Marquardsen et al., conducted an analysis of 17 protocols used by commercial companies and determined that the protocols were heavily redacted - which might be a substantial issue regarding the trustworthiness of published drug trials, data analysis and definition of outcomes. They could not find any legitimate rationale for the redactions and discovered the following sections were heavily redacted (in the order of seriousness): primary outcomes, secondary outcomes, other outcomes, adjusted analyses, subgroup analyses, handling of missing data, detections of adverse events, analyses of adverse events, sample size determination, interim analysis and termination of study. Access to the protocols was challenging as the companies took a long time to respond, or never provided the protocols to the research team. Even among the ones provided, they would refer to sections that were redacted or not provided to the researchers. Further, the right to terminate study rested with the commercial companies so, on some occasions, the studies were terminated if the outcomes were not in their favour. The main limitation for the research by Marquardsen et al. is that they did not have full access to the protocols. They could thus not address all the redactions in the protocols as no access to supplementary documents were submitted to the research ethics committees. In conclusion, Marquardsen et al. raise the importance of reproducibility, meticulous checking of results, and identification of bias because, without them, science ceases to exist. They advocate for the industry to become collaborative about everything that matters for patients, doctors and society at large, which include offering unconditional access to its trial protocols.
Link to study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5900834/
- Marina Kwak
Photo credit: Mutant669.