| ObjectiveAcquired immunodeficiency syndrome(AIDS)has become a severe public health problem and social problem which poses a serious threat to human health.With the development of society,economy and culture,as well as the constant revision of antiviral therapy(ART)standards,new HIV/AIDS patients receive ART almost immediately.The number of HIV/AIDS patients who do not accept ART is gradually decreasing,and limited sources of study subjects also would restrict the study of natural history of AIDS.Therefore,this study aimed to retrospectively collect the data of HIV/AIDS patients in a high-risk area of Henan Province,and screen out the data of disease progression of HIV-infected patients before receiving ART,and explore the related factors of natural progression of HIV to AIDS,so as to provide some reference for the study of natural history of AIDS.MethodsA retrospective analysis was conducted to collect the baseline information and follow-up data of HIV/AIDS patients diagnosed in a high-risk area of Henan Province from 1995 to 2016.According to the purpose,rapid progressors(RPs)and typical progressors(TPs)were included in this study.Group(0=RPs group,1=TPs group)was dependent variable,gender,residential area,nationality,contact with HIV/AIDS patients and experience as a peer educator or not were covariates.The two groups were matched at 1:1,1:2 and 1:3 using propensity score matching(PSM)model,selecting the nearest neighbor matching method and setting the caliper value at 0.02.Standard difference was used to test the balance of covariates between matched groups.The equilibrium distribution of covariates between two groups can be considered when the standard difference d is less than 0.1.After matching,unconditional logistic regression model and conditional logistic regression model were constructed based on the data before and after matching to explore the factors of natural progression of HIV infection.Microsoft Excel 2010,IBM SPSS Statistics22.0 and R 3.5.0 software were applied for data collation and statistical analysis.Results1.Baseline data and follow-up data of 13142 HIV/AIDS patient were collected in this study.A total of 2095 cases were selected,including 1716 cases in RPs group and 379 cases in TPs group.374 pairs,296 pairs and 227 pairs of study subjects were successfully matched after 1:1,1:2 and 1:3 propensity score matching.The standard difference picture showed that the standard difference value d of each covariate between the two groups was less than 0.1 after matching.2.The comparison of fitting results showed that the 1:1 conditional logistic regression model had the best fitting effect.The results showed that HIV-infected patients who had higher CD4~+T lymphocyte level at baseline(OR=4.14,95%CI:3.02-5.67)had a slower natural disease progression to AIDS;patients discovered from the unique survey/physical examination(OR=4.32,95%CI:2.50-7.47)had a slower natural disease progression to AIDS compared with discovering from voluntary counseling and testing;patients transmitted by sexual transmission(OR=0.35,95%CI:0.21-0.59)had a slower natural disease progression to AIDS than transmitting by blood transmission.In addition,patients with clinical symptoms at baseline(OR=0.50,95%CI:0.32-0.78)had a faster natural disease progression to AIDS,and patients with older age at diagnosis(OR=0.73,95%CI:0.60-0.90)also had a faster natural disease progression to AIDS.Conclusions1.The comparison of fitting in this study showed that propensity score matching method can effectively control the bias caused by different distribution of covariates between the groups in observational studies.2.Higher CD4~+T lymphocyte level at baseline and sample from unique survey/physical examination are protective factors for the natural progression of HIV infection.Sexual transmission,with clinical symptoms at baseline and older age at diagnosis are risk factors for the natural progression of HIV infection,which can also provide some reference for the study of natural history of AIDS. |
PDF Full Text Request