Harmonizing heterogeneous endpoints in COVID-19 trials without loss of information - an essential step to facilitate decision making

Maja von Cube; Marlon Grodd; Martin Wolkewitz; Derek Hazard; Jerome Lambert

Our review of registered interventional COVID-19 trials revealed high heterogeneity in the definition of primary endpoints. Most of the studies used a recovery endpoint for the primary analysis. We propose to accompany the primary analysis with multistate methodology. We highlight, that literature provides an extensive statistical tool box of suitable models to analyze clinical trials. All ongoing clinical trials, especially those with severe cases, should accommodate primary analysis with a stacked probability plot of the major events mechanical ventilation, discharge alive and death. The resulting informative plot is a powerful tool to harmonize the diversity of clinical endpoints and lengths of follow-up and thereby fastening accessibility of evidence and thus decision making. Additionally, treatment effects on a number of (potentially clinically opposite) endpoints can be studied simultaneously over time using cause-specific Cox regression and by estimating sojourn time spent in the various states. Moreover, while mechanical ventilation, hospitalization and mortality are objective outcomes, recovery and other states of severity are influenced by clinical judgement and may be additionally subject to measurement errors.