Definitions matter: Heterogeneity of COVID-19 disease severity criteria and incomplete reporting compromise meta-analysis

• DOI: 10.1371/journal.pgph.0000561
• Publication Date: 2022-07-19T17:24:04Z
• Authors: Philippe J. Guérin; Alistair R. D. Mc...; Sumayyah Rashan; AbdulAzeez Lawal; James A. Watson; Nathalie Strub-Wo...; Nicholas J. White
• Keywords: Epidemiology; MEDLINE; FOS: Political science; 610; FOS: Law; Infectious disease (medical specialty); FOS: Health sciences; Severity of illness; Coronavirus Disease 2019; 03 medical and health sciences; 0302 clinical medicine; Health Sciences; Neurological Manifestations of COVID-19 Infection; Intensive care medicine; Disease; Critically ill; Internal medicine; Political science; 3. Good health; Coronavirus disease 2019 (COVID-19); Meta-analysis; Infectious Diseases; Neurology; Randomized controlled trial; Medicine; Epidemiology and Management of Sepsis and Septic Shock; Public aspects of medicine; RA1-1270; Critical illness; Law; Clinical Characteristics; Research Article

Therapeutic efficacy in COVID-19 is dependent upon disease severity (treatment effect heterogeneity). Unfortunately, definitions of vary widely. This compromises the meta-analysis randomised controlled trials (RCTs) and therapeutic guidelines derived from them. The World Health Organisation 'living' for treatment are based on a network (NMA) published RCTs. We reviewed 81 studies included WHO living NMA compared their classifications with employed by international COVID-NMA initiative. two were concordant only 35% (24/68) trials. Of RCTs evaluated, 69% (55/77) considered group to include patients range severities (12 mild-moderate; 3 mild-severe; 18 mild-critical; 5 moderate-severe; 8 moderate-critical; 10 severe-critical), but distribution within these groups usually could not be determined, data duration illness and/or oxygen saturation values often missing. Where clear there was substantial overlap mortality across different strata. imprecision assessment validity some recommendations; notably extrapolation "lack benefit" shown hospitalised severely ill respiratory support ambulant mildly warranted. Both harmonised unambiguous individual patient (IPD) meta-analyses needed guide improve recommendations COVID-19. Achieving this goal will require improved coordination main stakeholders developing medicine regulatory agencies. Open science, including prompt sharing, should become standard allow IPD meta-analyses.