| Objectives:The purpose of this study is to explore the applicability of B-SHADE model in the estimation of HIV diagnosis rate,and the application of CD4 count back calculation in the estimation of HIV diagnosis rate.Methods:Using B-SHADE model and CD4 count back calculation to study the HIV Diagnosis rate at the municipal and county/district levels.The B-SHADE model uses the number of potentially biased new cases reported by a hospital to obtain the best linear unbiased estimate of the number of cases.Adopt study area a hospital report on the number of new cases to estimate the onset of the region,the total number of collection research area set to sample a hospital report on the number of new cases of weighted to correction of bias,the weighted values by research area sample hospital report on the number of new cases and the region the ratio of the total number of new cases and reflect the relevance of the covariance between all hospitals in the region combination.Finally,the corrected values were added together to obtain the minimum unbiased estimate as the total number of cases in the area.CD4 count back calculation is based on the characteristics of CD4 lymphocyte depletion with the extension of infection time of the infected person.The first CD4 count of newly diagnosed HIV infected person in the study area is collected to reverse estimate the time of HIV infection of the infected person.The Bayes theory was used to estimate the total number of infected persons,and then the number of reported infections was divided by the number of infections to get the HIV diagnosis rate.Through the cohort study,collect newly diagnosed infected CD4 count,for the first time according to the loss of CD4 count trajectories are a time linear function,to calculate the time of people living with HIV infection,in HIV diagnosis delay,calculate the diagnosis delay time distribution.The number of HIV infections in the cohort was estimated for each year,estimating the HIV Diagnosis rate during the study period.Results:Results showed that the B-SHADE model estimated the results of the study area could not obtain the HIV diagnosis rate.By the end of 2019,the number of people infected with HIV in Lincang was estimated to be 14,101 by Bayes theory,and the HIV diagnosis rate was 90.45%.The number of people infected with HIV in Wenshan Prefecture was 27,823,and the HIV diagnosis rate was 54.86%.In Lincang,14,868 people were infected with HIV,and the HIV diagnosis rate was 85.78%.The number of people infected with HIV in Wenshan Prefecture was 16,886,and the HIV diagnosis rate was 90.39%.Conclusions:In this study,B-SHADE model was used to estimate the detection rate of HIV diagnosis.However,the incidence mode of HIV/AIDS is mainly low prevalence,so it is impossible to report the number of cases in real time.The B-SHADE model is difficult to achieve the best unbiased estimation during calculation,and the number of cases in the final calculation area is low,leading to the failure to calculate the HIV diagnosis rate.Therefore,the B-SHADE model is not suitable for the measurement of HIV diagnosis rate.CD4 count back calculation was used to estimate the infected person’s infection time.According to the estimated infection time,Bayes theory and HIV diagnosis delay distribution were applied to estimate the total number of people infected with HIV,and the HIV diagnosis rate was obtained.It provides a theoretical basis and method for the exploration of the estimation method of HIV diagnosis rate at the state/city/district(county)level,and also provides a reference for the formulation of HIV/AIDS prevention and control measures in the future. |
