Analyses of repeatedly measured continuous outcomes in randomized controlled trials needed substantial improvements

Abstract

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Minimizing error in estimates of the effect of interventions by accounting for baseline measurements: A simulation study analyzing effects on child growth

“…Despite the benefits to ADJUST, DELTA+ADJUST and RESIDUALS, linear growth randomized trials have frequently not taken advantage of these methods (see example of Mangani et al, 2015 and Hill et al, 2020 , which used the DELTA method, and Maleta et al, 2004 and DeBoer et al, 2021 , which used the FINAL method; Ren et al, 2022 ). On face value, each of the five analytic methods, FINAL, ADJUST, DELTA, DELTA+ADJUST and RESIDUALS, appear as reasonable options.…”

“…Ren et al, 2022). On face value, each of the five analytic methods, FINAL, ADJUST, DELTA, DELTA+ADJUST and RESIDUALS, appear as reasonable options.…”

Efficacy of models of palliative care delivered beyond the traditional physician-led, subspecialty consultation service model: a systematic review and meta-analysis

“…Fourth, the quality of included studies was variable, with many at a high RoB. One particular issue was the analysis and reporting of repeated measures, which appears to be common across RCTs in general 67. Many studies suffered from missing data at follow-up among the surviving patients and, as expected in trials with seriously ill patients, mortality rates at 3 and 6 months was high, posing additional methodological challenges (eg, loss to follow-up).…”

Missing data were poorly reported and handled in randomized controlled trials with repeatedly measured continuous outcomes: a cross-sectional survey

“…This study was part of a larger initiative to investigate the statistical methods of analyzing the repeated measure data [22]. We included a study, if it was an RCT with a repeatedly measured continuous outcome as the primary outcome in the abstract, methods, or results.…”

“…The sample size was 200 studies. We estimated this sample size as that needed to achieve the desired confidence interval (0.43, 0.57) given the most conservative proportion (0.5) of RCTs which used advanced methods regarding repeatedly measured continuous outcomes, as described in the previous article [22]. This sample size could achieve a sufficient power when assessing missing data practices, and was larger than that in many similar studies [3,4,17].…”