Clinical Predictors of Non-Response to Initial Antibiotics in Pediatric Community-Acquired Pneumonia: A Risk Score Approach
University of Thi-Qar Journal Of Medicine · 04.06.2026 · 30 (1) · 19-27
DOI: 10.32792/jmed.2026.30.45 ↗Kurz gefasst
Fachartikel in der Zeitschrift University of Thi-Qar Journal Of Medicine, online erschienen am 04.06.2026. Verfasst von Roua Hameed Kadhem. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Community-acquired pneumonia (CAP) remains the leading cause of hospitalization and a major contributor to mortality amongchildren under five years of age globally. First-line antibiotic therapy fails in a substantial minority, with reported failure rates of 10–30% depending on definition and setting. Locally relevant predictor models that combine routine clinical variables with inflammatory biomarkers may improve early identification of children at risk of failure.
Objective
To identify clinical, radiographic, and biomarker predictors of first-line antibiotic treatment failure in children hospitalized with CAP at a single tertiary center, and to develop a multivariable risk model with internal validation. Patients and methods. A prospective cohort study was conducted at Bint Al-Huda teaching hospital from January 2023 through June2024. Eligible children were aged 2 months to 14 years admitted with World Health Organization (WHO)-defined CAP. Hospital-acquired pneumonia, immunocompromise, chronic lung disease, and aspiration were excluded. The primary outcome was treatment failure within 7 days, defined as antibiotic escalation before day 5, pediatric intensive care unit (PICU) transfer, or 30-day readmission. Univariable comparisons used the chi-squared, Fisher exact, Student t, or Mann–Whitney U test as appropriate. Multivariable logistic regression identified independent predictors; discrimination was assessed by the area under the receiver operating characteristic curve (AUC) with bootstrap internal validation.
Results
Of 512 children assessed, 348 formed the analytic cohort; 74 (21.3%) experienced treatment failure. Independent predictors wereprocalcitonin (PCT) ≥ 1.0 ng/mL (adjusted odds ratio [aOR] 3.18, 95% confidence interval [CI] 1.86–5.43), severe acute malnutrition (aOR 2.94, 1.61–5.36), SpO₂ < 92% on room air (aOR 2.86, 1.62–5.04), C-reactive protein (CRP) ≥ 80 mg/L (aOR 2.42, 1.42–4.12), multilobar consolidation (aOR 2.21, 1.30–3.78), and pleural effusion (aOR 2.04, 1.12–3.72). The combined model achieved an AUC of 0.85 (95% CI 0.80–0.89), with bootstrap-corrected AUC of 0.83.
Conclusions
Six routinely available variables three clinical, two biomarker, one radiographic discriminatefirst-line treatment failure in pediatric CAP with good performance. The model is suitable for institutional use after external validation.
Zitieren
APA
Kadhem, R. H. (2026). Clinical Predictors of Non-Response to Initial Antibiotics in Pediatric Community-Acquired Pneumonia: A Risk Score Approach. University of Thi-Qar Journal Of Medicine, 30(1), 19-27. https://doi.org/10.32792/jmed.2026.30.45
Vancouver
Kadhem RH. Clinical Predictors of Non-Response to Initial Antibiotics in Pediatric Community-Acquired Pneumonia: A Risk Score Approach. University of Thi-Qar Journal Of Medicine. 2026;30(1):19-27. doi:10.32792/jmed.2026.30.45
BibTeX
@article{Kadhem2026,
author = {Kadhem, Roua Hameed},
title = {Clinical Predictors of Non-Response to Initial Antibiotics in Pediatric Community-Acquired Pneumonia: A Risk Score Approach},
journal = {University of Thi-Qar Journal Of Medicine},
year = {2026},
volume = {30},
number = {1},
pages = {19--27},
doi = {10.32792/jmed.2026.30.45}
}
EUVIMED