.77to |per1,000 oso) | (fromss | MODERATE fewer to 41 fewer) Serious Adverse Events 7 — [randomized| serious“ | notserious not serious serious f none 210/1249 141/946 RR0.89 | 16fewer 6900 CRITICAL 1234567e| — trials (68% | (149% | (0.74to |per1,000 1.07) — | (from39 Low fewer to 10 more) GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate ofthe effect, but there is a possibility that itis substantially different Low certainty: Our confidence in the effect estimate ¡s limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplained heterogeneity across study findings Indirectness: Applicabilty or generalizability to the research question Imprecision: The confidence in the estimate of an effect to support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval, RR: Risk ratio Explanations Version 4.2.0 46 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. probability of obtaining a neutralizing antibody titer of >1:160 is highest (80% or greater) when the ELISA IgG titer is 21:1,350 [118]. In an analysis of the convalescent plasma expanded access program, higher levels of antibodies were associated with significant improvements in mortality compared to those receiving convalescent plasma with lower concentrations of neutralizing antibodies [116]. Regarding timing of treatment: Based on historical experience and emerging data, efficacy appears best when convalescent plasma is given at earlier stages of the disease and particularly prior to when patients become critically ¡ll [119, 120]. The analysis of the convalescent plasma expanded access program suggests the most benefit is seen when convalescent plasma is given in the first three days from diagnosis [116]. In August 2020, the FDA issued an emergency use authorization (EUA) for investigational convalescent plasma for the treatment of COVID-19 in hospitalized patients [121]. In early February 2021, the FDA issued a revision to the EUA to limit the authorization to the use of high- titer COVID-19 convalescent plasma for the treatment of hospitalized patients early in the disease course [122]. Summary of the evidence Our search identified and was informed by evidence from eleven RCTs and a large (n=20,000), single-arm registry study [111-115, 117-120, 123-125], as they provided the best available evidence for the outcomes of mortality, need for mechanical ventilation, serious adverse events, and adverse events. Ten of those RCTs reported on convalescent plasma transfusions for patients hospitalized with COVID-19 (Table 8) [111-114, 117-120, 123, 124] and one RCT reported on receipt of convalescent plasma by ambulatory persons with mild COVID-19 disease (Table 9) [115]. Ten trials randomized 13,026 patients hospitalized with COVID-19 to receive a transfusion with COVID-19 convalescent plasma [111-114, 117-120, 123, 124]. Several trials were open-label and/or had concerns with risk of bias due to lack of adjustment for critical confounders or potential for residual confounding (Table s12a). Timing of receipt of COVID-19 convalescent plasma during the clinical course of the patients” ¡llness varied across studies (Table s11). One trial reported on 160 persons who received high-titer convalescent plasma less than 72 hours after the onset of symptoms of COVID-19 (mean age: 77.2 years; standard deviation: +8.6 years) [115]. In addition, Joyner 2020 reported on Version 4.2.0 49 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. safety outcomes of over 20,000 patients enrolled in the same FDA Expanded Access Program for COVID-19 convalescent plasma study. Benefits Hospitalized patients Convalescent plasma transfusion failed to show or to exclude a beneficial or detrimental effect on mortality based on the body of evidence from RCTs (RR: 0.86; 95% Cl: 0.69, 1.06; moderate CoE). Receipt of COVID-19 convalescent plasma may not reduce the need for mechanical ventilation (RR: 1.11; 95% Cl: 0.95, 1.30; low CoE); however, the evidence is uncertain because of concerns with risk of bias and fragility of the estimate due to small number of events reported. Ambulatory persons Receipt of COVID-19 convalescent plasma may reduce progression to severe respiratory disease (RR: 0.52; 95% Cl: 0.29, 0.94; low CoE); however, the evidence is uncertain, as oxygenation and respiration rates are surrogate measures of need for ventilation, morbidity, and death, and because of the fragility of the estimate due to small number of events reported. Convalescent plasma transfusion may reduce mortality and clinical deterioration based on the body of evidence from an RCT (RR: 0.50; 95% Cl: 0.09, 2.65; low COoE and RR: 0.58; 95% Cl: 0.24, 1.40; very low CoE, respectively); however, the evidence is uncertain due to concerns with fragility of the estimate due to small number of events reported and the wide confidence interval failing to exclude a beneficial or detrimental effect. Harms In the largest safety study (n=20,000), within four hours of completion of convalescent plasma transfusion authors reported 146 serious adverse events classified as transfusion reactions (<1% of all transfusions) [125]. Of these, 63 deaths were reported (0.3%), 13 judged as possibly or probably related to the transfusion. The non-mortality serious adverse events include 37 reports of transfusion- associated circulatory overload, 20 cases of transfusion-related acute lung injury, and 26 cases of severe allergic transfusion reactions. Version 4.2.0 50 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Within seven days of transfusion, 1,711 deaths were reported (mortality rate: 8.56%; 95% Cl: 8.18, 8.95). In addition, 1,136 serious adverse events were reported: 643 cardiac events (569 judged as unrelated to the transfusion), 406 sustained hypotensive events requiring intravenous pressor support, and 87 thromboembolic or thrombotic events (55 judged as unrelated to the transfusion). Four trials among patients hospitalized for COVID-19 could not exclude an increase in mild-to- severe adverse events among patients receiving convalescent plasma (RR: 1.02; 95% Cl: 0.64, 1.62; low CoE) [113, 114, 118, 120]; however, the evidence was uncertain due to concerns with lack of blinding. In addition, included studies lacked a standard definition for what met the definition of an adverse event. One trial conducted among ambulatory persons receiving early, high-titer convalescent plasma did not report any serious adverse events [115]. Other considerations Hospitalized patients The panel agreed that the overall certainty of evidence is low due to concerns with risk of bias and imprecision. The guideline panel recognized the inability to exclude a meaningful beneficial or detrimental effect of convalescent plasma transfusion on mortality from the existing large body of evidence. Ambulatory persons The panel agreed that the overall certainty of evidence is low due to concerns with risk of