FemHealth is committed to transparency about how it works, where data comes from, and what the platform cannot do.
Live Databases
PubMed – NCBI E-utilities API. 35M+ citations. Medical literature indexed by the US National Library of Medicine.
Europe PMC – EMBL-EBI API. European biosciences literature, including preprints.
OpenAlex – Open scholarly graph. 250M+ works across all disciplines.
Gender Bias Detection – What Works
Abstract-based extraction works well when studies report participant sex explicitly (common in RCTs and systematic reviews). Pattern matching identifies phrases like "120 female participants", "n=80 women", and explicit exclusion statements.
Known Limitations
Older studies (pre-2000) rarely report sex breakdown in abstracts - extraction rate is low.
Some journals use "sex" and "gender" interchangeably; FemHealth flags both.
Studies not reporting sex breakdown show "data unavailable" - not equivalent to balanced data.
Abstracts without explicit counts cannot be analysed; this is common in qualitative research.
Traditional medicine database coverage is limited and opt-in only.
No Mock Data
All search results are retrieved live from real APIs. FemHealth does not generate, simulate, or interpolate research data.