HearShare: a multi-institution effort to make audiometric care data analyzable.

Audiometric data are collected at scale during routine clinical care, yet they remain difficult to study at population scale. Local schemas differ, vendor exports are inconsistent, key fields are buried in free text, and important context (test conditions, masking, transducers) is often documented heterogeneously across centers.

These barriers have historically limited research questions to single-site cohorts or hand-curated subsets, leaving fundamental questions about hearing-loss trajectories, normative references across the lifespan, and treatment effectiveness in real-world settings underpowered or unaddressed.

HearShare was assembled to change the substrate. Five academic medical centers have committed to contribute audiometric, demographic, and clinical data through a shared data-use framework. The plan is to harmonize those contributions into a single ontology-mapped record (HARK) hosted on the Brain Data Science Platform, alongside the analytics environment investigators would use to study it.

The resource is at the design stage. The schema, mappings, governance procedures, and access tiers shown here are proposals iterated with input from contributors and the broader research community; harmonized data have not yet been ingested. Where the work is provisional, we mark it as such on the data and access pages.