Comparison of Registered and Reported Outcomes in Randomized Clinical Trials Published in Anesthesiology Journals

Abstract

Despite trial registration being an accepted best practice, RCTs published in anesthesiology journals have a high rate of inadequate registration. While mandating trial registration has increased the proportion of adequately registered trials over time, there is still an unacceptably high proportion of inadequately registered RCTs. Among adequately registered trials, there are high rates of discrepancies between registered and reported outcomes, suggesting a need to compare a published RCT with its trial registry entry to be able to fully assess the quality of the study. If clinicians base their decisions on evidence distorted by primary outcome switching, patient care could be negatively affected.

Ensuring the quality and specificity of preregistrations

“…For example, if a researcher does not state anything about the handling of outliers, it may imply they will not remove any outlying values or that they will consider what to do with outliers after finding them, which means that the RDF is not restricted. Furthermore, studies on registrations of clinical trials show that other RDFs like outcome switching and HARKing are common in publications of preregistered studies (Chan, 2008;Chan & Altman, 2005;Chan, Hrobjartsson, Haahr, Gotzsche, & Altman, 2004;Chan, Hróbjartsson, Jørgensen, Gøtzsche, & Altman, 2008;Goldacre, 2016;Ioannidis, Caplan, & Dal-Ré, 2017;Jones et al, 2017;Lancee, Lemmens, Kahn, Vinkers, & Luykx, 2017;Mayo-Wilson et al, 2017;Rankin et al, 2017;Wayant et al, 2017). It is possible that increasing the explicit commitment to exhaustiveness will reduce this tendency.…”

Open season: positive changes for increased transparency in the biomedical literature

“…9 Clinical trial registration has been an expected practice for many years, yet unfortunately there is evidence that many anesthesia researchers still treat trial registration as an afterthought -often registering after recruitment has begun or even once recruitment has finished. 3 Registering an RCT after authors have access to patient data misses the point of pre-specifying interventions and outcomes before potential biases, such as detecting an early signal of effect in a partially completed RCT, have crept in. Currently, there is a very clear expectation for authors to adequately register RCTs before the recruitment of the first patient into the trial.…”

“…These elements have all remained private information-known only to the researchers and exposed to little external scrutiny or verification. This lack of transparency has facilitated selective trial reporting (also known as publication bias), 1,2 selective outcome reporting, 3 mistakes in statistical analyses, 4 and the conduct of suspected 5 or known 6 fraudulent studies. These factors represent threats to the goal of basing our clinical care on the highest quality evidence.…”

Publication Challenges, Past, Present, and Future

“…This has met with mixed results, with evidence suggesting that even after registering trials, authors frequently change the primary and secondary published outcomes. 7 Although there is ample evidence in the literature documenting publication bias in favor of positive trials, the impact of this bias has been harder to quantify, with only a few trials attempting to identify the magnitude of the effect on guidelines or meta-analyses. An article summarizing 7 methodological studies examining the impact of the addition of gray literature (unpublished data typically from registries or government reports) and treatment effects was able to show a reduction in the pooled estimate in 2 of the methodological studies and no significant effect in 5.…”