bias and imprecision, which recognized the limited events and concerns with fragility. The guideline panel recognized the uncertainty of potential benefit when high titer convalescent plasma is given early in the course of COVID-19 disease. Conclusions and research needs for this recommendation The guideline panel suggests against COVID-19 convalescent plasma for persons hospitalized with COVID-19. The guideline panel recommends COVID-19 convalescent plasma for ambulatory persons only in the context of a clinical trial. Additional clinical trials are needed to determine whether there is a benefit of treatment with COVID-19 convalescent plasma and at what dose (neutralizing Version 4.2.0 51 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Study [ 5 ñ isi (0401 ES + A A CE) CST) nconsistency mprecisión | e onsiderations MESE] Ca Ed (95% Cl) plasma observational extremely] — not serious notserious | notserious none SAEs from 20,000 transfused patients: Within 7 days of | SKOOO CRITICAL studies serious transfusion, 1,711 deaths (8.56%) and 1,136 serious VERY LOW adverse events (5.68%) were reported. Non-mortality SAEs included: 643 cardiac events (569 judged as unrelated to the transfusion); 406 sustained hypotensive events requiring intravenous pressor support and 87 thromboembolic or thrombotic events (55 judged as unrelated to the transfusion). Any adverse events (RCTs) 43458 randomized | serious ¿| notserious | notserious 9 | serious h none 102/433 — |52/240(21.7%)| RR1.02 | 4more 460900 IMPORTANT trials (23.6% (0.64to | per1,000| — LOW 1.62) | (from78 fewer to 134 more) GRADE Working Group grades of evidence High certainty: We are very confident that the true effectlies close to that of the estimate of the effect Moderate certainty: 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 itis substantialy different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially diferent from the estimate ofthe effect Very low certainty: We have very lítle confidence in the effect estimate: The true effectis likely to be substantiall diferent from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplaied heterogeneity across study findings Indirectness: Applicabilty or generalizabiity to the research question Imprecision: The confidence in the estimate of an effectto support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval; RR: Risk ratio; HR: Hazard Ratio; OR: Odds ratio Explanations a. Li2020 time between symptom onset and randomization was over 14 days for >90% (median 30 days), no adjustment for co-interventions, allocation concealment methods not reported and participants and healthcare professionals not blinded. Many trials had concerns due to open-label trial, allocation concealment not reported, and no adjustments for co-interventions. The 95%C1 includes the potential for appreciable benefit; however, cannot exclude the potential for no effect. Concerns include open-label trial design and assessment of outcome. The 95% Cl may not include a clinically meaningful reduction in need for mechanical ventilation. No comparative effects available. Some subjectivity in classification of outcomes as transfusion related. Lack standard definition for adverse events. Studies report on mild to severe events. erpasp Version 4.2.0 54 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. h. The 95%C1 includes the potential for both increased harms, as well as no increased harms. Few events suggests fragility of the estimate. References 1. LiL, ZhangW, Hu Y, et al. Effect of Convalescent Plasma Therapy on Time to Clinical Improvement in Patients With Severe and Life-threatening COVID-19: A Randomized Clinical Trial. JAMA 2020; 324(5): 460-70. 2. Gharbharan A, Jordans CC, GeurtsvanKessel C, et al. Convalescent Plasma for COVID-19. A randomized clinical trial. medRxiv 2020: Available at: https://doi.org/10.1101/2020.07.01.20139857 [Preprint 3 July 2020]. 3. AlQahtani M, Abdulrahman A, AlMadani A, et al. Randomized controlled trial of convalescent plasma therapy against standard therapy in patients with severe COVID-19 disease. medRxiv 2020: Available at https://doi. ory/10.1101/2020.11.02.20224303 [Preprint 4 November 2020]. 4. Avendaño-Solá C, Ramos-Martinez A, Muñez-Rubio E, et al. Convalescent plasma for COVID-19: a multicenter, randomized clinical trial. medRxiv 2020: Available at: https://doi.org/10.1101/2020.08.26.20182444 [Preprint 29 September 2020]. 5. Ray Y, Paul SR, Bandopadhyay P, et al. Clinical and immunological benefits of convalescent plasma therapy in severe COVID-19: insights from a single center open label randomised control trial. medRxiv 2020: Available at: https://doi.org/10.1101/2020.11.25.20237883 [Preprint 29 November 2020]. 6. Simonovich VA, Burgos Pratx LD, Scibona P, et al. A Randomized Trial of Convalescent Plasma in Covid-19 Severe Pneumonia. N Engl J Med 2021; 384(7): 619-29. 7. Agarwal A, Mukherjee A, Kumar G, et al. Convalescent plasma in the management of moderate covid-19 in adults in India: open label phase Il multicentre randomised controlled trial (PLACID Trial). BMJ 2020; 371: m4232. 8. O'Donnell MR, Grinsztejn B, Cummings MJ, et al. Arandomized, double-blind, controlled trial of convalescent plasma in adults with severe COVID-19. medRxiv 2021: Available at: https://doi.org/10.1101/2021.03.12.21253373 [Preprint 13 March 2021]. 9. Horby PW, Estcourt L, Peto L, et al. Convalescent plasma in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial. medRxiv 2021: Available at: https//doi. org/10.1101/2021.03.09.21252736 [Preprint 10 March 2021]. Balcells ME, Rojas L, Le Corre N, et al. Early versus deferred anti-SARS-CoV-2 convalescent plasma in patients admitted for COVID-19: A randomized phase ll clinical trial. PLoS Med 2021; 18(3): e1003415. 11. Joyner MJ, Bruno KA, Klassen SA, et al. Safety Update: COVID-19 Convalescent Plasma in 20,000 Hospitalized Patients. Mayo Clin Proc 2020; 95(9): 1888-97. 