Risk analysis of adverse immune reconstitution in HIV/AIDS patients after antiviral therapy based on random forest model
10.13431/j.cnki.immunol.j.20250041
- VernacularTitle:基于随机森林模型分析艾滋病患者抗病毒治疗后发生免疫重建不良的风险
- Author:
Xiaoshan HE
1
;
Hongbiao HOU
1
;
Yuting JIANG
1
Author Information
1. 514031,梅州市人民医院感染性疾病科
- Publication Type:Journal Article
- Keywords:
Random forest model;
Human immunodeficiency virus;
Acquired immune deficiency syndrome;
Immune non-responder
- From:
Immunological Journal
2025;41(4):274-278
- CountryChina
- Language:Chinese
-
Abstract:
Objective A random forest model was used to analyze the risk of immune reconstitution dysfunction(INR)after antiviral therapy(HAART)in HIV/AIDS patients.Methods A total of 67 HIV/AIDS patients treated in our hospital from January 2019 to December 2022 were selected and all of them received HAART treatment.After 18 months of follow-up,patients were divided into INR group and non-INR group according to prognosis.The clinical data of the two groups were compared,and the importance of the indicators affecting INR in patients was ranked by random forest model,and the related factors affecting INR in HIV/AIDS patients after HAART were analyzed by Logistic regression model.Results During the follow-up period of 18 months,a total of 13(19.40％)of 67 HIV/AIDS patients developed INR.There were statistically significant differences in the time interval from diagnosis to treatment and baseline CD4+T lymphocyte count between INR group and non-INR group(P＜0.05).Random-forest model results showed that variables were ranked in descending order of importance as baseline CD4+T lymphocyte count and time interval between diagnosis and treatment.Logistic multivariate regression analysis showed that more than 1 year between diagnosis and treatment and baseline CD4+T lymphocyte count≤ 200/μl are all risk factors for INR in HIV/AIDS patients after HAART treatment(P＜0.05).Conclusion More than 1 year between diagnosis and treatment and baseline CD4+T lymphocyte count≤200/μl are all risk factors for INR in HIV/AIDS patients after HAART.In addition,the random forest model is of great significance for health care workers to identify high-risk groups of INR after HAART and formulate intervention plans.
