Survival Analysis And Prediction Model Construction Of 733 Elderly HIV/AIDS Patients

Objective:By analyzing the survival influencing factors of 733 elderly HIV/AIDS patients in Zhuhai and Luoding from 2008 to 2019,the risk factors leading to the death of patients due to AIDS and related diseases were evaluated,and a nomogram was drawn based on the risk factors for prediction the prognosis of individual elderly HIV/AIDS patients.Methods:Using the historical cohort research method,the subjects of the study were elderly HIV/AIDS patients diagnosed by the Centers for Disease Control and Prevention in Zhuhai and Luoding city from January 1,2008 to December 31,2019.Collect the time of diagnosis and cause of death of all patients Data such as time,loss to follow-up,withdrawal and referral time,demographic characteristics,baseline blood creatinine,hemoglobin,CD₄⁺T lymphocyte counts,etc.The study observation endpoint was death from AIDS and related diseases.Descriptive analysis of the epidemiological characteristics of the study object,using the life table method to calculate the survival rate of the patient,using the Kaplan-Meier method to draw the survival rate curve and log-rank test for univariate analysis,selecting the difference in univariate analysis（P<0.05）were incorporated into the COX proportional hazard regression model for multi-factor analysis.According to the results of the multi-factor analysis,the R language was used to draw a nomogram to construct a predictive model and conduct internal verification.Results:A total of 733 elderly HIV/AIDS patients were included in this study.The average age of the patients were 61.6±8.0 years old,ranging from 50 to 86 years old.There were 525 males and 208 females.The ratio of male to female was 2.5:1.Most of the patients were heterosexual transmission.Mainly junior high school and below,married or cohabiting.The cumulative survival probabilities for 1,3,5,and 10 years of follow-up were 94%,91%,89%,and 86%respectively.1.The results of univariate analysis showed age,gender,route of infection,and whether the disease was affected in the last year pulmonary tuberculosis,current WHO clinical stage,BMI,e GFR,baseline CD₄⁺T lymphocyte counts,anemia,type of disease in the last 3 months,and past use of compound trimethoprim had statistically significant differences（P<0.05）;2.More the results of factor COX regression analysis showed that the differences in age,whether they had tuberculosis in the last year,BMI,e GFR,baseline CD₄⁺T lymphocyte counts,and anemia were statistically significant（P<0.05）;3.Draw a nomogram based on the above 6 influencing factors,and conduct internal verification.The result was showed that the model discrimination C-index was 0.822（95%CI:0.780～0.864）.The calibration curve indicates that the model was in good agreement,indicating that the nomogram model had better predictive ability.Conclusion:Old age,pulmonary tuberculosis in the last year,BMI<18.5kg/m²,e GFR<60ml/min·1.73m²,baseline CD₄⁺T lymphocyte counts and anemia were independent risk factors for AIDS-related death in elderly HIV/AIDS patients.The nomogram model had good discrimination and consistency,and can more accurately predict the prognosis of individual elderly HIV/AIDS patients.