Back

COVID-19: Why SOLIDARITY and DisCoVeRy trials may fail to bring informative and timely results

TOUMI, M.; LIANG, S.; DABBOUS, M.; WANG, Y.; QIU, T.; HAN, R.

2020-06-02 infectious diseases
10.1101/2020.06.01.20118927 medRxiv
Show abstract

ObjectiveThe SOLIDARITY and DisCoVeRy trials were launched to facilitate the rapid worldwide comparison of the efficacy and safety of treatments against COVID-19. This study aimed to review the trial designs of SOLIDARITY and DisCoVeRy and their feasibility to generate high-quality evidence. MethodA systematic search of the European Clinical trial registry, the U.S. National Library of Medicine ClinicalTrials.gov, and the World Health Organizations (WHO) International Clinical Trials Registry Platform (ICTRP) was conducted on May 10th, 2020 to identify the study details of the SOLIDARITY and DisCoVeRy trials. A supplementary search of PubMed, WHOs website, French authorities websites, and Google search engine was conducted. A critical review was performed on the findings. ResultsThe DisCoVeRy trial design was detailed consistently in both the European and the US clinical rial registries. SOLIDARITY was registered on ICTRP, with country-specific information reported on country-level registry platforms. The DisCoVeRy trials design appears to be ideal from the methodological perspective. Both trials appear difficult to implement, impractical, and disconnected from the pandemic reality. This is consistent with the apparent failure of the trials to deliver conclusions before the end of the pandemic. ConclusionBoth trials constitute an interesting initiative yet may lack the resources to support a high-quality implementation. The authors call for a pandemic task force, with various experts on the front-line of COVID-19, to inform policy-makers to make effective decisions that may not be based on traditional, methodological state-of-the-art evidence, but rather pragmatic and revisable decisions reflecting emerging evidence for the benefit of patients and society.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.