Predicting Post‐Discharge Frailty in Older Adults With a Hip Fracture: An Analysis Using the Decision Tree Model.
Purpose: To compare the predictive performance of a decision tree analysis against that of a logistic regression in identifying the frailty status 3 months post‐discharge among older adults hospitalized for a hip fracture. Methods: A prospective, longitudinal study was conducted at two medical cente...
| Published in: | Health & Social Care in the Community Vol. 2026; pp. 1 - 15 |
|---|---|
| Main Authors: | , , , , |
| Format: | algorithm research tables/charts Journal Article |
| Published: |
Wiley-Blackwell
6/25/2026
|
| Online Access: | View this record in EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=194870497&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 194870497 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 09660410 EVX jtl: Health & Social Care in the Community issn: 09660410 maglogo: Y pubinfo: dt: 6/25/2026 vid: 2026 pid: 480 pub: Wiley-Blackwell place: Malden, Massachusetts artinfo: ui: 194870497 194870497 194870497 10.1155/hsc/1969735 194870497 ppf: 1 ppct: 14 formats: fmt: – @attributes: type: T – @attributes: type: C – @attributes: type: P tig: atl: Predicting Post‐Discharge Frailty in Older Adults With a Hip Fracture: An Analysis Using the Decision Tree Model. aug: au: Chen, I-Hui Yeh, Tzu-Pei Hashimoto, Mitsuo Lin, Yen-Kuang Chen, Qing-Wei affil: School of Nursing,, College of Nursing,, Taipei Medical University,, 250 Wuxing St, Xinyi District, Taipei, 11031,, Taiwan, tmu.edu.tw sug: subj: Hip Fractures In Old Age Frailty Syndrome Diagnosis Decision Trees Hospitalization of Older Persons Patient Discharge Logistic Regression Human Taiwan Female Male Middle Age Aged Prospective Studies Convenience Sample Academic Medical Centers Weight Loss Walking Speed Self Report Physical Activity Scales Grip Strength Dynamometry Geriatric Depression Scale Psychological Tests Sleep Quality Serum Albumin Interviews Questionnaires Calcium Blood Aged, 80 and Over Cognition Confidence Intervals Sensitivity and Specificity ROC Curve Odds Ratio Data Analysis Software Descriptive Statistics Middle Aged: 45-64 years Aged: 65+ years Aged, 80 & over Female Male ab: Purpose: To compare the predictive performance of a decision tree analysis against that of a logistic regression in identifying the frailty status 3 months post‐discharge among older adults hospitalized for a hip fracture. Methods: A prospective, longitudinal study was conducted at two medical centers in northern Taiwan between April 2022 and February 2023. Participants (N = 238) were recruited from orthopedic wards using convenience sampling. The study examined various admission‐based predictors, including physiological, psychological, and social factors, along with the initial frailty status. Results: The decision tree model achieved 96.2% accuracy, with 95.5% sensitivity and 98.3% specificity in predicting the frailty status; across 100 repeated stratified 70:30 hold‐out splits, mean AUROC was 0.829 (SD = 0.059) for the decision tree and 0.890 (SD = 0.041) for logistic regression, and the difference between models was not statistically significant (DeLong p = 0.906). Depression emerged as the root node, with a threshold Geriatric Depression Scale (GDS)‐15 score of ≥ 2.5, which was significantly associated with frailty, followed by the St. Louis University Mental Status (SLUMS) score, grip strength, 25‐hydroxyvitamin D level, age, frailty status at admission, sleep quality, serum albumin, and pain as additional predictive factors. The multivariate logistic regression analysis identified several significant predictors of frailty at 3 months post‐discharge: female sex (odds ratio (OR) = 0.04, p = 0.004), grip strength (OR = 0.85, p = 0.048), serum albumin (OR = 0.11, p = 0.045), a higher depression score (OR = 2.18, p = 0.009), SLUMS cognitive function score (OR = 0.63, p = 0.003), a higher sleep quality score (OR = 1.93, p = 0.004), and frailty at admission (OR = 80.87, p = 0.002). Conclusions: Both predictive models demonstrated strong performances in identifying post‐discharge frailty risk, with the decision tree analysis offering particularly high accuracy. The findings suggest that multiple factors, especially the psychological status at admission, play crucial roles in the development of post‐discharge frailty among older adults with a hip fracture. External validation of the identified thresholds in independent multicenter cohorts is required before clinical application. pubtype: Academic Journal doctype: algorithm research tables/charts Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|