Leveraging Artificial Intelligence to Inform Care Coordination by Identifying and Intervening in Patients' Unmet Social Needs: A Scoping Review.
Victoria H Davis, Andrew D Pinto, Minal R Patel
PMID 40062487WHAT IT FOUND
AI tools in this review mostly flag unmet social needs from records.
Little evidence shows whether care coordination improved, and equity risks remain. Treat outputs as prompts, not proof that a patient will get help.
Key findings
01Eight studies met inclusion criteria; all were based in the United States, and the designs were five retrospective chart reviews, two case studies, and one secondary analysis of a longitudinal study.
02Four of the eight studies described using combined machine learning and natural language processing to identify or classify unmet social needs from existing data.
03Only two of the eight studies examined AI used during care coordination or care management to foster social resources to address unmet social needs.
STILL TO COME
How it was doneWhat they foundWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The review did not appraise study quality, and a single reviewer screened the articles, so relevant studies may have been missed. Only eight studies were included, all from the United States, so the findings may not apply to other health systems. The included studies were retrospective chart reviews, case studies, or a secondary analysis, so they show how AI was used, not whether care coordination improved. One included case study reported that a care platform served 77 individuals, but the review says it is unclear whether this produced clinically meaningful outcomes. Only two of six articles focusing on identification or resource provision explicitly considered co-occurring needs or intersectionality. Many platforms are proprietary, and only two included articles explicitly highlighted the dangers of using AI to identify or inform social care.
Declared interests
The authors declared no conflicts of interest.
The easy way to misread this
Do not read the 21% rehospitalization reduction mentioned in the introduction as a result of this review. It is cited background from another study. The eight included studies are retrospective chart reviews, case studies, or a secondary analysis, and the review says it is unclear whether the care platform had clinically meaningful outcomes.