Mapping eligibility criteria in oncology target trial emulations using real-world data: a scoping review

This scoping review systematically characterized how eligibility criteria from oncology clinical trials are mapped to real-world data in target trial emulations, analyzing 74 individual emulations across 47 studies published between 2016 and 2025. Among emulations of actual clinical trials, only a median of 35.5% of eligibility criteria were successfully mapped to available real-world data, compared to 91.7% in emulations of hypothetical trials, which tend to specify far fewer and simpler criteria by design. Demographic criteria were most frequently mapped at 93.9%, while safety and concomitant risk-related criteria were mapped in only 3.7% of cases, and the dominant mapping method across studies was direct mapping using administrative coding systems, with no studies reporting the use of common data models or computational methods.

These findings expose a significant and underappreciated limitation of real-world evidence in oncology: when target trial emulations can only represent a fraction of the eligibility criteria that define the intended study population, the resulting cohorts may systematically differ from the populations they are meant to approximate, with meaningful consequences for the validity and comparability of findings. The review calls for greater transparency in how mapping decisions are reported and for the development of more sophisticated computational methods that can move beyond administrative codes to capture the full complexity of clinical trial eligibility in real-world datasets.