Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission
Background: Acute respiratory infections (ARIs) are common in children. We developed machine learning models to predict pediatric ARI pathogens at admission. Methods: We included hospitalized children with respiratory infections between 2010 and 2018. Clinical features were collected within 24 h of...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-08-01
|
| Series: | Journal of Microbiology, Immunology and Infection |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1684118223000890 |
| _version_ | 1797749193408774144 |
|---|---|
| author | Tu-Hsuan Chang Yun-Chung Liu Siang-Rong Lin Pei-Hsin Chiu Chia-Ching Chou Luan-Yin Chang Fei-Pei Lai |
| author_facet | Tu-Hsuan Chang Yun-Chung Liu Siang-Rong Lin Pei-Hsin Chiu Chia-Ching Chou Luan-Yin Chang Fei-Pei Lai |
| author_sort | Tu-Hsuan Chang |
| collection | DOAJ |
| description | Background: Acute respiratory infections (ARIs) are common in children. We developed machine learning models to predict pediatric ARI pathogens at admission. Methods: We included hospitalized children with respiratory infections between 2010 and 2018. Clinical features were collected within 24 h of admission to construct models. The outcome of interest was the prediction of 6 common respiratory pathogens, including adenovirus, influenza virus types A and B, parainfluenza virus (PIV), respiratory syncytial virus (RSV), and Mycoplasma pneumoniae (MP). Model performance was estimated using area under the receiver operating characteristic curve (AUROC). Feature importance was measured using Shapley Additive exPlanation (SHAP) values. Results: A total of 12,694 admissions were included. Models trained with 9 features (age, event pattern, fever, C-reactive protein, white blood cell count, platelet count, lymphocyte ratio, peak temperature, peak heart rate) achieved the best performance (AUROC: MP 0.87, 95% CI 0.83–0.90; RSV 0.84, 95% CI 0.82–0.86; adenovirus 0.81, 95% CI 0.77–0.84; influenza A 0.77, 95% CI 0.73–0.80; influenza B 0.70, 95% CI 0.65–0.75; PIV 0.73, 95% CI 0.69–0.77). Age was the most important feature to predict MP, RSV and PIV infections. Event patterns were useful for influenza virus prediction, and C-reactive protein had the highest SHAP value for adenovirus infections. Conclusion: We demonstrate how artificial intelligence can assist clinicians identify potential pathogens associated with pediatric ARIs upon admission. Our models provide explainable results that could help optimize the use of diagnostic testing. Integrating our models into clinical workflows may lead to improved patient outcomes and reduce unnecessary medical costs. |
| first_indexed | 2024-03-12T16:15:45Z |
| format | Article |
| id | doaj.art-ab936143751349ed9f2b5f7390dd3da5 |
| institution | Directory Open Access Journal |
| issn | 1684-1182 |
| language | English |
| last_indexed | 2024-03-12T16:15:45Z |
| publishDate | 2023-08-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Microbiology, Immunology and Infection |
| spelling | doaj.art-ab936143751349ed9f2b5f7390dd3da52023-08-09T04:32:19ZengElsevierJournal of Microbiology, Immunology and Infection1684-11822023-08-01564772781Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admissionTu-Hsuan Chang0Yun-Chung Liu1Siang-Rong Lin2Pei-Hsin Chiu3Chia-Ching Chou4Luan-Yin Chang5Fei-Pei Lai6Department of Pediatrics, Chi Mei Medical Center, Tainan City, TaiwanDepartment of Pediatrics, National Taiwan University Hospital, College of Medicine, National Taiwan University, Taipei City, TaiwanInstitute of Applied Mechanics, National Taiwan University, Taipei City, TaiwanInstitute of Applied Mechanics, National Taiwan University, Taipei City, TaiwanInstitute of Applied Mechanics, National Taiwan University, Taipei City, Taiwan; Corresponding author. Institute of Applied Mechanics, National Taiwan University, No. 1, Sec. 4, Roosevelt Rd., Da'an Dist., Taipei City, 10617, Taiwan.Department of Pediatrics, National Taiwan University Hospital, College of Medicine, National Taiwan University, Taipei City, Taiwan; Corresponding author. Department of Pediatrics, National Taiwan University Hospital and College of Medicine, National Taiwan University, No. 8, Chung-Shan South Road, Taipei City, 10041, Taiwan.Graduate Institute of Biomedical Electronics and Bioinformatics, Taipei City, National Taiwan University, Taiwan; Department of Computer Science and Information Engineering, National Taiwan University, Taipei City, Taiwan; Department of Electrical Engineering, National Taiwan University, Taipei City, TaiwanBackground: Acute respiratory infections (ARIs) are common in children. We developed machine learning models to predict pediatric ARI pathogens at admission. Methods: We included hospitalized children with respiratory infections between 2010 and 2018. Clinical features were collected within 24 h of admission to construct models. The outcome of interest was the prediction of 6 common respiratory pathogens, including adenovirus, influenza virus types A and B, parainfluenza virus (PIV), respiratory syncytial virus (RSV), and Mycoplasma pneumoniae (MP). Model performance was estimated using area under the receiver operating characteristic curve (AUROC). Feature importance was measured using Shapley Additive exPlanation (SHAP) values. Results: A total of 12,694 admissions were included. Models trained with 9 features (age, event pattern, fever, C-reactive protein, white blood cell count, platelet count, lymphocyte ratio, peak temperature, peak heart rate) achieved the best performance (AUROC: MP 0.87, 95% CI 0.83–0.90; RSV 0.84, 95% CI 0.82–0.86; adenovirus 0.81, 95% CI 0.77–0.84; influenza A 0.77, 95% CI 0.73–0.80; influenza B 0.70, 95% CI 0.65–0.75; PIV 0.73, 95% CI 0.69–0.77). Age was the most important feature to predict MP, RSV and PIV infections. Event patterns were useful for influenza virus prediction, and C-reactive protein had the highest SHAP value for adenovirus infections. Conclusion: We demonstrate how artificial intelligence can assist clinicians identify potential pathogens associated with pediatric ARIs upon admission. Our models provide explainable results that could help optimize the use of diagnostic testing. Integrating our models into clinical workflows may lead to improved patient outcomes and reduce unnecessary medical costs.http://www.sciencedirect.com/science/article/pii/S1684118223000890Machine learningChildrenRespiratory infectionsPathogens predictionCommunity-acquired pneumonia |
| spellingShingle | Tu-Hsuan Chang Yun-Chung Liu Siang-Rong Lin Pei-Hsin Chiu Chia-Ching Chou Luan-Yin Chang Fei-Pei Lai Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission Journal of Microbiology, Immunology and Infection Machine learning Children Respiratory infections Pathogens prediction Community-acquired pneumonia |
| title | Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission |
| title_full | Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission |
| title_fullStr | Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission |
| title_full_unstemmed | Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission |
| title_short | Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission |
| title_sort | clinical characteristics of hospitalized children with community acquired pneumonia and respiratory infections using machine learning approaches to support pathogen prediction at admission |
| topic | Machine learning Children Respiratory infections Pathogens prediction Community-acquired pneumonia |
| url | http://www.sciencedirect.com/science/article/pii/S1684118223000890 |
| work_keys_str_mv | AT tuhsuanchang clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT yunchungliu clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT siangronglin clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT peihsinchiu clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT chiachingchou clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT luanyinchang clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission AT feipeilai clinicalcharacteristicsofhospitalizedchildrenwithcommunityacquiredpneumoniaandrespiratoryinfectionsusingmachinelearningapproachestosupportpathogenpredictionatadmission |