1 e Version 4.2.0 55 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 9. GRADE evidence profile, Recommendation 9 Question: Early convalescent plasma with high titers compared to no early convalescent plasma for persons not hospitalized for COVID-19 compared to no convalescent plasma for hospitalized patients with COVID-19 New evidence profile developed 4/7/2021 EA p CEA A (ela UT n em Other convalescent Relative | Absolute El design El IS pao prin considerations EEE with AO (95% Cl) | (95% CI) da Mortality (follow up: 25 days) 11 |randomized| not not serious not serious | very serious ? none 2/80 (2.5% 4/80 (5.0%) | RR0.50 | 25 fewer e6900 CRITICAL trials — | serious (0.09to | per 1,000 LOW a 2.65) | (from 46 fewer to 83 more) Progression to severe respiratory disease (follow up: 15 days; assessed with: defined as a respiratory rate of 2 30 breaths per minute, Sa02 < 93% on RA, or both) 11 |randomized| not not serious serious serious none 13/80 (16.3%) | 25/80 (31.3%) | RR 0.52 | 150 fewer e6900 CRITICAL trials — | serious (0.29to | per 1,000 LOW a 0.94) | (from 222 fewer to 19 fewer) Clinical deterioration (follow up: 25 days; assessed with: life-threatening respiratory disease, critical systemi Iness, or death, alone or in combination) 11 [randomized| not not serious serious * — | very serious > none 7/80 (8.8% | 12/80 (15.0%) | RR 0.58 | 63 fewer 000 CRITICAL trials — [serious (0.24to | per 1,000 VERY LOW a 1.40) | (from 114 fewer to 60 more) Serious Adverse Events (follow up: 25 days) 11 |[randomized| not not serious not serious | very serious none 0/79 (0.0% 0/80 (0.0%) not e6900 CRITICAL trials [serious estimable LOW GRADE Working Group grades of evidence High certainty: We are very confidentthat the true effectlies close to that ofthe estimate of the effect Moderate certainty: We are moderately corfident 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 substantialy different Version 4.2.0 56 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Summary of the evidence Hospitalized patients with oxygen saturation >94% without supplemental oxygen Three RCTs compared treatment with five days of remdesivir (200 mg day one, 100 mg daily days 2-5), 10 days of remdesivir (200 mg day one, 100 mg daily days 2-10), or no remdesivir for patients hospitalized with oxygen saturation >94% on room air [32, 130, 131] (Table 10). The outcomes assessed were mortality, clinical improvement, and serious adverse events. ACTT-1 and SOLIDARITY provided post-hoc analyses among patients with mild-to-moderate disease [32, 130]. Randomization and lack of blinding failed to control for or balance receipt of co-interventions (e.g., treatment with dexamethasone, tocilizumab, hydroxychloroquine, and lopinavir/ritonavir) equally across arms in Spinner et al (2020) [131]. In addition, the Spinner et al did not adjust for severity of disease. Hospitalized patients with SpO, <94% on room air, including patients on supplemental oxygen, on mechanical ventilation, and ECMO Three RCTs comparing treatment with remdesivir (200 mg day one, 100 mg daily days 2-10) against no remdesivir treatment [32, 130, 132], and one RCT comparing five days of treatment (200 mg day one, 100 mg daily days 2-5) against 10 days (200 mg day one, 100 mg daily days 2-10) of treatment [133] served as the best available evidence among hospitalized persons with severe COVID-19 (Tables 10-12). The outcomes assessed were mortality, time to clinical improvement, need for mechanical ventilation, serious adverse events, and adverse events leading to treatment discontinuation. All trials used different definitions of severe disease for participants. Adaptive Covid-19 Treatment Trial (ACTT-1) participants were considered to have severe disease if they required mechanical ventilation, supplemental oxygen, ¡if SpOz was 94% or lower while breathing ambient air, or if they had tachypnea (respiratory rate >24 breaths per minute) [130]. Within the SOLIDARITY trial (available only as a pre-print at this time), participants with severe disease were receiving mechanical ventilation [32]. In Wang 2020, severe participants had a SpO, <94% while breathing room air or a ratio of arterial oxygen partial pressure to fractional inspired Oz of <300 mm Hg and radiologically confirmed pneumonia. Version 4.2.0 59 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Updated analyses include the final analysis from the ACTT-1 and the interim analysis of the SOLIDARITY trial [32, 130]. SOLIDARITY reported mortality among persons remaining in hospital up to the duration of the study; however, among patients discharged before the end of the study, mortality may not have been collected completely. The study by Wang et al (2020) was stopped early due to lack of recruitment into the trial due to decreased incidence in China. Randomization performed in Goldman 2020 failed to establish prognostic balance between baseline clinical status among the 397 patients randomized into the treatment arms, with patients in the 10-day arm more severely ill at study entry. Even with the adjusted analysis, residual confounding is possible. In addition, participants, healthcare workers, and outcome assessors were not blinded to the treatment arms. Benefits Hospitalized patients with oxygen saturation >94% without supplemental oxygen Treatment with a five- or ten-day course of remdesivir failed to show or to exclude a reduction in mortality when compared with no remdesivir (RR: 0.69; 95% Cl: 0.36, 1.34; Very low COE). A five-day course of remdesivir may increase clinical improvement over no remdesivir (RR: 1.16; 95% Cl: 1.00, 1.34; Very low CoE) but a 10-day course of remdesivir was not associated with improved clinical status as compared with no remdesivir. Patients with mild-to-moderate disease receiving treatment with remdesivir had similar median time to recovery (median 5 vs. 5 days; Rate ratio: 1.22; 95% Cl: 0.82, 1.81; Very low CoE). Hospitalized patients with SpO> <94% on room air, including patients on supplemental oxygen, on mechanical ventilation and ECMO The pooled analysis failed to show a mortality benefit at 28 days (RR: 0.92; 95% Cl: 0.77, 1.10; Low CoE) [32, 130, 132]. Patients receiving treatment with remdesivir trend toward greater clinical improvement at 28 days than patients not receiving remdesivir (RR: 1.13; 95% Cl: 0.91, 1.41; Low CoE) [132]. In addition, based on a post-hoc analysis of patients with severe COVID-19 receiving treatment with remdesivir had a shorter median time to recovery (median 11 vs 18 days; Rate ratio: 1.31; 95% Cl: Version 4.2.0 60 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. 1.12, 1.52; Low CoE) and decreased need for mechanical ventilation (RR: 0.57; 95% Cl: 0.42, 0.79; Moderate CoE) [130]. In the study by Goldman et al that compared five and ten days of treatment, the shorter course of remdesivir showed a trend toward decreased mortality (RR: 0.75; 95% Cl: 0.51, 1.12; Low CoE) and increased clinical improvement at 14 days (RR: 1.19; 95% Cl: 1.01, 1.40; Low CoE); however, the evidence is uncertain because the persons in the 10-day group had more severe disease at baseline and there is the possibility of residual confounding despite the adjusted analysis [133]. Harms Hospitalized patients with oxygen saturation >94% without supplemental oxygen Patients treated with five days of remdesivir do not appear to experience greater serious adverse events than those not receiving remdesivir (RR: 0.64; 95% Cl: 0.31, 1.31; Very low CoE). Hospitalized patients with SpO, <94% on room air, including patients on supplemental oxygen, on mechanical ventilation and ECMO Patients treated with remdesivir do not appear to experience greater SAEs (grade 3/4) than those not receiving remdesivir (RR: 0.87; 95% Cl: 0.59, 1.28; Moderate CoE) [130, 132]. Patients receiving five days of remdesivir