Target Trial Emulation in Observational Researches: Clinical Applications, Methodological Gaps, and Recommendations for Design and Implementation
Author(s)
Xin Sun, PhD, Yulong Jia, MSc, Peng Zhao, MSc, Yiquan Xiong, PhD, Yan Ren, PhD, Jing Tan, PhD.
Chinese Evidence-based Medicine Center, West China Hospital, Sichuan University, Chengdu, China.
Chinese Evidence-based Medicine Center, West China Hospital, Sichuan University, Chengdu, China.
OBJECTIVES: Target trial emulation (TTE) is increasingly used to estimate causal effects using real-world data (RWD), but the appropriateness of its clinical application and adherence to key methodological principles remain unclear. We aimed to evaluate the clinical contexts in which TTE is applied, identify methodological gaps in trial design and implementation, and propose recommendations for improving practice.
METHODS: We conducted a comprehensive literature search in PubMed and supplemented it with three existing scoping reviews, focusing on studies published between 2017 and 2023 in Journal Impact Factor Quartile 1 (JIF Q1) clinical journals. Eligible studies explicitly aimed to emulate a target trial using RWD. Screening was supported by a machine learning framework. A standardized extraction tool, developed through expert consensus, captured study objectives, clinical scenarios, target trial design, and adherence to seven core methodological components. Data were extracted in duplicate and verified.
RESULTS: Of 237 included studies, most focused on drug interventions for infectious diseases, cardiology, oncology, or endocrinology. TTE was commonly used to evaluate effectiveness (69.6%), long-term outcomes (13.5%), emergency treatments, or comparative effectiveness (12.7%). While time-varying strategies were frequent (58.6%). Methodological shortcomings were common: 56.5% lacked a prespecified trial protocol; 53.6% did not review existing RCTs; and few justified the type of emulated trial. In implementation, only 14.8% used sequential designs, 33.3% did not emulate randomisation, 83.1% failed to define time zero through a follow-up diagram, and 69.2% did not address unmeasured confounding.
CONCLUSIONS: TTE is widely applied across clinical domains, but many studies fall short in defining target trials and applying core methodological principles. We propose a structured five-step framework and practical considerations to guide future TTE design and implementation. Improved transparency and standardisation are needed to enhance the credibility of causal inference using RWD.
METHODS: We conducted a comprehensive literature search in PubMed and supplemented it with three existing scoping reviews, focusing on studies published between 2017 and 2023 in Journal Impact Factor Quartile 1 (JIF Q1) clinical journals. Eligible studies explicitly aimed to emulate a target trial using RWD. Screening was supported by a machine learning framework. A standardized extraction tool, developed through expert consensus, captured study objectives, clinical scenarios, target trial design, and adherence to seven core methodological components. Data were extracted in duplicate and verified.
RESULTS: Of 237 included studies, most focused on drug interventions for infectious diseases, cardiology, oncology, or endocrinology. TTE was commonly used to evaluate effectiveness (69.6%), long-term outcomes (13.5%), emergency treatments, or comparative effectiveness (12.7%). While time-varying strategies were frequent (58.6%). Methodological shortcomings were common: 56.5% lacked a prespecified trial protocol; 53.6% did not review existing RCTs; and few justified the type of emulated trial. In implementation, only 14.8% used sequential designs, 33.3% did not emulate randomisation, 83.1% failed to define time zero through a follow-up diagram, and 69.2% did not address unmeasured confounding.
CONCLUSIONS: TTE is widely applied across clinical domains, but many studies fall short in defining target trials and applying core methodological principles. We propose a structured five-step framework and practical considerations to guide future TTE design and implementation. Improved transparency and standardisation are needed to enhance the credibility of causal inference using RWD.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
MSR194
Topic
Methodological & Statistical Research, Study Approaches
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
No Additional Disease & Conditions/Specialized Treatment Areas