Abstract: Background: Some critically ill children with influenza A virus (IAV) infection may have severe sequelae or even die, but its early clinical manifestations are non-specific. At present, there is a lack of relevant prediction models at home and abroad. Objective: To establish a nomogram prediction model for critical IAV infection in children to help early clinical identification of critical IAV infection. Design: Case-control study. Methods: Consecutive patients with IAV infection who were hospitalized in Children's Hospital Affiliated to Zhengzhou University from January 2018 to November 2023 were enrolled. According to the discharge diagnosis and clinical data, they were divided into critically ill children and non-critically ill children. The demographic data, symptoms on admission, laboratory tests on admission, and co-infection with other pathogens were collected. According to the incidence of IAV in Henan Province, the samle size should be greater than 320 cases. The enrolled children were randomly divided into a modeling group and a validation group at a ratio of 7∶3. The influencing factors of critical IAV infection were screened in the modeling group, and the R 4.3.2 software package was used to construct a nomogram prediction model for critical IAV infection. Main outcome measures: Risk factors for critically ill children with influenza A virus (IAV) infection. Results: Among 391 hospitalized children with IAV infection, 134 cases were critically ill, of whom 20 cases (14.9%) had sequelae, all of them were nervous system damage, and 12 cases (9.0%) died. All the non-critically ill children were cured and discharged. There were 274 cases in the modeling group and 117 cases in the validation group. There was no significant difference in clinical data between the two groups. Multivariate Logistic regression analysis showed that neurological symptoms (OR=6.923, 95%CI: 2.569-18.656), co-infection with other pathogens (OR=3.092, 95%CI: 1.379-6.934), and elevated NLR (OR=1.404, 95%CI: 1.029-1.914) and increased IL-6 (OR=1.009, 95%CI: 1.000-1.018) are risk factors, and propagated rise (OR=0.925, 95%CI: 0.862-0.992) is a protection factor. Taking critical IAV infection as the prediction outcome, a nomogram prediction model was constructed based on neurological symptoms, combined with other pathogen infections, and laboratory indicators such as NLR, IL-6, and ALB. The AUC of the model was 0.949 (95%CI: 0.915-0.982) in the modeling group and 0.912 (95%CI: 0.871-0.952) in the validation group. The nomogram model fitted well (χ2=5.077，P=0.749), the predicted probability was in good agreement with the actual probability, and had a high net clinical benefit rate. Conclusion: The nomogram prediction model based on neurological symptoms, infection with other pathogens and laboratory indexes of NLR, IL-6 and ALB is effective and has good discriminative ability.

Key words: Influenza A, Critically ill, Children, Nomogram model, Influencing factors