may experience fewer SAEs and AEs leading to treatment discontinuation than patients receiving 10 days of remdesivir (RR: 0.61; 0.44, 0.85; Low CoE and RR: 0.44; 95% Cl: 0.21, 0.95; Low CoE, respectively); however, this evidence is uncertain because of the increased severity of disease among patients in the 10-day arm [133]. Other considerations Hospitalized patients with oxygen saturation >94% without supplemental oxygen The panel agreed that the overall certainty of the evidence for treatment of patients with an oxygen saturation >94% with remdesivir compared to no remdesivir was very low due to concerns with study limitations and imprecision. Because of the study limitations and the relatively small effect of remdesivir in patients with moderate COVID-19, the panel suggests remdesivir not be used routinely in Version 4.2.0 61 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 10. GRADE evidence profile, Recommendation 10 Question: Remdesivir compared to no antiviral treatment for hospitalized patients with COVID-19 and oxygen saturation >94% without supplemental oxygen Lastreviewed and updated 4/5/2021 (e assessment Ev Study Mortality (follow up: range 11 days to 29 days) 3122 randomized] very notserious | notserious | serious * none 15/1100 20/914 RR0.69 | 7fewer 000 CRITICAL trials — [serious 2: (1.4%) (2.2%) — (0.36 to 1.34)| per1,000| VERY LOW (from 14 fewer to 7 more) Time to recovery (follow up: 29 days) 12 [randomized] serious“ | notserious not serious [very serious none 54/55 46/50 Rate ratio | 34more 6000 CRITICAL trials (98.299 | (92.0% 122 |per1,000| VERYLOW (0.82 to 1.81)| (from 46 fewer to 70 more) Clinical improvement at day 11 (assessed with: >=2-pt improvement on 7-pt scale; higher = better) 11 [randomized| very notserious | notserious | serious * none 134/191 121/200 RR1.:16 | 97more | S6ODO CRITICAL tials — | serious 20 (10.299 | (60.5% |(1.00to 1.34)| per 1,000 | VERYLOW r (from 0 fewer to 206 more) Serious adverse events 212 [randomized] very notserious | notserious | serious none 11/246 18/249 RR0.64 |26fewer | 6000 CRITICAL trials — [serious abc (45% (1.2% — |(0.31 to 1.31)| per1,000 | VERYLOW (from 50 fewer to 22 more) GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate ofthe effect, but there is a possibility that itis substantially different Low certainty: Our confidence in the effect estimate ¡s limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Version 4.2.0 64 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Risk of bias: Study limitations Inconsistency: Unexplained heterogeneity across study findings Indirectness: Applicability or generalizability to the research question Impre Publication bias: Selective publication of studies : The confidence in the estimate of an effect to support a particular decision Cl: Confidence interval, RR: Risk ratio Explanations a. Spinner et al. co-treatments were not balanced between arms: 45%of patients randomized to control arm received HCQ or CQ compared to 11% in 10-day arm or 8%in 5- day arm; lopinavir/ritonavir was 22% in control arm, 6% in 10-day arm, and 5% in 5-day arm. b. Open-label trial design may have led to different clinical practices (co-interventions and time of hospital discharge). c. Post-hoc analysis of patients with mild-moderate disease from ACTT-1 (Beigel 2020) and SOLIDARITY (Pan 2020) may introduce bias. d. The 95%CI includes the potential for both appreciable benefit as well as the potential for harm. Few events reported do not meet the optimal information size and suggest fragility in the estimate. e. The 95%CI may not include a clinically meaningful benefit. f.. Spinner 2020 reported an odds ratio of 1.65 (95%CI: 1.09, 2.48); however, compared to relative risks, odds ratios tend to overestimate the effect with baseline risk is high. References 1. Spinner CD, Gottlieb RL, Criner GJ, et al. Effect of Remdesivir vs Standard Care on Clinical Status at 11 Days in Patients With Moderate COVID-19: A Randomized Clinical Trial. JAMA 2020; 324(11): 1048-57. 2. Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for the Treatment of Covid-19 - Final Report. N Engl J Med 2020; 383(19): 1813-26. 3. PanH, Peto R, Karim QA, et al. Repurposed antiviral drugs for COVID-19; interim WHO SOLIDARITY trial results. MedRxiv 2020. Available at: https://doi. org/10.1101/2020.10.15.20209817 [Preprint 15 October 2020]. Version 4.2.0 65 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 11. GRADE evidence profile, Recommendation 11 Question: Remdesivir compared to no antiviral treatment for hospitalized patients with severe COVID-19 Lastreviewed and updated 11/22/2020 Mortality (follow up: range 28 days to 29 days) ad Inconsistenet Impr ” nl 1) Relative | Absolute di y P ETE remdesi 0 1 ME 1) 0 3123 [randomized| serious **| not serious not serious serious Y none 369/2726 | 374/2593 RR0.92 | 12fewer 00 O O CRITICAL trials (13.5% | (144% |(0.77 to 1.10)| per 1,000 Low (from 33 fewer to 14 more) Time to recovery (follow up: 29 days) 12 |randomized| serious* | notserious not serious | notserious none 345/486 306/471 Rate ratio | 97 more 009 O CRITICAL trials (110% | (65.0% o TE (4.12 to 1.52) | (from 41 more to 147 more) Clinical improvement (follow up: 23 days) 11 - [randomized| not serious | not serious not serious [very serious Y none 103/158 45/78 RR1.13 | 75 more 00 OO CRITICAL trials ab (65.2% | (57.7% |(0.91 to 1.41)| per 1,000 0% (from 52 fewer to 237 more) Need for mechanical ventilation (follow up: 29 days) 12 |randomized| not serious| — not serious not serious serious * none 52/402 82/364 RR0.57 | 97fewer 009 O CRITICAL trials (12.9% | (22.5% |(0.42 to 0.79)| per 1,000 Grom1a1 | MODERATE fewer to 47 fewer) Serious adverse events (grade 3/4) Version 4.2.0 66 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 12. GRADE evidence profile, Recommendation 12 Question: Remdesivir 5 days compared to remdesivir 10 days for hospitalized patients with severe (not critically il) COVID-19 Last updated 9/10/2020; last reviewed 11/22/2020 Study DEIEES (nx D e remdesi AC o Meson Exeliy di y P MO Earn 5 days EN (9 1) Mortality 11 [randomized| serious? | notserious not serious serious 2 none 16/200 21/1197 HR 0.75 | 27fewer 6900 CRITICAL trials (8.0% (10.7% - |(0.40to 1.39)| per 1,000 LOW (from 64 fewer to 42 more) Clinical improvement at 14 days 11 |randomized| serious? | notserious not serious serious * none 129/200 107/197 RR1.19 [103 more 6900 CRITICAL trials (645% | (543% |(1.01to1.40)| per 1,000 Low (from 5 more to 217 more) SAEs 11 [randomized| serious? | notserious not serious serious * none 42/200 68/197 RR0.61 |135fewer 6900 CRITICAL trials (10% | (345% |(0.44to0.85)| per 1,000 Low (from 193 fewer to 52 fewer) AEs leading to treatment discontinuation 11 — [randomized| serious bd | notserious not serious serious none 9/200(45%)| 20/197 RR0.44 | 57fewer 6900 CRITICAL trials (10.2% - |(0.21to.0.95)| per 