aceTackling algorithmic bias and promoting transparency in health datasets: the STANDING Together consensus recommendations
Without careful dissection of the ways in which biases can be encoded into artificial intelligence (AI) health technologies, there is a risk of perpetuating existing health inequalities at scale. One major source of bias is the data that underpins such technologies. The STANDING Together recommendations aim to encourage transparency regarding limitations of health datasets and proactive evaluation of their effect across population groups. Draft recommendation items were informed by a systematic review and stakeh…
STANDING Together consensus recommendations provide guidance for documenting health datasets and for using them to identify and mitigate algorithmic biases, aiming to support AI health technologies that are safe and effective across population groups.
Without careful dissection of how biases are encoded into AI health technologies, underlying health dataset limitations risk perpetuating existing health inequalities at scale.
No health dataset is free of limitations, so absence of transparent communication about limitations should itself be treated as a limitation, and the recommendations are meant to prompt inquiry rather than serve as a checklist.
Evidence
- Peer-reviewedThe Lancet Digital Health2024-12-18
How should this claim be treated?
Truvace Impact Record TRV-2026-0551, v1: “Tackling algorithmic bias and promoting transparency in health datasets: the STANDING Together consensus recommendations.” Truvace, 2026-07-24. /record/TRV-2026-0551 (accessed at citation time). sha256 055dc9f7e98f80e6…
Calibration history
Every change to this record since certification, in the open. None yet — the reading has held since it entered the record.
Certified into the record
How to verify without trusting this page
Fetch the canonical text of any version from /api/record/TRV-2026-0551 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.