A target trial emulation evaluating paediatric respiratory support produced a higher estimated risk ratio for liberation from support than the published randomised controlled trial it was benchmarked against. At the median follow-up, the target trial emulation yielded an estimated risk ratio of 1.33 with a 95 percent confidence interval from 1.27 to 1.40, whereas the benchmark randomised controlled trial reported an estimated risk ratio of 1.03 with a 95 percent confidence interval from 0.90 to 1.18.
Target trial emulations offer a structured framework for assessing comparative effectiveness to complement randomised controlled trials, but researchers note that little attention has previously focused on the challenges that arise when emulating a published trial from routine healthcare data. To investigate these challenges, the study evaluated the validity of an emulation by benchmarking its results directly against a published randomised controlled trial and assessing the potential biases stemming from limitations in the emulation process.
Following methodological guidance, the researchers emulated the FIRST-ABC step-up randomised controlled trial using routine data obtained from PICANet. The study evaluated children admitted to paediatric intensive care, comparing two modes of non-invasive respiratory support based on the primary outcome of time to liberation from respiratory support.
The complete-case analysis for the target trial emulation included a cohort of 5,405 children. To address differences in patient characteristics and reduce the risk of confounding inherent in observational data, the investigators applied doubly robust statistical methods.
To evaluate potential biases arising from departures of the emulation from the benchmark trial, the authors conducted a structured risk of bias assessment using the ROBINS-I framework. The assessment applied ROBINS-I domains systematically across each individual component of the target trial framework.
The structured risk of bias assessment identified important limitations in the emulation, including an inability to fully replicate the eligibility criteria and outcome measurement of the randomised controlled trial, as well as the potential for residual unmeasured confounding. The researchers concluded that these findings highlight key challenges in replicating an ideal trial from routine data and offer practical insights to improve the conduct and interpretation of target trial emulations.
Subject of Research: Benchmarking a paediatric target trial emulation against a randomised controlled trial using routine intensive care data
Article Title: Target trial emulation of randomised controlled trials: lessons from benchmarking and risk of bias assessment
Article References: Giallongo, E., Carroll, O., Ramnarayan, P., Mitting, R., Seaton, S. E., Shortt, D., Mason, A. J., Harrison, D. A., & Grieve, R. (2026). Target trial emulation of randomised controlled trials: lessons from benchmarking and risk of bias assessment. BMC Medical Research Methodology. https://doi.org/10.1186/s12874-026-03091-w
Image Credits: AI Generated
DOI: 10.1186/s12874-026-03091-w
Keywords: target trial emulation, randomised controlled trials, FIRST-ABC step-up trial, PICANet, paediatric intensive care, non-invasive respiratory support, liberation from respiratory support, ROBINS-I framework, risk of bias assessment, doubly robust methods, benchmarking, routine healthcare data, comparative effectiveness, residual unmeasured confounding, eligibility criteria replication, complete-case analysis
News Source: Harold Sullivan. (November 13, 2026). Target Trial Emulation Found Divergent Outcomes When Benchmarked Against Paediatric Trial. Scienmag.



