Assessing the impact of attrition in randomized controlled trials

Prior notification of trial participants by newsletter increased response rates: a randomized controlled trial

“…Many, if not most, RCTs suffer some element of attrition. In particular, trials that rely on selfcompleted outcome measures, often delivered by post, from patients can have high levels of attrition and a 20% loss to follow-up in such trials is not uncommon [2]. Consequently, it is crucial that attrition is kept to a minimum.…”

Retrospective analysis of a South African cardiovascular trial site’s patient retention rates

“…4,6 However, it is important to state that discontinuation rates may vary depending on the therapeutic area, investigational drug, inclusion/exclusion criteria, and patient characteristics. 6 Death was the predominant reason for discontinuation (39.6%), as reported by this analysis.…”

“…2 High patient attrition rates also pose a risk to the interpretation and validity of the intended research findings. [2][3][4] Furthermore, the cost to recruit a patient is significantly higher than to retain an existing patient in a clinical trial. 5 Reported patient attrition rates range from 15% to 40% of enrolled subjects depending on the therapeutic area, investigational drug, inclusion/exclusion criteria, and patient characteristics.…”

“…5 Reported patient attrition rates range from 15% to 40% of enrolled subjects depending on the therapeutic area, investigational drug, inclusion/exclusion criteria, and patient characteristics. 4,6 Disappointing statistics show that as many as 26% of patients drop out after providing consent. 7,8 A recent study conducted at a Texas medical school retrospectively analyzed 541 randomly selected charts from clinical trials conducted between 2000 and 2006.…”

Addressing missing participant outcome data in dental clinical trials

“…With proper randomization any imbalances of known and unknown prognostic factors between treatment arms are attributed simply to chance. [1][2][3][4] In addition, implementation of blinding, where applicable, is likely to reduce performance, detection and attrition bias. There is empirical evidence that implementation of blinding is associated with improved adherence to the assigned interventions and reduced losses to follow-up compared to unblinded trials.…”