Feasibility Assessment of Real-World Data for Trial Emulations: A Methodological Framework
Author(s)
Olga Volodina, MD, MSc1, Riley Geason, MPH2, Evie Merinopoulou, MSc3, Vartika Savarna, MPP4, Jiaxuan Wang, MSc5, Anupama Vasudevan, MPH, PhD2, Marco Ghiani, PhD4, Jonas Haggstrom, PhD2.
1Cytel, Pincourt, QC, Canada, 2Cytel, Waltham, MA, USA, 3Cytel, London, United Kingdom, 4Cytel, Berlin, Germany, 5Cytel, Vancouver, BC, Canada.
1Cytel, Pincourt, QC, Canada, 2Cytel, Waltham, MA, USA, 3Cytel, London, United Kingdom, 4Cytel, Berlin, Germany, 5Cytel, Vancouver, BC, Canada.
OBJECTIVES: Real-world data (RWD) offer a valuable opportunity to emulate clinical trials, especially in scenarios where placebo-controlled trials are not ethically feasible. To ensure regulatory compliance, data must be assessed for fitness-for-purpose (FFP) by evaluating its reliability and relevance to the study population and research objectives.
METHODS: A structured multi-domain framework is essential to evaluate RWD: 1.Contextual fit assessment: Alignment on the clinical question, disease specification and treatment landscape 2.Data requirements/Trial design translation: Operationalization of key data elements: Patient population, study eligibility, covariates, confounders, endpoints, study period, geographical scope 3.Data source evaluation: Identification and prioritization of RWD sources by conducting targeted literature search. This may include electronic health records(EHR), claims and registries and typically exclude case reports, case series, single cohort studies. 4.Data and Variables Quality assessment: In-depth patient feasibility (ex. assessment of the missingness of the variables/temporal relevance of covariates). 5.Sample size assessment: Final sample size based on the study eligibility criteria 6.Heat map scoring system: Databases are evaluated based on relevance and reliability criteria ranked from 5(highest importance) to 0(lowest importance) to develop a prioritization score. 7.Feasibility report: Tabulate the findings, rank the sources and identify the “FFP” RWD.
RESULTS: We assessed nine RWD sources in the USA that capture administrative claims, EHR or linked claims and EHR data to build a cohort of patients with drug-resistant bacterial infections, focusing on treatment eligibility, endpoints, and patient characteristics. Three datasets met essential criteria, and one was identified as most FFP for comparative effectiveness studies.
CONCLUSIONS: RWD feasibility assessment is a foundational step while designing trial emulations. Employing a structured methodological framework enables data-driven decisions regarding data source selection, trial alignment and analytic strategy. These practices are essential for enhancing the scientific credibility, reproducibility and regulatory acceptance of RWD based evidence generation.
METHODS: A structured multi-domain framework is essential to evaluate RWD: 1.Contextual fit assessment: Alignment on the clinical question, disease specification and treatment landscape 2.Data requirements/Trial design translation: Operationalization of key data elements: Patient population, study eligibility, covariates, confounders, endpoints, study period, geographical scope 3.Data source evaluation: Identification and prioritization of RWD sources by conducting targeted literature search. This may include electronic health records(EHR), claims and registries and typically exclude case reports, case series, single cohort studies. 4.Data and Variables Quality assessment: In-depth patient feasibility (ex. assessment of the missingness of the variables/temporal relevance of covariates). 5.Sample size assessment: Final sample size based on the study eligibility criteria 6.Heat map scoring system: Databases are evaluated based on relevance and reliability criteria ranked from 5(highest importance) to 0(lowest importance) to develop a prioritization score. 7.Feasibility report: Tabulate the findings, rank the sources and identify the “FFP” RWD.
RESULTS: We assessed nine RWD sources in the USA that capture administrative claims, EHR or linked claims and EHR data to build a cohort of patients with drug-resistant bacterial infections, focusing on treatment eligibility, endpoints, and patient characteristics. Three datasets met essential criteria, and one was identified as most FFP for comparative effectiveness studies.
CONCLUSIONS: RWD feasibility assessment is a foundational step while designing trial emulations. Employing a structured methodological framework enables data-driven decisions regarding data source selection, trial alignment and analytic strategy. These practices are essential for enhancing the scientific credibility, reproducibility and regulatory acceptance of RWD based evidence generation.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
RWD84
Topic
Methodological & Statistical Research, Real World Data & Information Systems, Study Approaches
Disease
Infectious Disease (non-vaccine), No Additional Disease & Conditions/Specialized Treatment Areas