Reporting of Rurality, Deprivation, and Socioeconomic Characteristics in Cancer Clinical Trials: A Systematic Review.

Abstract

The composition of clinical trial populations influences the interpretation and generalizability of cancer trial results. The extent to which geographic and socioeconomic characteristics are reported in primary cancer clinical trial publications is unknown. To evaluate the frequency and temporal trends in reporting of geographic, socioeconomic, and demographic characteristics in primary publications of US-based randomized cancer clinical trials. This systematic review included US-based randomized phase II or III cancer treatment trials published in high-impact peer-reviewed journals from 2020 to 2025. Using prespecified search strategies and eligibility criteria, primary trial reports were identified. Data were abstracted on reporting of rurality, area-level deprivation, insurance status, income, education, age, sex, race, and ethnicity. Overall prevalence and temporal trends in reporting of geographic, socioeconomic, and demographic characteristics. Among 6681 reviewed trials, 441 met inclusion criteria. Reporting of age and sex was nearly universal. Reporting of race and ethnicity was common, with 365 trials (82.8%) reporting more than 1 category. The most frequently reported race variables were White (356 [80.7%]), Asian (330 [74.8%]), and Black (320 [72.6%]); Hispanic ethnicity was reported in 140 trials (31.8%). Reporting of any race or ethnicity increased from 62.3% (38 of 62) in 2020 to 93.9% (77 of 82) in 2025. In contrast, no trials reported rurality, income, or area deprivation; 2 reported education, and 1 reported insurance. In this systematic review of 441 US-based randomized cancer clinical trials, geographic and socioeconomic characteristics were rarely reported despite increasing attention to social determinants of health. Limited reporting of these characteristics may constrain assessment of trial generalizability and interpretation across diverse populations and care settings.