Risk factor and construction of nomogram prediction model for pertussis in children
- VernacularTitle:儿童百日咳发病的危险因素分析及列线图预测模型构建
- Author:
Qianqian SHI
1
;
Jingjing LIU
Author Information
- Keywords: children; pertussis; influencing factors; nomogram model; passive smoking; pertussis vaccine; prolonged cough; predictive efficiency
- From: Journal of Clinical Medicine in Practice 2025;29(6):56-61
- CountryChina
- Language:Chinese
- Abstract: Objective To explore the risk factors for pertussis in children and construct a nomo-gram prediction model for the risk of pertussis in children.Methods A total of 175 children with pro-longed cough admitted from February to June 2024 were selected as modeling group and divided into pertussis group and control group based on whether pertussis was confirmed.The general conditions,clinical characteristics,and laboratory examination indicators of the two groups were compared.Logis-tic regression analysis was used to screen risk factors for pertussis in children,and a nomogram predic-tion model was constructed and internally validated.Additionally,53 children with prolonged cough admitted from July to August 2024 were selected as validation group to externally validate the nomo-gram model.Results Among 175 children in the modeling group,52(29.71％)were diagnosed with pertussis.The proportions of children with poor indoor ventilation,contacting with coughing pa-tients in 3 weeks before onset,passive smoking in 3 weeks before onset,and those who had not re-ceived the pertussis vaccine,as well as platelet levels,were higher in the pertussis group than in the control group(P＜0.05).Multivariate Logistic regression analysis showed that poor indoor ventila-tion,contacting with coughing patients in 3 weeks before onset,passive smoking in 3 weeks before on-set,and not receiving the pertussis vaccine were independent risk factors for pertussis in children(OR=2.983,4.943,3.998,5.943;P＜0.05).Based on these results,a nomogram model for predicting pertussis in children was constructed.The internal validation results showed that the area under the curve(AUC)of this nomogram model for predicting pertussis in children was 0.824,with a sensi-tivity of 82.70％,a specificity of 74.80％,and a goodness-of-fit Hosmer-Lemeshow test result of x2=7.591,P=0.425.The external validation results showed an AUC of 0.799,with sensitivity of 80.80％,with specificity of 63.40％,and a goodness-of-fit Hosmer-Lemeshow test result of x2=10.369,P=0.169.Conclusion Poor indoor ventilation,contacting with coughing patients in 3 weeks before onset,passive smoking in 3 weeks before onset,and not receiving the pertussis vaccine are independent risk factors for pertussis in children.The nomogram prediction model constructed based on these factors can effectively predict the risk of pertussis in children.