1,000 LOW (from 80 fewer to 5 fewer) Cl: Confidence interval, RR: Risk ratio Explanations a. The 95%CI includes the potential for both appreciable benefit, as well as appreciable harm. Few events reported do not meet the optimal information size and suggest fragility in the estimate. Version 4.2.0 69 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. b. Goldman 2020 did not blind participants, healthcare workers or outcome assessors. After randomization, disease severity was greater in the 10-day arm; while the analysis adjusted for baseline characteristics including disease severity, there is still the potential for residual confounding. c. The lower boundary of the 95%.Cl may not include a clinically meaningful effect. Few events reported do not meet the optimal information size and suggest fragility in the estimate. d. Goldman stratified adverse events by days 1-5, 6-10. AEs leading to treatment discontinuation during days 1-5 were 9 (4% in the 5-day arm and 14 (7% in the 10-day arm. Reference 1. Goldman JD, Lye DCB, Hui DS, et al. Remdesivir for 5 or 10 Days in Patients with Severe Covid-19. N Engl J Med 2020; 383: 1827-37. Version 4.2.0 70 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Famotidine Section last reviewed and updated 6/22/2020 Last literature search conducted 6/18/2020 Recommendation 13: Among hospitalized patients with severe COVID-19, the IDSA panel suggests against famotidine use for the sole purpose of treating COVID-19 outside of the context of a clinical trial. (Conditional recommendation, very low certainty of evidence) The last literature search was conducted on June 18, 2020 and we identified one non- randomized study in OVID. There were no new non-indexed RCTSs available. Why is famotidine considered for treatment? Anecdotal reports from China suggest that patients infected with coronavirus who were receiving famotidine, a H2 receptor antagonist to treat conditions such as acid reflux and peptic ulcer disease, had improved survival vs. those receiving proton pump inhibitors (PPIs) [135]. This post hoc finding summarized below has led to interest in the drug, though no predominant theory describing a mechanism for its efficacy yet exists. One theory is that famotidine, like many other compounds, binds and therefore inhibits the coronavirus main protease, 3C-like main protease (3CLpro) [136]. Summary of the evidence Our search identified one cohort study that compared 84 patients treated with famotidine against 1,536 patients not receiving treatment with famotidine [137] (Table 13). Fifteen percent of patients in the famotidine group (13/84) started famotidine at home before presenting to the hospital. In addition, a subset of 420 patients not treated with famotidine were matched on baseline characteristics to the treated patients. Benefits Famotidine may decrease the composite outcome of death or intubation (HR: 0.42; 95% Cl: 0.21, 0.85; Very low CoE); however, the evidence is very uncertain (Table 13). Version 4.2.0 71 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. c. Concerns about selective reporting due to unavailability of disaggregated data for outcomes of mortality or intubation, missing supplemental files, and raw data for primary outcome from propensity-matched control group. Reference 1. Freedberg DE, Conigliaro J, Wang TC, et al. Famotidine use is associated with improved clinical outcomes in hospitalized COVID-19 patients: A propensity score matched retrospective cohort study. Gastroenterology 2020; 159(3): 1129-31. Version 4.2.0 74 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Neutralizing Antibodies Section last reviewed and updated 4/11/2021 Last literature search conducted 3/31/2021 Recommendation 14: Among ambulatory patients with mild to moderate COVID-19 at high risk for progression to severe disease, the IDSA guideline panel suggests bamlanivimab/etesevimab or casirivimab/imdevimab rather than no neutralizing antibodies. (Conditional recommendation, low certainty of evidence) + Remarks: o Patients with mild to moderate COVID-19 who are at high risk of progression to severe disease admitted to the hospital for reasons other than COVID-19 may also receive bamlanivimab/etesevimab or casirivimab/imdevimab. o Local variant susceptibility may be considered in the choice of the most appropriate neutralizing antibody therapy. o There are limited data on efficacy of bamlanivimab/etesevimab or casirivimab/imdevimab in high-risk patients between 12 and 18 years of age. Recommendation 15: Among hospitalized patients with severe COVID-19, the IDSA guideline panel recommends against bamlanivimab monotherapy. (Strong recommendation, Moderate certainty of evidence) Version 4.2.0 75 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Figure 2. FDA EUA criteria for the use of casirivimab/imdevimab and bamlanivimab/etesevimab 12a This EUA is for the use of the unapproved products bamlanivimab, casirivimab and imdevimab, and/or bamlanivimab and etesevimab for the treatment of mild to moderate COVID-19 in adults and pediatric patients (12 years of age and older weighing at least 40 kg) with positive results of direct SARS-CoV-2 viral testing, and who are at high risk for progressing to severe COVID-19 and/or hospitalization. High risk is defined as patients who meet at least one of the following criteria: + Have a body mass index (BMI) >35 + Have chronic kidney disease + Have diabetes + Have immunosuppressive disease + Are currently receiving immunosuppressive treatment + Are265 years of age + Are255 years of age AND have o cardiovascular disease, OR o hypertension, OR o chronic obstructive pulmonary disease/other chronic respiratory disease. + Are 12-17 years of age AND have o BMI 285th percentile for their age and gender based on CDC growth charts, OR sickle cell disease, OR congenital or acquired heart disease, OR neurodevelopmental disorders, for example, cerebral palsy, OR a medical-related technological dependence, for example, tracheostomy, 0000 gastrostomy, or positive pressure ventilation (not related to COVID-19), OR o asthma, reactive airway or other chronic respiratory disease that requires daily medication for control. a. These criteria refer to Recommendation 14. Reference 1. U.S. Food and Drug Administration. Fact Sheet for Health Care Providers: Emergency Use Authorization (EUA) of Bamlanivimab and Etesevimab. Available at: https://www.fda.gov/media/145802/download. Accessed 9 April 2021. 