Internal Medicine

Pivotal device trials report demographics but rarely integrate equity

Seventy-four pivotal device studies were reviewed; most (68/74, 91.9%) were randomized trials. Age and sex were almost always reported, but only 18.9% performed age subgroup analyses and 14.8% did sex analyses. Race/ethnicity appeared in 35.1% of trials and PROGRESS-Plus variables in 9.5%, with only 2.7% explicitly framing equity. No study applied CONSORT‑Equity, and none reported population benchmarking or representativeness analyses, limiting external validity. 1

AI trial consent documents often lack key disclosures and participant control

Content analysis of 114 AI-involved trials found 55% of informed consents omitted AI type or usage. Risk information was missing in 16.4%, and only 14% met combined brevity and readability criteria. Data-handling after withdrawal varied: 51 provided no information, 29 allowed continued use, and only 3.5% offered participant choice. Authors propose minimum consent requirements but note current practices undermine transparency and participant control. 2

Women underrepresented in lung cancer RCTs; female enrollment linked to AE reporting

Systematic review of 636 RCTs (265,989 participants) found a pooled enrolment incidence disparity (EID) of -3.17% for women. Underrepresentation was pronounced in very high‑HDI regions and select subgroups like younger and predominantly Caucasian populations. Trial-level differences in female enrolment were associated with differences in reported adverse event incidence. Authors conclude underrepresentation may limit generalisability and call for sex-disaggregated safety reporting. 3