Differential dropout and bias in randomised controlled trials: when it matters and when it may not

Abstract

Dropout in randomised controlled trials is common and threatens the validity of results, as completers may differ from people who drop out. Differing dropout rates between treatment arms is sometimes called differential dropout or attrition. Although differential dropout can bias results, it does not always do so. Similarly, equal dropout may or may not lead to biased results. Depending on the type of missingness and the analysis used, one can get a biased estimate of the treatment effect with equal dropout rates and an unbiased estimate with unequal dropout rates. We reinforce this point with data from a randomised controlled trial in patients with renal cancer and a simulation study.

Improving Quality of Life and Depression After Stroke Through Telerehabilitation

“…For the purposes of this intent-to-treat analysis, data were assumed to be missing at random. The mixed analysis approach was used, with no ad hoc imputation (Bell, Kenward, Fairclough, & Horton, 2013;Chakraborty & Gu, 2009). The estimate of primary interest was the time (baseline, EOT) · treatment (HEP, robot 1 HEP) interaction.…”

Reducing Distress in Mothers of Children With Autism and Other Disabilities: A Randomized Trial

“…Follow-up 232 ANOVAs (MBSR versus PAD by dropouts versus completers) revealed no significant main or interaction effects in baseline assessments or demographics, indicating no apparent differential dropout biases. 41 Baseline Characteristics …”

“…Parents in our study may not represent the population of all parents of children with developmental disabilities, but they were more diverse in age, race, and income than those typically included in MBSR or positive psychology interventions. 21,41 Given the number of study outcomes, significant findings with smaller effect sizes could be due to chance and thus need to be replicated in future work. A final limitation, inherent in "real-world" effectiveness studies, is dropout owing to life circumstances, and in our case to caregiving demands, and dwindling questionnaire compliance during follow-up.…”

Effect of Clinical Nurses Specialists intervention on rehabilitation outcomes in patients with heart failure

“…[24,62] In this study the dropout rate was 17.3%. The drop out analysis showed, that there were no differences between patients who dropped out of the study and patients who completed the study in relation to demographic data.…”