2. U.S. Food and Drug Administration. Fact Sheet for Health Care Providers: Emergency Use Authorization (EUA) of Casirivimab and Imdevimab. Available at: https://www.fda.gov/media/143892/download. Accessed 9 April 2021. Version 4.2.0 76 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. relative risk of persistently high viral load at day seven compared to no bamlanivimab/etesevimab (RR: 0.34; 95% Cl: 0.25-0.46; low CoE). Casirivimab/imdevimab Among ambulatory patients with at least one risk factor for severe disease, casirivimab/imdevimab demonstrated a lower relative risk of mortality compared to no casirivimab/imdevimab (RR: 0.28; 95% Cl: 0.05, 1.40; moderate CoE). In ambulatory patients with at least one risk factor for severe disease, there was a lower relative risk of hospitalization in patients treated with casirivimab/imdevimab compared to no casirivimab/imdevimab (RR:0.45; Cl: 0.30, 0.67; low CoE). Hospitalization was defined as a medically attended visit and included telemedicine visits, in-person visits, urgent care or emergent department visits, and hospitalization. The low certainty of evidence was due to indirectness, as hospitalization is a surrogate for mortality, and imprecision, due to few events recorded. Bamlanivimab monotherapy Among ambulatory patients, bamlanivimab demonstrated a lower relative risk of hospitalization, including visits to the emergency room, compared to no bamlanivimab (RR: 0.26; 95% Cl: 0.09, 0.75; very low CoE). The very low certainty of evidence was due to indirectness, as the treatment may not have been provided to enough patients at risk of developing severe disease to be representative of the general population, and imprecision, due to few events recorded. Bamlanivimab may increase viral clearance at three days (mean difference: -0.49; 95% Cl: -0.87, -0.11; low CoE); however, there may not be a meaningful difference at 11 days as measured by change from baseline SARS-CoV-2 viral load (mean difference: -0.22; 0.95: -0.60, 0.15; low CoE). Among patients hospitalized for COVID-19, treatment with bamlanivimab compared to placebo failed to show or exclude a beneficial effect on mortality (HR: 2.00; 95% Cl: 0.67, 5.99; moderate CoE). Clinical improvement, as defined as a decrease in a pulmonary ordinal scale, may not be meaningfully different among patients hospitalized for COVID-19 who received treatment with bamlanivimab or placebo (OR: 0.85; 0.56, 1.29; moderate CoE). Version 4.2.0 79 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Harms Bamlanivimab/etesevimab Patients receiving bamlanivimab/etesevimab experienced more serious adverse events. However, this may not be meaningfully different from those receiving placebo (RR: 1.40; 95% Cl: 0.45, 4.37; moderate CoE). Casirivimab/imdevimab Serious adverse events were less frequent among patients receiving casirivimab/imdevimab compared to those receiving placebo (RR: 0.31; 95% Cl: 0.20, 0.46; low CoE). Bamlanivimab monotherapy Serious adverse events among ambulatory patients receiving bamlanivimab monotherapy may not be meaningfully different from those receiving placebo (RR: 0.15; 95% Cl: 0.01, 3.78; low CoE). Patients receiving bamlanivimab did experience more infusion-related adverse events, including pruritus, flushing, rash, and facial swelling (RR: 1.62; 95% Cl: 0.34, 7.70; low CoE). Similarly, serious adverse events at five and 28 days among patients hospitalized for COVID-19 receiving bamlanivimab may not be meaningfully different from those receiving placebo (RR: 1.85; 95% Cl: 0.34, 9.97; moderate CoE and RR: 0.93, 95% Cl: 0.27, 3.14; moderate CoE, respectively). Similarly, infusion-related adverse events may not be meaningfully different between patients hospitalized for COVID-19 receiving bamlanivimab or placebo (OR: 1.64, 95% Cl: 0.79, 3.44; moderate CoE). Other considerations Bamlanivimab/etesevimab The panel agreed that the overall certainty of evidence for the treatment with bamlanivimab/etesevimab in ambulatory patients with COVID-19 was low due to imprecision Version 4.2.0 80 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. and indirectness of outcomes (i.e., different dose recommended in the EUA when compared to the dose used in the trial; viral loads or hospitalizations are indirect markers of severity of disease and death). The guideline panel made a conditional recommendation for using bamlanivimab/etesevimab in mild to moderate COVID-19 at high risk for developing severe disease as the expected benefits likely outweigh any potential harms. The FDA EUA for bamlanivimab/etesevimab was issued for the dosage of 700 mg instead of 2,800 mg of bamlanivimab, and 1,400 mg instead of 2,800 mg for etesevimab. Summary rationale for the lower dose is provided in the bamlanivimab/etesevimab FDA EUA FAQ stating “based on analysis of the available nonclinical, clinical, and virologic data, as well as supportive data from pharmacokinetic/pharmacodynamics modeling, the authorized dosage of 700 mg bamlanivimab and 1,400 mg of etesevimab is expected to have similar clinical effects to a dosage of 2,800 mg bamlanivimab and 2,800 mg of etesevimab used in the clinical study.” The lower dose from the EUA was not studied in the phase lll trial BLAZE-1. The extrapolation that the lower dose has the same efficacy as the higher dose, based on surrogate outcomes from other study lowers certainty of evidence due to indirectness. Casirivimab/imdevimab The panel agreed that the overall certainty of evidence for the currently available data on the treatment with casirivimab/imdevimab for ambulatory patients with COVID-19 is low dueto concerns with indirectness and imprecision due to sparseness of events. Bamlanivimab monotherapy The panel agreed that the overall certainty of evidence for treatment with bamlanivimab for ambulatory patients with COVID-19 is very low due to concerns with indirectness and imprecision. Though the treatment with combination bamlanivimab/etesevimab is preferred compared to bamlanivimab monotherapy due to the more robust data supporting it, the panel recognized that short-term supply constraints for bamlanivimab/etesevimab may necessitate the use of bamlanivimab monotherapy in some situations. Version 4.2.0 81 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 14. GRADE evidence profile, Recommendation 14 Question: Bamlanivimab/etesevimab compared to no bamlanivimab/etesevimab for ambulatory patients with mild to moderate COVID-19 at high risk for progression to severe disease Last updated 3/2/2021; last reviewed 4/11/2021 A EE Study INES (CS INTO design 0 n US CEUEnEN [E (95% CI) etesevimab Mortality (follow up: 29 days) 11 [randomize| not not serious | not serious | very serious none 0/518 (0.0%) 10/517 (1.9%) RR 0.05 19 fewer per 06900 | CRITICAL dtrials | serious a b (0.00 to 0.80) 1,000 Low (from 31 fewer to 7 fewer) < Hospitalization (>1=24 hours of acute care) with COVID-19 (follow up: 29 days) 11 [randomize| not notserious | serious 24| serious? none 11/518 (2.1%) 36/517 (7.0%) RR 0.30 49 fewer per 06900 | CRITICAL dtrials | serious (0.16 to 0.59) 1,000 Low (from 58 fewer to 29 fewer) Persistently high viral load (PHVL) at day 7 (follow up: 7 days; assessed with: RT-PCR) 11 [randomize| not notserious | serious 3* | serious? none 50/508 (9.8%) | 145/499 (29.1%) RR0.34 192 fewer per 06000 [IMPORTAN dtrials | serious (0.25 to 0.46) 1,000 Low T (from 218 fewer to 157 fewer) Serious adverse events 11 [randomize| not notserious |notserious| serious? none 7/518 (1.4%) 5/517 (1.0%) RR1.40 |4more per 1,000 000 CRITICAL dtrials | serious a (0.45 to 4.37) | (from 5 fewer to 33 more) MODERATE GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: 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 Low certainty: Our confidence in the effect estimate ¡s limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Version 4.2.0 84 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Risk of bias: Study limitations Inconsistency: Unexplaied heterogeneity across study findings Indirectness: Applicabilty or generalizabiity to the research question Imprecision: The confidence in the estimate of an effectto support a particular decision Publication bias: Selective publication of studies Ct: Confidence interval; RR: Risk ratio Explanations a. Estimate reflects the use of a higher dose than treatment dose approved by the FDA. Fragility present, low number of events Zero events preventing RR estimate Hospital admission is an intermediary outcome for morbidity, ICU admission, and need for ventilation Measure of viral clearance is a surrogate outcome for hospital admission, need for intensive care, intubation and death. f. RR estimated by using continuity correction of 0.5. References 1. Dougan M, Nirula A, Azizad M, et al. The Impact of Bamlanivimab + Etesevimab Neutralizing Antibody Combination Treatment on Hospitalization Rates and Deaths Among High-Risk Patients Presenting With Mild-to-Moderate COVID-19 Illness. 2021: [Under review]. PDPaSoDs Version 4.2.0 85 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 15. GRADE evidence profile, Recommendation 14 Question: Casirivimab/imdevimab compared to no casirivimab/imdevimab for ambulatory patients with mild to moderate COVID-19 at high risk of progression to severe disease New evidence profile developed 4/11/2021 UT design Inconsistency Imprecision Mortality (1,200 mg + 2,400 mg; mITTIMFAS) (follow up: 29 days) USES (150) 1 randomized not not serious not serious serious none 212091 5/1341 (0.4%) | RR0.28 | 3fewer 9000 CRITICAL trials — | serious* (0.19% (0.05 to per | MODERATE 1.40) 1,000 (from 4 fewer to 1 more) Hospitalization (medically attended visits; 1,200 mg + 2,400 mg; mITTImFAS) (follow up: 29 days) 1 randomized not not serious serious > serious none 40/2091 57/1341 RR0.45 | 23fewer | 06960OD CRITICAL trials — | serious* (1.99% (43% (0.30 to per LOW 0.67) 1,000 (from 30 fewer to 14 fewer) Serious adverse events (1,200 mg + 2,400 mg) (follow up: 29 days) 1 randomized not not serious not serious serious none 33/2676 74/1843 RR0.31 | 28 fewer 0000 CRITICAL trials serious 2 (12%) (40%) (0.20 to per MODERATE 0.46) 1,000 (from 32 fewer to 22 fewer) GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: 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 ¡tis substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Version 4.2.0 86 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. E 2 [espe li Ev 0) Masren 11 [randomized not not serious not serious | very serious none 7/309(23%) | 2143 (1.4% RR 1.62 9 more e6900 CRITICAL trials — | serious (0.34to | per1,000 LOW 770) | (from9 fewer to 94| more) GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true efectis likely to be close to the estimate of the effect, but there is a possibility that itis substantialy diferent Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantally differentfrom the estimate of the effect Very low certainty: We have very litle confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplaied heterogeneity across study findings Indirectness: Applicabilty or generalizability to the research question Imprecision: The confidence in the estimate of an effect to support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval, RR: Risk ratio; MD: Mean difference Explanations a. Uncertain that the treatment was provided in enough patients at risk of developing severe disease to be representative of the general population. b. The 95%CI may not include a meaningful difference. Few events reported suggests fragility of the estimate. c. Measure of viral clearance is a surrogate outcome for hospital admission, need for intensive care, intubation and death. d. The 95%CI includes values that suggest either an increase or decrease in harm. Few events reported suggests fragility of the estimate. References 1. ChenP, Nirula A, Heller B, et al. SARS-CoV-2 Neutralizing Antibody LY-CoV555 in Outpatients with Covid-19. N Engl J Med 2021; 384(3): 229-37. Version 4.2.0 89 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 17. GRADE evidence profile, Recommendation 15 Question: Bamlanivimab compared to no bamlanivimab for patients hospitalized for COVID-19 Last updated 1/29/2021; last reviewed 4/11/2021 LL n (eva Study A np Other Relative | Absolute BH AREA En A TS EEES Mortality 11 randomized| not not serious not serious serious 2 none 9/163 (5.5%) 5/151 (3.3%) | HR2.00 | 32 more | 96900 CRITICAL trials — [serious (0.67 to | per 1,000 | MODERATE 5.99) | (from 11 fewer to 150 more) Clinical improvement at day 5 (assessed with: pulmonary ordinal outcome [scale 1-7; 1 = least severe)) 11 randomized| not not serious not serious serious 2 none 161 150 OR 0.85 - 0000 CRITICAL trials — | serious (0.56 to MODERATE 1.29)» Serious adverse events (follow up: 5 days) 11 randomized| not not serious not serious serious 2 none 4/163 (2.5%) 2/151 (13%) | RR1.85 | 11 more | 99060 CRITICAL trials — [serious (0.34 to | per 1,000 | MODERATE 9.97) | (trom9 fewer to 119 more) Serious adverse events (follow up: 28 days) 11 randomized| not not serious not serious serious 2 none 5/163(31% | 5/151(3.3%) | RR0.93 | 2fewer | H0ODO | IMPORTANT trials — [serious (0.27 to | per 1,000 | MODERATE 3.14) | (from24 fewer to 71 more) Infusion-related adverse event 11 randomized| not not serious not serious serious 2 none 23/163 (14.1%) | 21/151 (139%) | OR 1.64 | 70 more | S60S0 [IMPORTANT trials — [serious (0.79 to | per 1,000 | MODERATE 3.44)< | (from 26 fewer to 218 more) Version 4.2.0 90 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of Ihe estimate of the efect Moderate certainty: 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 itis substantialy diferent Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantaly diferent from the estimate ofthe effect Very low certainty: We have very litle corfidence in the effect estimate: The true effect is likely to be substantialy diferent from the estimate of effect Risk of bias: Study limitations Inconsistency: Unexplaed heterogeneity across study findings Indirectness: Applicabilty or generalizabiity to the research question Imprecision: The confidence in the estimate of an effect to support a particular