Outcome reporting bias in nephrology randomized clinical trials: Examining outcomes represented by graphical illustrations.

BACKGROUND

Outcome reporting bias (ORB) is widely reported in the medical literature, but the contribution from published graphical illustrations is unknown. The aim of this study was to investigate the occurrence of ORB in contemporary nephrology clinical trials relating to the choice of outcomes reported through graphical illustrations.

METHODS

An observational study was conducted using nephrology clinical trials searched from five high-impact medical journals from 2015 to 2020. Eligible trials reported a phase 2, 3 or 4 trial, contained at least one published outcome graphical illustration and were registered on a clinical trial registry. The primary outcome was the occurrence of ORB based on the choice of graphical illustrations in published trial manuscripts, deemed to be present if a graphical illustration displayed a secondary or unregistered outcome ahead of a trial's primary outcome, or if any unregistered trial outcome was presented as a graphical illustration.

RESULTS

In 75 eligible clinical trials, the primary outcome for ORB was present in 60% of the trials (n = 45). Occurrence of the primary outcome did not differ significantly based on trial sample size, funding model, trial phase, individual medical journal or publication year. An unregistered trial outcome was graphically illustrated in 93% (n = 42) of those clinical trials with ORB present.

CONCLUSION

Outcome reporting bias based on the choice of graphical illustration is common, driven primarily by graphical illustration of unregistered trial outcomes. More appropriate choice of outcomes for graphical illustrations by authors, coupled with both increased enforcement of CONSORT guidelines by medical journals and specific guidelines for graphical illustrations choice, are desirable to address these findings.