Data for Long-Term Care Research

Why It Matters

Robust, linked long‑term‑care data unlocks evidence‑based policy and investment decisions, while data access hurdles risk delaying solutions to a sector serving millions of vulnerable Americans.

Key Takeaways

• Public datasets like MDS and Medicare claims enable resident-level analysis.
• Payroll‑Based Journal (PBJ) provides auditable, daily staffing data for nursing homes.
• Assisted‑living research lacks standardized data; nine‑digit ZIP method offers workaround.
• Linking multiple sources creates Residential History Files for comprehensive longitudinal studies.
• Data access barriers and outdated DUAs hinder timely long‑term‑care research.

Summary

The meeting opened by announcing an October 2 in‑person conference on long‑term‑care data and invited paper submissions, underscoring the field’s growing research appetite. Speakers highlighted the core data ecosystems: the Minimum Data Set (MDS) for resident assessments, Medicare fee‑for‑service and Advantage claims, and the newer Payroll‑Based Journal (PBJ) that captures daily, auditable staffing information across every U.S. nursing home. Key insights included the power of linking MDS with claims to build Residential History Files, enabling precise tracking of resident trajectories. For assisted‑living facilities, where no MDS exists, researchers have devised a nine‑digit ZIP‑code matching technique to identify large facilities within claims data. The PBJ revealed stark staffing gaps—average weekend staffing below weekday levels and turnover rates exceeding 100% in many homes—information unattainable from older survey‑based sources. Notable examples featured David Gowski’s citation of the Brown team’s residential history algorithm and the 2019 Fang Leong paper demonstrating PBJ’s ability to detect systematic staffing inflation around survey windows. Another striking finding was the distribution of staff turnover, with some facilities reporting annual rates of 200‑300%, highlighting workforce instability. The discussion concluded that richer, linked datasets open new avenues for evaluating Medicare Advantage versus traditional Medicare, and for assessing Medicaid services across home‑based and institutional settings. However, researchers urged CMS to streamline data use agreements, expand access to newer variables, and improve the quality of facility‑reported measures, warning that current barriers could stall critical policy‑relevant research.