decision Publication bias: Selective publication of studies Cl: Confidence interval; HR: Hazard Ratio; OR: Odds ratio; RR: Risk ratio Explanations a. The 95%C1 includes the potential for both appreciable benefit as well as the potential for harm. Few events reported do not meet the optimal information size and suggest fragility of the estimate b. Study-provided odds ratio adjusted for baseline ordinal category and trial pharmacy. c. Study-provided odds ratio adjusted for the trial pharmacy. References 1. ACTIV-3/TICO LY-Co0V555 Study Group, Grund B, Barkauskas CE, et al. A Neutralizing Monoclonal Antibody for Hospitalized Patients with Covid-19. N Engl J Med 2020: [Epub ahead of print 22 December 2020]. Version 4.2.0 91 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Harms In ACTT-2, patients receiving baricitinib and remdesivir had a lower risk of developing any serious adverse events through day 28 (16% vs. 21%; RR 0.76; 95% Cl 0.59,0.99; Moderate CoE) whether or not thought to be related to the study drug. In this trial, the overall rate of new infections was lower in the baricitinib plus remdesivir group compared with remdesivir alone (30 patients [5.9%] versus 57 patients [11.2%]) [157]. However, patients who received concomitant glucocorticoids had a higher incidence of serious or non-serious infection as compared with those who did not: 25.1% and 5.5% respectively. lt was not specified what proportion of these patients were in the baricitinib combination group versus the control group in the study. Based on experience in clinical trials for RA, baricitinib has been found to be associated with an increased risk of adverse effects including infections (especially upper respiratory tract infections), thrombosis, lymphopenia, anemia, increases in lipids, elevations in liver enzymes, and elevations in creatinine phosphokinase [158]. Many of these side effects are thought to be dose related, with increased incidence in patients taking baricitinib 4 mg compared with 2 mg. Patients in ACTT-2 received baricitinib 4 mg daily for two weeks or until discharge, a shorter duration than those taking the drug for RA. In clinical trials for RA, baricitinib was associated with a numerically higher risk of upper respiratory tract infections and herpes simplex and herpes zoster infections compared with placebo [159]. Opportunistic infections have been reported in patients taking baricitinib. Patients with COVID-19 have been found to have abnormalities in coagulation parameters and might have an elevated risk of trombosis [160]. Baricitinib has been found to increase incidence of thrombosis compared with placebo in clinical trials for its FDA approval for rheumatoid arthritis, especially at a higher dose of 4 mg daily [158]. During the 16-week treatment period in RA trials, VTE occurred in five patients treated with baricitinib 4 mg daily, compared with zero in the 2 mg daily and placebo groups. Arterial thrombosis occurred in two patients treated with baricitinib 4 mg, two patients treated with baricitinib 2 mg, and one patient on placebo. In ACTT-2, the percentage of patients reported to have VTE was numerically higher in the combination group (21 patients [4.1%] vs 16 patients [3.1%]) although it was Version 4.2.0 94 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. similar overall (absolute difference 1%, 95% Cl -1.3 to 3.3) [157]. Of note, all patients in the trial were recommended to receive VTE prophylaxis if they had no contraindication. Other considerations The panel agreed on the overall certainty of evidence as low due to concerns with risk of bias, driven by the use of data from post-hoc analyses, and imprecision, which recognized the limited events and concerns with fragility in the group who likely benefited most (those requiring supplemental oxygen or non-invasive ventilation). The guideline panel recognized the uncertainty of potential harms of baricitinib with remdesivir given in combination with corticosteroids (dexamethasone is recommended for patients with severe or critical disease); however, noted the importance of suggesting baricitinib plus remdesivir as an option for persons unable to receive corticosteroids. Conclusions and research needs for this recommendation The guideline panel suggests baricitinib with remdesivir for persons who cannot receive corticosteroids. Baricitinib plus remdesivir should be reserved for patients who cannot take corticosteroids because dexamethasone has been proven to reduce mortality in patients hospitalized with COVID-19 who require supplemental oxygen or mechanical ventilation and, for this reason, dexamethasone is recommended by the panel for this group. It is uncertain whether baricitinib plus remdesivir will have the same benefit as dexamethasone, and there are insufficient data to recommend the use of dexamethasone along with baricitinib plus remdesivir. The guideline panel recommenads baricitinib with remdesivir in combination with corticosteroids only in the context of a clinical trial. Additional clinical trials are needed to better understand the potential harms when baricitinib is given with corticosteroids and whether there is a benefit of treatment for patients with COVID-19 (Table s2). Version 4.2.0 95 Last updated April 14, 2021 and posted online at www.idsociety.org/COVID19guidelines. Please check website for most updated version of these guidelines. Table 18. GRADE evidence profile, Recommendation 16 Question: Baricitinib with remdesivir compared to remdesivir for hospitalized patients with COVID-19 New evidence profile developed 12/23/2020 (EST ca SO Impr n l y 1d O Mortality (follow up: 23 days) 11 — [randomized| not serious | not serious not serious serious 2 none 24/515 37/518 HR 0.65 | 24fewer a) CRITICAL trials (4.7% (7.1% — |(0.39to 1.09) | per 1,000 (from 43 MODERATE fewer to 6 more) Clinical recovery - hospitalized requiring supplemental 02/r ing noninvasive ventilation or high-flow 02 (ordinal 5+6) (assessed with: Ordinal scale <4) 11 [randomized| serious* | not serious not serious serious * none 344/391 316/389 RR1.08 | 65more 0 00 CRITICAL trials (e80% | (81.2% |(1.02to 1.15)| per 1,000 Low (from 16 more to 122 more) Clinical recovery - receiving noninvasive ventilation or high-fiow 02, invasive mechanical ventilation or ECMO (ordinal 6+7; stratified) (assessed with: Ordinal scale <4) 11 — [randomized| not serious | not serious not serious serious * none 1221176 114/191 HR 1.29 | 93more 00 0 CRITICAL trials a (69.3% | (597% |(1.00to1.66)| per 1,000 a (from O MODERATE fewer to 182 more) New use of mechanical ventilation or ECMO (follow up: 29 days) 11 [randomized] serious" | not serious not serious serious 9 none 46/461 70/461 RR0.66 | 52fewer 00 O O CRITICAL trials (10.0% | (152% |(0.46to0.93)| per 1,000 Low (from 82 fewer to 11 fewer) Serious adverse events (follow up: 28 days) Version 4.2.0 96