What does this research mean for the field?

Data-driven estimation of between-trial heterogeneity in Bayesian meta-analytic predictive models substantially increases the effective sample size derived from historical controls without altering treatment effect estimates, outperforming default conservative choices. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To assess the feasibility of constructing historical controls using Bayesian meta-analytic predictive priors derived from systematic review data in type 2 diabetes mellitus.

August 30, 2026Clinical TrialsOpen Access

A practical application for the efficient use of historical data in randomized controlled design using systematic review data

View Full Paper

Ask AI

Bookmark

Authors

TSTim J. Schouten AGAndrea Gabrio JCJochen W. L. Cals

Discussion

Member takes

Overview

Meta-analytic modeling demonstrates that data-derived heterogeneity estimates increase historical control information borrowing in diabetes trials, indicating a practical route to reduce sample sizes.

Key Points

To assess the feasibility of constructing historical controls using Bayesian meta-analytic predictive priors derived from systematic review data in type 2 diabetes mellitus.
Extracted baseline and end point glycated hemoglobin values from 25 randomized trials comprising 36 comparisons of metformin.
Constructed meta-analytic predictive priors using the RBesT package and quantified data contribution using the effective sample size metric.
Evaluated a default conservative between-trial heterogeneity prior (τ = 0.5) against empirically derived τ values from the metafor package, with sensitivity analyses stratified by follow-up duration and publication year.
Initial models with high between-trial heterogeneity (τ = 0.55–0.85) yielded low effective sample size values of 1–4, showing negligible information borrowing from historical trial data.
Excluding four outlier studies lowered the data-derived τ to 0.0935 and increased the effective sample size from 18 to 52, while preserving stable treatment effect estimates of −1.34% to −1.36% glycated hemoglobin.

Cite This Study

Schouten et al. (2026) studied this question.

synapsesocial.com/papers/6a93f1726c1a8fb52e79e563 https://doi.org/10.1177/17407745261467982

View Full Paper

Ask AI

Bookmark

August 30, 2026Clinical TrialsOpen Access

A practical application for the efficient use of historical data in randomized controlled design using systematic review data

View Full Paper

Ask AI

Bookmark

Authors

TSTim J. Schouten AGAndrea Gabrio JCJochen W. L. Cals

Discussion

Member takes

Overview

Key Points

To assess the feasibility of constructing historical controls using Bayesian meta-analytic predictive priors derived from systematic review data in type 2 diabetes mellitus.
Extracted baseline and end point glycated hemoglobin values from 25 randomized trials comprising 36 comparisons of metformin.
Constructed meta-analytic predictive priors using the RBesT package and quantified data contribution using the effective sample size metric.
Evaluated a default conservative between-trial heterogeneity prior (τ = 0.5) against empirically derived τ values from the metafor package, with sensitivity analyses stratified by follow-up duration and publication year.
Initial models with high between-trial heterogeneity (τ = 0.55–0.85) yielded low effective sample size values of 1–4, showing negligible information borrowing from historical trial data.
Excluding four outlier studies lowered the data-derived τ to 0.0935 and increased the effective sample size from 18 to 52, while preserving stable treatment effect estimates of −1.34% to −1.36% glycated hemoglobin.

Cite This Study

Schouten et al. (2026) studied this question.

synapsesocial.com/papers/6a93f1726c1a8fb52e79e563 https://doi.org/10.1177/17407745261467982

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark