Misinformation in COVID-19: nimble innovation within Cochrane to appraise the source

Article type
Authors
Fox T1, Walshe D1, Kuehn R1, Wiebel S2, Popp M2, Singh B3, Fletcher T1, Garner P1
1Liverpool School of Tropical Medicine
2University Hospital Würzburg
3University of Liverpool
Abstract
Background:
During COVID-19, the public was put in danger by erroneous beliefs derived from poor-quality research amplified through social media.

In the Cochrane Infectious Diseases Group (CIDG), we needed to be innovative with methods to address COVID-19 misinformation. Here, we will present this evolving response, leading to an approach to appraise research that caused unsubstantiated hypotheses. In this process, we adapted Cochrane methods to appraise comparative serological studies.

Objectives:
1. To present the CIDG strategy to address the COVID-19 misinformation pandemic.
2. To outline novel methods for critically appraising comparative serological studies.

Methods:
The CIDG identified important areas of misinformation where our methods could provide an authoritative voice. This started with Cochrane Reviews of trials to help end the use of unapproved drugs for COVID-19 (chloroquine and ivermectin). Later, we identified a potentially misinformed hypothesis derived from laboratory research (“microclots” cause the post-COVID-19 syndrome [PCS]) leading to public demand for the treatment plasmapheresis. Because there were no randomized controlled trials, we rigorously appraised these laboratory studies using Cochrane principles in collaboration with laboratory scientists and thrombotic disorder specialists.

Results:
Cochrane Reviews of trials of chloroquine and ivermectin challenged fake and poorly conducted science in the public narrative; some trials were excluded from the ivermectin review based on a failure to fulfill the expected ethical and scientific criteria. These reviews were highly cited and used as the gold standard to state there was no evidence these treatments worked. Conducting a Cochrane Review of plasmapheresis for PCS was problematic, as trials had not been conducted and would be expensive. So we used novel methods to appraise the comparative serological studies that generated the hypothesis that microclots were associated with PCS. We identified bias across three domains: collection and handling of samples; experimental methodology; and reporting and interpretation of results.

Conclusions:
Sometimes policy needs systematic reviews that evaluate the rationale for a treatment to help justify whether a trial is worth doing. These can be laboratory studies with sera or animal studies. Cochrane principles and methods can be adapted to meet this need, but flexibility within Cochrane formats and editorial policies could facilitate this.