Nursing-record features improve estimation of home-care use
Source check: Source checked. Not independent scientific confirmation.
Finding: Random-forest analysis of 4,323 Dutch home-care clients found that recorded nursing diagnoses and symptoms added information beyond existing case-mix groups. Reported internal cross-validated R-squared for weekly care hours increased from 22.4% to 32.1% when nursing-record features were added. The result concerns observed service use at one provider, not objectively necessary care or improved payment policy. Documentation differences and underprediction of high-use clients limit generalization.
Technology: Random forest; recursive feature elimination
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Forest ensembles learn interactions among demographic, case-mix and nursing-assessment variables to estimate recorded care-hour use; comparing feature sets measures the information added by nursing records.
- Test and result
- One randomly selected assessment per client was retained to prevent repeated-client leakage. Forest models were compared using ten-fold validation against average weekly care hours in the 28 days after assessment, with demographic/case-mix and nursing-record feature sets.
- Limitation / still unknown
- Performance at other providers, clinically appropriate care need and reimbursement-policy effects remain unestablished. The primary body describes feature elimination but its exact nesting/tuning details are in an unreviewed supplement. Care hours combine nursing and personal care, and documentation practices can vary.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC7561562/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Primary relevant client selection, outcome and forest/validation Methods, Results and limitations; supplementary parameter-tuning and feature-elimination details not independently reviewed. Exact primary identity/date and fresh primary PubMed correction/retraction relationships checked..
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 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.