Development and Internal Validation of a Model for Predicting Intensive Care Unit Admission in Patients with Pulmonary Tuberculosis and Concurrent Klebsiella pneumoniae Pneumonia: A Retrospective Single-Center Study.
5 affiliations
- 1Intensive Care Unit, Public Health Clinical Center of Chengdu, Chengdu, Sichuan, 610061, People's Republic of China.
- 2Public Health Clinical Center of Chengdu & Public Health Clinical Center of Chengdu University of Traditional Chinese Medicine, Chengdu, 610061, People's Republic of China.
- 3Department for Tuberculosis, Public Health Clinical Center of Chengdu, Chengdu, Sichuan, 610061, People's Republic of China.
- 4College of Basic Medical Sciences, Chengdu University of Traditional Chinese Medicine, Chengdu, 611137, People's Republic of China.
- 5Beijing Chest Hospital, Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Research Institute, Beijing, 101149, People's Republic of China.
Abstract
Purpose
Patients with pulmonary tuberculosis (PTB) complicated by Klebsiella pneumoniae (KP) pneumonia are prone to acute deterioration, frequently leading to intensive care unit (ICU) admission and higher mortality. This study aimed to identify independent risk factors for ICU admission among PTB patients with KP pneumonia and develop and internally validate a concise nomogram for early risk prediction.Patients and methods
This single-center retrospective cohort study enrolled 102 PTB patients complicated by KP pneumonia (50 ICU group, 52 non-ICU group) at Public Health Clinical Center of Chengdu from June 2024 to June 2025. Least Absolute Shrinkage and Selection Operator (LASSO) regression with 1000-bootstrap stability assessment was used to screen candidate variables. Firth penalized logistic regression was applied to identify independent risk factors and construct the nomogram. Internal validation was performed via 1000 bootstrap resampling. Multiple statistical approaches, including the area under the curve (AUC), calibration slope, Brier score, Hosmer-Lemeshow test, calibration plots, decision curve analysis (DCA), and clinical impact curves (CIC), were adopted to quantify model discrimination, calibration, and clinical utility.Results
Screened by LASSO regression and further validated with Firth logistic regression, six-month pre-admission Gram-negative bacterial infection history, pre-admission carbapenem exposure within six months, respiratory virus infection, fungal infection, and systemic glucocorticoid use at admission constituted independent risk factors for ICU admission (all P<0.05). The nomogram had an apparent AUC of 0.935 and bootstrap-corrected AUC of 0.920 (optimism = 0.015), suggesting minimal overfitting within the development cohort, though external validation is needed to confirm generalizability. The apparent and optimism-corrected calibration slopes were 1.138 and 0.914, respectively. The apparent Brier score was 0.103, and the corrected Brier score was 0.083. The Hosmer-Lemeshow test indicated favorable calibration (P=0.832). DCA and CIC confirmed stable clinical net benefit across reasonable threshold probabilities.Conclusion
Constructed from five readily-available clinical variables, this nomogram shows good discrimination and satisfactory calibration for predicting ICU-admission risk in PTB patients with KP pneumonia. External validation is still needed to verify its generalizability. Before such validation is performed, the tool should serve only as a supportive risk-stratification aid instead of a standalone clinical decision-making instrument. It may help identify patients who deserve intensified monitoring owing to high risk of clinical deterioration.
Related
- Nutritional factors in predicting severe myelosuppression during concurrent chemoradiotherapy for esophageal cancer: development and internal validation of a pre-treatment nomogram
- Development and internal validation of a nomogram for predicting flexible bronchoscopy utilization in children with lobar pneumonia caused by Mycoplasma pneumoniae: a single-center retrospective cohort study
- Development and validation of machine learning models for predicting pancreatic injury observed after tigecycline initiation: a retrospective real-world study
- Development and internal validation of a nomogram for predicting mechanical ventilation requirement in adult tetanus patients: a dual-center retrospective study
- Development and internal validation of a nomogram-based screening tool for concurrent autoimmune hepatitis in children with intestinal diseases: a retrospective cohort study.
