WHO COVID-19 therapeutic guidelines - Authors' reply.

We thank Bram Rochwerg and colleagues for information on the WHO therapeutic guideline development process. Unfortunately, they do not address our main concern: the unjustified extrapolation of evidence from randomised controlled trials in severe COVID-19 to therapeutic guidelines for uncomplicated illness. Pooling summary data from studies with different severity definitions, deciding on inappropriate primary outcomes, and extrapolating from results in hospitalised patients to ambulant individuals with mild infections suggests a worrying lack of clinical judgement. The recent separate guidelines on COVID-19 chemoprophylaxis, which contain judgements on mortality prevention derived from trials with no mortality, have only magnified concerns about the WHO assessments.4, 5 The world looks to WHO for guidance and leadership in these difficult times. Therapeutic guidelines should be based on an understanding of the disease process, the health needs and health-system capabilities, the clinical pharmacology of the drugs, and the quality and weight of evidence. When advising on potential treatments, evidence from randomised clinical trials with patients who have severe COVID-19 should not be extrapolated to prevention and early treatment. Despite the undoubted equitability, impartiality, and rigour of the WHO COVID-19 therapeutic guideline process, there is something fundamentally wrong with it.

We declare no competing interests.