Non-invasive respiratory support strategies in COVID-19

In hospitalised patients with COVID-19, an increase in oxygen requirements prompts the clinician to decide how and when escalate treatment. A key treatment goal is avoid, where possible, need for invasive mechanical ventilation. However, up 20% of UK require admission critical care units, around 40% those requiring ventilation COVID-19 pneumonitis do not survive.1Docherty AB Harrison EM Green CA et al.Features 20 133 hospital covid-19 using ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study.BMJ. 2020; 369m1985Crossref PubMed Scopus (1821) Google Scholar To date, only treatments that have been shown reduce are dexamethasone interleukin-6 blockade. Non-invasive respiratory support strategies, such as continuous positive airway pressure (CPAP) or high-flow nasal (HFNO), attractive options might avoid its inherent risks. context concern has raised these strategies cause harm through delays tracheal intubation exacerbation lung injury, health-care workers nosocomial infection, systems high demand devices. This uncertain balance harms benefits resulted marked variation international practice. survey 1132 participants across 85 countries used a case vignette previously healthy patient severe hypoxaemia;2Azoulay E de Waele J Ferrer R al.International management patients.Crit Care. 24: 486Crossref (39) choice initial strategy included HFNO (47%), CPAP non-invasive (26%), immediate (7%), remaining respondents opting optimise conventional therapy.2Azoulay Variability practice was associated country, rurality, intensive unit bed availability, individual characteristics. There paucity high-quality evidence COVID-19. One multicentre, randomised controlled trial (RCT)3Grieco DL Menga LS Cesarano M al.Effect helmet noninvasive vs on days free moderate hypoxemic failure: HENIVOT randomized clinical trial.JAMA. 2021; (published online March 25.)https://doi.org/10.1001/jama.2021.4682Crossref (195) reported no difference support-free 109 hypoxaemia who were treated either oxygen, although limitation this study absence control group receiving standard therapy management. Other direct remains limited retrospective series studies inconsistent findings risk bias design.4Ashish Unsworth Martindale al.CPAP first quantitative analysis from inpatient service evaluation.BMJ Open Respir Res. 7e000692Crossref (35) Scholar, 5Demoule Vieillard Baron Darmon al.High-flow cannula critically ill COVID-19.Am Crit Care Med. 202: 1039-1042Crossref (147) 6Arina P Baso B Moro V al.Discriminating between success failure failure.Intensive 47: 237-239Crossref (16) For example, study6Arina rates 66% CPAP, mortality (55%) after failure. Evidence HFNO, effective acute hypoxaemic drawn populations without systematic review network meta-analysis7Ferreyro BL Angriman F Munshi L al.Association oxygenation all-cause adults meta-analysis.JAMA. 324: 57-67Crossref (200) concluded delivered by both mask interface reduced intubation, intubation. presenting community-acquired pneumonia. novel disease generalising data other causes inherently problematic. viral influenzae coronaviruses, excess 70% reported,8Esquinas AM Egbert Pravinkumar S Scala al.Noninvasive high-risk pulmonary infections: review.Eur Rev. 2014; 23: 427-438Crossref (40) serve delay, rather than regarding use more compliant lungs potential large tidal volume breathing self-induced which similar pathogenesis ventilator-induced injury. converse argument liberal likely ventilator-associated complications mortality.9Tobin Basing physiological principles.Am 201: 1319-1320Crossref (174) The transmission delivering centres aerosol generation. Early mechanistic evaluations droplet spread suggested risks comparable therapy. Generation aerosols be influenced device, settings, interface, but also characteristics, load coughing profile. substantive does indicate risk. Further research needed understand patients. International guidelines prolific (appendix pp 1–3). UK, clinicians informed recommendations NHS England care–anaesthesia communities, well global organisations. Across guidelines, there variability transparency development, process synthesising evidence, recommended approach. November, 2020, p 1) preferred form advised against basis perceived efficacy, use, infection workers, guidance under NICE. By contrast, Surviving Sepsis Campaign they acknowledge strength recommendation weak based low-certainty evidence. adopts balanced recommendation, including all justified inadequate base any Others, Australia New Zealand Intensive Society, moved away previous position favouring one over another, now their living 2) suggest decisions assessment setting, emphasis reducing workers. RCTs urgently evaluate effectiveness At present, driven personal preference influence, prior experience, local availability methods supplies. But backdrop safety, effectiveness, optimal approach failure, it essential demonstrate equipoise randomise into available trials jurisdictions. number reports pneumomediastinum pneumothorax support.10Elhakim TS Abdul HS Pelaez Romero C Rodriguez-Fuentes Y Spontaneous pneumomediastinum, subcutaneous emphysema pneumonia: rare literature review.BMJ Case Rep. 13e239489Crossref (62) These concern, confounded many unmeasured factors owing nature. Because being part usual settings harm, further compared care. should exclude contraindication ensure harms, incidence pneumothorax, reported. far largest area RECOVERY-Respiratory Support trial,11Perkins GD Couper K Connolly al.RECOVERY-Respiratory Support: suspected proven failure; pressure, care: structured summary protocol trial.Trials. 21: 687PubMed funded prioritised National Institute Health Research urgent public health study. adaptive, multicentre RCT evaluates primary outcome within 30 randomisation. As April 12, 2021, 1200 had randomly assigned, enrolment features two pathways 1). During pandemic, resources exceeds capacity,12Coppadoro Benini Fruscio al.Helmet treat hypoxic pneumonia outside ICU: during outbreak.Crit 25: 80Crossref (44) established viewed possible treatment, particularly if option participate trial. capacity issues exist, mindful provision rigorous infrastructure represents random empirical If interventions beneficial, will delayed answering question at hand. intervention shows favourable effect (or worse, harm), justify continued unproven framework, decision deny opportunity nationally research. Understanding most requires investigation relative wider system, can addressed randomisation trials. All authors responsible conduct delivery trial, (NIHR) referenced Comment. BC educational fees Fisher & Paykel, her institution receives funds NIHR she Director Society. GDP grants NIHR. DFM consultancy GlaxoSmithKline, Boehringer Ingelheim, Bayer, Novartis, Eli Lilly, sitting monitoring ethics committee undertaken Vir Biotechnology. DFM's received several funders distress syndrome COVID-19; addition, he patent (US8962032) issued his inflammatory disease. Society Programme Efficacy Mechanism Evaluation programme. Download .pdf (.23 MB) Help pdf files Supplementary appendix