ADL Trajectory Clusters for 265,530 U.S. Veterans Affairs Nursing Home Residents

Thirteen distinct functional trajectory clusters derived from 1.3 million Activities of Daily Living assessments. The dataset was created by Ghazal Khalili and published on figshare in May 2026. It results from a hybrid trajectory clustering framework applied to longitudinal data from U.S. Veterans Affairs nursing homes.

Use Cases

• Identify high-risk patient phenotypes for targeted resource allocation based on trajectory clusters.
• Study associations between functional trajectory patterns and mortality risk based on Kaplan-Meier and Cox model results.
• Benchmark new trajectory analytics methods against the described hybrid clustering framework.
• Inform personalized care planning in long-term care settings based on interpretable trajectory phenotypes.

Strengths

• Derived from a substantial underlying dataset of 1.3 million ADL assessments across 265,530 residents.
• Clusters are clinically interpretable, differing in dominant disability states, duration, and recurrence.
• Mortality risk associations are quantified, with the highest-risk cluster showing approximately 48% cumulative mortality within the first year.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the final cluster set is unknown, which may limit suitability assessment.
• Data is specific to U.S. Veterans Affairs nursing home residents and may not generalize to other populations.

Provenance

Source

U.S. Veterans Affairs nursing homes.

Collection Method

A hybrid trajectory clustering framework combining spell-based sequence construction, optimal-matching-based dissimilarity, and scalable quality-guided clustering.

Freshness

Last updated 2026-05-27 17:50:57; freshness should be verified.

Geography

United States