AI Science Tracker
EN

Exploration of the correlation between clinical indicators and prognosis in hospitalized children with pneumonia and construction of a risk prediction model based on machine learning algorithms.

School of Medicine, Shaanxi University of International Trade and Commerce · 28 Jan 2026 · medicine

Discovery

Exploration of the correlation between clinical indicators and prognosis in hospitalized children with pneumonia and construction of a risk prediction model based on machine learning algorithms.

Source check: Source checked. Not independent scientific confirmation.

Finding: In a single-center retrospective cohort of 582 hospitalized children aged 1 month to 5 years with community-acquired pneumonia, XGBoost predicted a composite adverse outcome: stay longer than seven days, PICU admission, or in-hospital death. On a random 175-child validation split, it had AUC 0.84 (95% CI 0.78–0.90), accuracy 81.1%, sensitivity 78.6%, and specificity 82.3%, outperforming the compared random-forest and logistic-regression models. Etiology was unavailable for 41.2% of records; no external or prospective evaluation was reported.

Technology: XGBoost, compared with random forest and logistic regression; SHAP was used to rank indicators.

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
XGBoost integrated admission clinical and laboratory indicators to classify the study-defined adverse prognosis; SHAP analyzed feature contributions.
Test and result
Retrospective data from 582 children were split into 407 training and 175 validation records. XGBoost validation AUC was 0.84 (95% CI 0.78–0.90), with 81.1% accuracy, 78.6% sensitivity and 82.3% specificity. Top indicators included admission procalcitonin, C-reactive protein, respiratory rate, age under six months and oxygen saturation.
Limitation / still unknown
This was a retrospective internal split, not a prospective clinical-impact evaluation. Etiologic data were missing for 240/582 children, and no external validation was reported. This was a single-center retrospective internal split, not a prospective clinical-impact evaluation.
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/PMC12891092/ · https://europepmc.org/article/MED/41684921

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: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

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.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.