Truace
Subscribe

Detecting Language-Related Disparities in Spiritual Care Using Artificial Intelligence

Introduction: Language barriers perpetuate health care disparities, yet traditional data sources overlook inequities in relationship‑based care. This study uses artificial intelligence to examine language‑related disparities in spiritual care, a domain largely documented in unstructured clinical notes. Methods: We analyzed electronic health records of 7,360 deceased adults aged ≥65 years from a tertiary hospital (2010-2023). Natural language processing identified spiritual care documentation, and structured flow…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-2091· cite & verify
Detecting Language-Related Disparities in Spiritual Care Using Artificial Intelligence
WHSF- Returned, 67-12 - DPLA - 3f4bc2b9476a1d46e1e7400ac8576f5e by Richard Nixon Library & Birthplace (1990 - 7/11/2007). Public domain

In brief

Researchers reviewed electronic health records of 7,360 adults aged 65 and older who died at one tertiary hospital between 2010 and 2023. Natural language processing found spiritual care documentation in clinical notes, while structured flowsheets recorded formal chaplain visits. The team then compared access and timing by patient language.

The findings suggest that inequities in relationship-based care can stay hidden when only structured data are examined, and that text analysis can make them visible. It remains uncertain how far the results apply beyond this hospital and patient group, and the abstract does not say why the gaps arose or whether fixing them changes patient outcomes.

Main points

  1. The sample was 7,360 deceased adults aged 65 or older from a tertiary hospital, with records spanning 2010 to 2023.
  2. Non-English speakers had an odds ratio of 0.41 (p = .003) for receiving formal spiritual care.
  3. The median time to support was 136 days for non-English speakers versus 432 days for others (p = .020).
  4. Among non-English speakers, providers other than chaplains more often first documented spiritual needs.

The gain

Natural language processing of unstructured clinical notes found language-based inequities in spiritual care that structured data alone did not capture.

The rundown

The study combined two data sources. Natural language processing identified spiritual care documentation in unstructured clinical notes, and structured flowsheets captured formal chaplain visits. Logistic regression then tested links between patient language and both access to and timing of spiritual care.

Non-English speakers had markedly lower odds of formal spiritual care and received it nearly ten months later at the median. The authors report that, for these patients, spiritual needs were more often first noted by providers other than chaplains, which points to how needs are identified before a chaplain gets involved.

The authors conclude that AI-supported analysis of unstructured documentation revealed disparities that structured data missed. They describe it as a culturally grounded way to identify inequities in relationship-based care.

Sources

  1. Peer-reviewedJournal of Transcultural Nursing2026-10-20

The debate