ICMRA COVID-19 Treatments and Clinical Trials Workshop #2 - Summary

July 2020


The SARS-CoV-2 pandemic has affected to date more than 14,000,000 people worldwide and continues to present an extraordinary challenge to global health. Many applicants developing medicines for the treatment of COVID-19 are in the process of conducting clinical trials and/or approached National Regulatory Authorities (NRAs) with their proposals for Phase 3 clinical trials.
The July 20, 2020 teleconference of global regulators convened under the auspices of ICMRA, discussed, based on experience accumulated by various regulators from proposals received from developers, the acceptability of various endpoints that could be used as primary endpoints in the clinical trials conducted for the treatment of COVID-19.

Key topics:

  • Putative primary endpoints for randomized controlled clinical trials involving medicines used for the treatment of COVID-19.

The following represents generally agreed positions among global regulators in attendance.

Primary endpoints for RCTs involving medicines used for the treatment of COVID-19

  • A range of different primary endpoints have been proposed by sponsors of clinical trials for COVID therapeutics, depending on the trial population, e.g. mild disease in outpatients vs. severe disease in hospitalized patients.
  • It was felt important to discuss and possibly agree on which endpoints would be deemed acceptable by regulators, to facilitate rapid and consistent implementation of future clinical trials across the world of medicines developed for COVID-19 treatment.
  • The workshop started with presentations given by some of the participating Agencies illustrating various primary endpoints included in some of the ongoing/completed clinical trials and deemed acceptable by the reviewing agencies.
  • It was reflected that in only 5 months into this pandemic with a new virus our understanding of the natural history of COVID-19 disease has progressed swiftly, but still a more granular characterization of the disease course and how to manage different patient’s phenotypes would be highly informative for clinical studies design of investigational COVID-19 treatments.
  • The primary endpoint ideally should be clinically meaningful (capturing patient function as well as survival), but also it needs to be measurable, to be sufficiently sensitive to allow realistic sample sizes. A sensible and suitable way of handling the missing data / intercurrent events and mortality also needs to be defined.


  • For hospitalized patients with moderate/severe COVID-19, it was agreed that there is a portfolio of options that can be considered suitable as primary endpoint to inform on the clinical benefit of investigational therapeutics for COVID-19 and that could support regulatory approval. Among these, time to recovery through D28/29, clinical status as improvement of 2 points on an ordinal scale (primary analysis using a proportional odds model, adjusting for baseline status and covariates), mortality within 28 days after randomization (especially for severe COVID-19) or time to sustained recovery up to day 90 (defined as alive and no relapse).
  • While it was agreed that mortality is not the only acceptable primary endpoint, still it should be collected as a key secondary endpoint in all studies that don’t plan to use this outcome as primary. In the context of time to event endpoints, the way recorded deaths are handled should be discussed and clarified with regulators. For time to event endpoints, censoring deaths at D28 (or last timepoint in the window of investigation) as worst outcome was considered an acceptable approach, providing mortality is evaluated as a key secondary endpoint.
  • Other endpoints that were discussed were: progression of disease, number of days not on ventilator from randomization to a specified timepoint, recovery rates.
  • For endpoints based on proportional odds at a fixed timepoint, it remains to be discerned how to select the timepoint that would be mostly informative in capturing a treatment effect that would be clinically meaningful.


  • For outpatients with mild COVID-19, it was agreed that mortality as the primary endpoint may not be suitable; rate of progression to severe disease, proportion of patients not hospitalized at a prespecified time point may be appropriate, depending on the primary objective of the study. It was felt warranted that objective criteria for disease progression should be included in the endpoint definition. Some endpoints such as rate of hospitalization are not considered objective enough as to be accepted.
  • It was basically agreed that a virological endpoint could be useful for proof-of-concept studies, but primary endpoint for phase III clinical trials should be decided, reflecting the primary objective of the study.
  • The evolving landscape of standard-of-care in the treatment of COVID-19 will need to be duly considered in the design of future studies.


Due to time constraints it was agreed to discuss other important aspects (patient populations enrolled, definition of control arm) in a future workshop.