Exploring the Impact of Factors Affecting the Lifespan of HIVs/AIDS Patient’s Survival: An Investigation Using Advanced Statistical Techniques

Abstract

This study investigates the impact of various factors on the lifespan and diagnostic time of HIV/AIDS patients using advanced statistical techniques. The Power Chris-Jerry (PCJ) distribution is applied to model CD4 counts of patients, and the goodness-of-fit test confirms a strong fit with a p-value of 0.6196. The PCJ distribution is found to be the best fit based on information criteria (AIC and BIC) with the smallest negative log-likelihood, AIC, and BIC values. The study uses datasets from St. Luke hospital Uyo, Nigeria, containing HIV/AIDS diagnosis date, age, CD4 count, gender, and opportunistic infection dates. Multiple linear regression is employed to analyze the relationship between these variables and HIV/AIDS diagnostic time. The results indicate that age, CD4 count, and opportunistic infection significantly impact the diagnostic time, while gender shows a nonsignificant relationship. The F-test confirms the model's overall significance, indicating the factors are good predictors of HIV/AIDS diagnostic time. The R-squared value of approximately 72% suggests that administering antiretroviral therapy (ART) can improve diagnostic time by suppressing the virus and protecting the immune system. Cox proportional hazard modeling is used to examine the effects of predictor variables on patient survival time. Age and CD4 count are not significant factors in the hazard of HIV/AIDS diagnostic time, while opportunistic infection is a significant predictor with a decreasing effect on the hazard rate. Gender shows a strong but nonsignificant relationship with decreased risk of death. To address the violation of the assumption of proportional hazard, the study employs an assumption-free alternative, Aalen ’ s model. In the Aalen model, all predictor variables except age and gender are statistically significant in relation to HIV/AIDS diagnostic time. The findings provide valuable insights into the factors influencing diagnostic time and survival of HIV/AIDS patients, which can inform interventions aimed at reducing transmission and improving early diagnosis and treatment. The Power Chris-Jerry distribution proves to be a suitable fit for modeling CD4 counts, while multiple linear regression and survival analysis techniques provide insights into the relationships between predictor variables and diagnostic time. These results contribute to the understanding of HIV/AIDS patient outcomes and can guide public health interventions to enhance early detection, treatment, and care.