tetano
Editor, Senior Moderator
Heliyon
. 2024 Jul 31;10(15):e35571.
doi: 10.1016/j.heliyon.2024.e35571. eCollection 2024 Aug 15. Development and validation of a nomogram for predicting occurrence of severe case in children hospitalized with influenza A (H1N1) infection during the post-COVID-19 era
Hai-Feng Liu[SUP] 1 [/SUP], Xiao-Zhong Hu[SUP] 2 [/SUP], Cong-Yun Liu[SUP] 3 [/SUP], Zheng-Hong Guo[SUP] 4 [/SUP], Rui Lu[SUP] 5 [/SUP], Mei Xiang[SUP] 6 [/SUP], Ya-Yu Wang[SUP] 7 [/SUP], Zhao-Qing Yin[SUP] 8 [/SUP], Min Wang[SUP] 9 [/SUP], Ming-Ze Sui[SUP] 1 [/SUP], Jia-Wu Yang[SUP] 1 [/SUP], Hong-Min Fu[SUP] 1 [/SUP]
Affiliations
Background: The significant rebound of influenza A (H1N1) virus activity, particularly among children, with rapidly growing number of hospitalized cases is of major concern in the post-COVID-19 era. The present study was performed to establish a prediction model of severe case in pediatric patients hospitalized with H1N1 infection during the post-COVID-19 era.
Methods: This is a multicenter retrospective study across nine public tertiary hospitals in Yunnan, China, recruiting pediatric H1N1 inpatients hospitalized at five of these centers between February 1 and July 1, 2023, into the development dataset. Screening of 40 variables including demographic information, clinical features, and laboratory parameters were performed utilizing Least Absolute Shrinkage and Selection Operator (LASSO) regression and logistic regression to determine independent risk factors of severe H1N1 infection, thus constructing a prediction nomogram. Receiver operating characteristic (ROC) curve, calibration curve, as well as decision curve analysis (DCA) were employed to evaluate the model's performance. Data from four independent cohorts comprised of pediatric H1N1 inpatients from another four hospitals between July 25 and October 31, 2023, were utilized to externally validate this nomogram.
Results: The development dataset included 527 subjects, 122 (23.1 %) of whom developed severe H1N1 infection. The external validation dataset included 352 subjects, 72 (20.5 %) of whom were eventually confirmed as severe H1N1 infection. The LASSO regression identified 19 candidate predictors, with logistic regression further narrowing down to 11 independent risk factors, including underlying conditions, prematurity, fever duration, wheezing, poor appetite, leukocyte count, neutrophil-lymphocyte ratio (NLR), erythrocyte sedimentation rate (ESR), lactate dehydrogenase (LDH), interleukin-10 (IL-10), and tumor necrosis factor-α (TNF-α). By integrating these 11 factors, a predictive nomogram was established. In terms of prediction of severe H1N1 infection, excellent discriminative capacity, favorable accuracy, and satisfactory clinical usefulness of this model were internally and externally validated via ROC curve, calibration curve, and DCA, respectively.
Conclusion: Our study successfully established and validated a novel nomogram model integrating underlying conditions, prematurity, fever duration, wheezing, poor appetite, leukocyte count, NLR, ESR, LDH, IL-10, and TNF-α. This nomogram can effectively predict the occurrence of serious case in pediatric H1N1 inpatients during the post-COVID-19 era, facilitating the early recognition and more efficient clinical management of such patients.
Keywords: Children; Influenza A (H1N1) virus; Machine learning; Nomogram; Post-COVID-19 era.
. 2024 Jul 31;10(15):e35571.
doi: 10.1016/j.heliyon.2024.e35571. eCollection 2024 Aug 15. Development and validation of a nomogram for predicting occurrence of severe case in children hospitalized with influenza A (H1N1) infection during the post-COVID-19 era
Hai-Feng Liu[SUP] 1 [/SUP], Xiao-Zhong Hu[SUP] 2 [/SUP], Cong-Yun Liu[SUP] 3 [/SUP], Zheng-Hong Guo[SUP] 4 [/SUP], Rui Lu[SUP] 5 [/SUP], Mei Xiang[SUP] 6 [/SUP], Ya-Yu Wang[SUP] 7 [/SUP], Zhao-Qing Yin[SUP] 8 [/SUP], Min Wang[SUP] 9 [/SUP], Ming-Ze Sui[SUP] 1 [/SUP], Jia-Wu Yang[SUP] 1 [/SUP], Hong-Min Fu[SUP] 1 [/SUP]
Affiliations
- PMID: 39170375
- PMCID: PMC11336715
- DOI: 10.1016/j.heliyon.2024.e35571
Background: The significant rebound of influenza A (H1N1) virus activity, particularly among children, with rapidly growing number of hospitalized cases is of major concern in the post-COVID-19 era. The present study was performed to establish a prediction model of severe case in pediatric patients hospitalized with H1N1 infection during the post-COVID-19 era.
Methods: This is a multicenter retrospective study across nine public tertiary hospitals in Yunnan, China, recruiting pediatric H1N1 inpatients hospitalized at five of these centers between February 1 and July 1, 2023, into the development dataset. Screening of 40 variables including demographic information, clinical features, and laboratory parameters were performed utilizing Least Absolute Shrinkage and Selection Operator (LASSO) regression and logistic regression to determine independent risk factors of severe H1N1 infection, thus constructing a prediction nomogram. Receiver operating characteristic (ROC) curve, calibration curve, as well as decision curve analysis (DCA) were employed to evaluate the model's performance. Data from four independent cohorts comprised of pediatric H1N1 inpatients from another four hospitals between July 25 and October 31, 2023, were utilized to externally validate this nomogram.
Results: The development dataset included 527 subjects, 122 (23.1 %) of whom developed severe H1N1 infection. The external validation dataset included 352 subjects, 72 (20.5 %) of whom were eventually confirmed as severe H1N1 infection. The LASSO regression identified 19 candidate predictors, with logistic regression further narrowing down to 11 independent risk factors, including underlying conditions, prematurity, fever duration, wheezing, poor appetite, leukocyte count, neutrophil-lymphocyte ratio (NLR), erythrocyte sedimentation rate (ESR), lactate dehydrogenase (LDH), interleukin-10 (IL-10), and tumor necrosis factor-α (TNF-α). By integrating these 11 factors, a predictive nomogram was established. In terms of prediction of severe H1N1 infection, excellent discriminative capacity, favorable accuracy, and satisfactory clinical usefulness of this model were internally and externally validated via ROC curve, calibration curve, and DCA, respectively.
Conclusion: Our study successfully established and validated a novel nomogram model integrating underlying conditions, prematurity, fever duration, wheezing, poor appetite, leukocyte count, NLR, ESR, LDH, IL-10, and TNF-α. This nomogram can effectively predict the occurrence of serious case in pediatric H1N1 inpatients during the post-COVID-19 era, facilitating the early recognition and more efficient clinical management of such patients.
Keywords: Children; Influenza A (H1N1) virus; Machine learning; Nomogram; Post-COVID-19 era.