Attrition Bias

Q: What is the usual threshold to be considered as a high rate of attrition in a Randomized Control Trial (RCT)? 

Answer: 20%

Attrition refers to participants dropping out of a study, which can occur in any medical study. Contrary to popular belief, it's not only the participants leaving the study, but it also includes patients who die or are lost to follow-up during the study. 

The authors are obliged to report the level of attrition so readers and critics can assess the validity and bias (attrition bias), and the effect on statistical power and confidence intervals of the (probably skewed) results. Attrition bias may overestimate the results. There is a high probability that participants who leave the study may differ from those who remain or survive during the study period.

A minor level, i.e., up to 5%, is expected in long-term studies, but >20% is considered high. Attrition may occur due to higher mortality, negative experiences, complicated protocols, side effects, logistical issues, relocations, and the length of the study.

#statistics

References:

1. Dumville JC, Torgerson DJ, Hewitt CE. Reporting attrition in randomised controlled trials. BMJ. 2006 Apr 22;332(7547):969-71. doi: 10.1136/bmj.332.7547.969. PMID: 16627519; PMCID: PMC1444839.

2. Nunan D, Aronson J, Bankhead C. Catalogue of bias: attrition bias. BMJ Evid Based Med. 2018 Feb;23(1):21-22. doi: 10.1136/ebmed-2017-110883. PMID: 29367321.

3. Rees JS, Somi S. A guide to the clinical management of attrition. Br Dent J. 2018 Mar 9;224(5):319-323. doi: 10.1038/sj.bdj.2018.169. Epub 2018 Mar 2. PMID: 29495028.

4. Linardon J. Rates of attrition and engagement in randomized controlled trials of mindfulness apps: Systematic review and meta-analysis. Behav Res Ther. 2023 Nov;170:104421. doi: 10.1016/j.brat.2023.104421. Epub 2023 Oct 14. PMID: 37862854.