Screening for frequent hospitalization risk among community-dwelling older adult between 2016 and 2023: machine learning-driven item selection, scoring system development, and prospective validation
Source check: Source checked. Not independent scientific confirmation.
Finding: In 1,611 community-service clients, DeepBoost selected 17 items and achieved AUC 0.90 across operational definitions. A prospective cohort of 329 older adults recruited from a different district during the COVID-19 period validated the score with AUC 0.82 for identifying two or more hospitalizations over two years. Primary Methods/Results document prospective validation in a different district and later time period (329 adults; 13.1% reported at least two hospitalizations over two years). This is a specific clinical retrospective outcome with prospective score validation; no causal treatment effect is claimed.
Technology: DeepBoost selects a 17-item short form from comprehensive geriatric assessment items and derives a risk score for frequent hospitalization among community-dwelling older adults.
Publication: Peer-reviewed paper · Test context: prospective or independent check. Catalog categories, not an evidence grade.
- AI contribution
- DeepBoost selects a 17-item short form from comprehensive geriatric assessment items and derives a risk score for frequent hospitalization among community-dwelling older adults.
- Test and result
- In 1,611 community-service clients, DeepBoost selected 17 items and achieved AUC 0.90 across operational definitions. A prospective cohort of 329 older adults recruited from a different district during the COVID-19 period validated the score with AUC 0.82 for identifying two or more hospitalizations over two years.
- Limitation / still unknown
- Primary Methods/Results document prospective validation in a different district and later time period (329 adults; 13.1% reported at least two hospitalizations over two years). This is a specific clinical retrospective outcome with prospective score validation; no causal treatment effect is claimed.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2024.1413529/full · https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11632619/fullTextXML
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 22 Nov 2026. Review scope: Primary source full text Methods and Results; version/date checked against bibliographic record..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.