BibTeX
@article{teng2026development,
title = {{Development and Internal Validation of a Model for Predicting Intensive Care Unit Admission in Patients with Pulmonary Tuberculosis and Concurrent Klebsiella pneumoniae Pneumonia: A Retrospective Single-Center Study.}},
author = {Shuangqin Teng and Yi Mao and Haixia Luo and Yao Cheng and Yuke Li and Jiao Yang and Xin Li and Huan Liu and Guihui Wu and Jibin Liu and Shenjie Tang},
journal = {Infection and drug resistance},
year = {2026},
month = nov,
doi = {10.2147/idr.s647201},
url = {https://europepmc.org/article/MED/43017515},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - Development and Internal Validation of a Model for Predicting Intensive Care Unit Admission in Patients with Pulmonary Tuberculosis and Concurrent Klebsiella pneumoniae Pneumonia: A Retrospective Single-Center Study.
AU - Shuangqin Teng
AU - Yi Mao
AU - Haixia Luo
AU - Yao Cheng
AU - Yuke Li
AU - Jiao Yang
AU - Xin Li
AU - Huan Liu
AU - Guihui Wu
AU - Jibin Liu
AU - Shenjie Tang
PY - 2026
DA - 2026/11/07
JO - Infection and drug resistance
DO - 10.2147/idr.s647201
UR - https://europepmc.org/article/MED/43017515
AB - Purpose Patients with pulmonary tuberculosis (PTB) complicated by Klebsiella pneumoniae (KP) pneumonia are prone to acute deterioration, frequently leading to intensive care unit (ICU) admission and higher mortality. This study aimed to identify independent risk factors for ICU admission among PTB patients with KP pneumonia and develop and internally validate a concise nomogram for early risk prediction.Patients and methods This single-center retrospective cohort study enrolled 102 PTB patients complicated by KP pneumonia (50 ICU group, 52 non-ICU group) at Public Health Clinical Center of Chengdu from June 2024 to June 2025. Least Absolute Shrinkage and Selection Operator (LASSO) regression with 1000-bootstrap stability assessment was used to screen candidate variables. Firth penalized logistic regression was applied to identify independent risk factors and construct the nomogram. Internal validation was performed via 1000 bootstrap resampling. Multiple statistical approaches, including the area under the curve (AUC), calibration slope, Brier score, Hosmer-Lemeshow test, calibration plots, decision curve analysis (DCA), and clinical impact curves (CIC), were adopted to quantify model discrimination, calibration, and clinical utility.Results Screened by LASSO regression and further validated with Firth logistic regression, six-month pre-admission Gram-negative bacterial infection history, pre-admission carbapenem exposure within six months, respiratory virus infection, fungal infection, and systemic glucocorticoid use at admission constituted independent risk factors for ICU admission (all P<0.05). The nomogram had an apparent AUC of 0.935 and bootstrap-corrected AUC of 0.920 (optimism = 0.015), suggesting minimal overfitting within the development cohort, though external validation is needed to confirm generalizability. The apparent and optimism-corrected calibration slopes were 1.138 and 0.914, respectively. The apparent Brier score was 0.103, and the corrected Brier score was 0.083. The Hosmer-Lemeshow test indicated favorable calibration (P=0.832). DCA and CIC confirmed stable clinical net benefit across reasonable threshold probabilities.Conclusion Constructed from five readily-available clinical variables, this nomogram shows good discrimination and satisfactory calibration for predicting ICU-admission risk in PTB patients with KP pneumonia. External validation is still needed to verify its generalizability. Before such validation is performed, the tool should serve only as a supportive risk-stratification aid instead of a standalone clinical decision-making instrument. It may help identify patients who deserve intensified monitoring owing to high risk of clinical deterioration.
SN - 1178-6973
ER -