Integrated clinical and laboratory assessment for severity classification in pediatric pertussis
- 1Pediatrics Department, Guizhou Medical University, Guizhou, China
- 2Pediatric Respiratory Department, The Affiliated Hospital of Guizhou Medical University, Guizhou, China
Abstract
Background: Early identification of children with severe pertussis is essential for timely intervention and improved clinical outcomes.
Objective: To assess clinical and laboratory characteristics associated with severe pertussis and to evaluate their ability to classify severe disease at admission among hospitalised children.
Methodology: This retrospective cohort study included 722 children aged 1 month–13 years with laboratory-confirmed pertussis admitted from January 2020 to December 2025. Clinically diagnosed, laboratory-unconfirmed cases were excluded. Demographic, clinical, laboratory, radiological, and outcome data were extracted from electronic medical records. Univariable and multivariable logistic regression identified factors associated with severe pertussis, with model performance assessed by AUC. Because several predictors contributed to the definition of severe disease, the integrated model was interpreted as a concurrent classification model. A second model using demographic, historical, and admission laboratory variables was interpreted as an admission-time severity classification model. Bootstrap internal validation and calibration were performed.
Results: Severe pertussis occurred in 157 children (21.7%). In the concurrent model, apnea and respiratory distress were strongly associated with severe disease but were not interpreted as independent predictors because they formed part of the severity definition. Age <6 months, prematurity, incomplete-for-age or absent vaccination, radiological pneumonia, marked leukocytosis, hypoalbuminemia, and low immunoglobulin G (IgG) levels were also independently associated with severe disease (all p < 0.05). The apparent AUC was 0.89 (sensitivity 83.4%, specificity 82.1%, accuracy 82.4%), although this performance may be overestimated because of outcome-information leakage. In the admission-time model, age <6 months, incomplete-for-age or absent vaccination, higher white blood cell and absolute lymphocyte counts, and lower albumin and IgG levels were independently associated with severe pertussis. For the admission-time model, AUC was 0.961 (95% CI 0.947–0.975), with an optimism-corrected figure of 0.957, an apparent measure of discrimination that should not be taken as a measure of prediction of clinical deterioration.
Conclusion: Severe pertussis was associated with distinct demographic and laboratory abnormalities, with admission-time variables showing strong discrimination for severe disease. Given that many severe cases met severity criteria at the index event, the admission-time model should not be interpreted as prognostic for subsequent deterioration.
Related
- Integrated field and laboratory assessment of Swiss grapevine cultivar susceptibility to flavescence doree reveals a central role for plant-vector interactions
- Matching laboratory and real-world listening environments with ecological momentary assessment improves prediction of speech perception.
- Integrated multi-omics reveals methylation-linked biological heterogeneity in intermediate-risk pediatric acute myeloid leukemia
- Computational alternative to cardiac biopsy in multiple myeloma: utilizing the shape of routine laboratory data for the classification of AL cardiac amyloidosis
- Pediatric Sleep Assessment through Spectral Sleep Scoring: A Comparison with Polysomnography
BibTeX
@article{wang2026integrated,
title = {{Integrated clinical and laboratory assessment for severity classification in pediatric pertussis}},
author = {Xuexin Wang and Xiaoping Zhu},
journal = {Frontiers in Medicine},
year = {2026},
month = oct,
doi = {10.3389/fmed.2026.1961021},
url = {https://doi.org/10.3389/fmed.2026.1961021},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Integrated clinical and laboratory assessment for severity classification in pediatric pertussis
AU - Xuexin Wang
AU - Xiaoping Zhu
PY - 2026
DA - 2026/10/21
JO - Frontiers in Medicine
DO - 10.3389/fmed.2026.1961021
UR - https://doi.org/10.3389/fmed.2026.1961021
AB - Background: Early identification of children with severe pertussis is essential for timely intervention and improved clinical outcomes. Objective: To assess clinical and laboratory characteristics associated with severe pertussis and to evaluate their ability to classify severe disease at admission among hospitalised children. Methodology: This retrospective cohort study included 722 children aged 1 month–13 years with laboratory-confirmed pertussis admitted from January 2020 to December 2025. Clinically diagnosed, laboratory-unconfirmed cases were excluded. Demographic, clinical, laboratory, radiological, and outcome data were extracted from electronic medical records. Univariable and multivariable logistic regression identified factors associated with severe pertussis, with model performance assessed by AUC. Because several predictors contributed to the definition of severe disease, the integrated model was interpreted as a concurrent classification model. A second model using demographic, historical, and admission laboratory variables was interpreted as an admission-time severity classification model. Bootstrap internal validation and calibration were performed. Results: Severe pertussis occurred in 157 children (21.7%). In the concurrent model, apnea and respiratory distress were strongly associated with severe disease but were not interpreted as independent predictors because they formed part of the severity definition. Age <6 months, prematurity, incomplete-for-age or absent vaccination, radiological pneumonia, marked leukocytosis, hypoalbuminemia, and low immunoglobulin G (IgG) levels were also independently associated with severe disease (all p < 0.05). The apparent AUC was 0.89 (sensitivity 83.4%, specificity 82.1%, accuracy 82.4%), although this performance may be overestimated because of outcome-information leakage. In the admission-time model, age <6 months, incomplete-for-age or absent vaccination, higher white blood cell and absolute lymphocyte counts, and lower albumin and IgG levels were independently associated with severe pertussis. For the admission-time model, AUC was 0.961 (95% CI 0.947–0.975), with an optimism-corrected figure of 0.957, an apparent measure of discrimination that should not be taken as a measure of prediction of clinical deterioration. Conclusion: Severe pertussis was associated with distinct demographic and laboratory abnormalities, with admission-time variables showing strong discrimination for severe disease. Given that many severe cases met severity criteria at the index event, the admission-time model should not be interpreted as prognostic for subsequent deterioration.
SN - 2296-858X
ER -