According to a new analysis, only 29.1% of current COVID-19 studies have sufficient methodological strength to produce high quality evidence.
The researchers from Stanford University, USA, examined the characteristics and expected strength of evidence of all currently registered studies on ClinicalTrials.gov, in terms of the 2011 Oxford Centre for Evidence-based Medicine (OCEBM) level of evidence framework.
A total of 1551 studies identified met the inclusion criteria; 59% of which were interventional trials (including 664 randomised trials) and 41% were observational studies. However, from the total, only 451 studies (29.1%) could potentially yield OCEBM level 2 evidence, which is the highest level of individual study evidence. For example, across the 664 randomised controlled trials (RCTs), blinding (required for a level 2 study) was reported in only 364 and only 74 RCTs were placebo-controlled. For the 640 observational studies, 81% were single-centred and only 87 (13.6%) were of sufficient quality to yield level 2 OCEBM evidence.
The authors expressed concern that given the huge number of studies being made available via preprint (that is, prior to peer-review), the rapid dissemination of low quality evidence may influence public opinion and clinical practice in ways that were potentially harmful, especially at a time when high quality knowledge is critically needed.
Pundi K et al. Characteristics and strength of evidence of COVID-19 studies registered on ClinicalTrials.gov. JAMA Int Med 2020; 27 Jul doi:10.1001/jamainternmed.2020.2904