Methods for Estimating Relative Risks and Risk Differences in Trials
Author Information
Author(s): Thompson Jacqueline J.Y., Watson Samuel I., Middleton Lee, Hemming Karla
Primary Institution: Department of Applied Health Sciences, College of Medicine and Health, University of Birmingham
Hypothesis
What methods are used to estimate covariate-adjusted relative risks and risk differences in randomized controlled trials?
Conclusion
Approximately half of the RCTs report either a covariate-adjusted relative risk or risk difference, but many lack adequate details on the methods used.
Supporting Evidence
- 49% of RCTs reported a covariate-adjusted relative risk or risk difference.
- 65% of studies used the log-binomial model for adjusted relative risks.
- 48% of studies used the binomial model for adjusted risk differences.
Takeaway
This study looked at how often trials report adjusted risks and found that about half do, but many don't explain how they calculated them.
Methodology
Systematic review of two-arm parallel RCTs published in high-impact journals from January 1, 2018, to March 11, 2023.
Potential Biases
Potential bias due to unclear reporting of methods and selection of covariates.
Limitations
The review only included trials from high-impact journals and may not represent all RCTs.
Participant Demographics
Most trials were multi-centre with an average sample size of 357 in the control arm and 367 in the intervention arm.
Statistical Information
Confidence Interval
95% CI: 43–54%
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website