Use of Quintile Regression to Find Factors Influencing CD4 Cell Count in Iranian Newly Recognized HIV-Infected People (1987–2016)

Abstract

BackgroundThe number of CD4 cell counts is one of the markers that are widely used to assess the level of disease progression. Potential risk factors may have a different relationship at different points in the CD4 distribution. For example, some factors may have a weak link with the CD4 mean. While their relationship at the lower levels of CD4 is very strong. Understanding such a different relationship to identify a population at risk based on different levels of CD4 is very beneficial. Accordingly, this study aimed to use quantile regression to provides a more comprehensive analysis of the factors associated across different quantiles of CD4 counts at the time of diagnosis in Iranian newly recognized HIV-infected people. MethodsThis retrospective cohort study included 4402 HIV-infected patients from 31 provinces of Iran from 1987 to 2016. Patients' information included baseline CD4 cell count, age, sex, marital status, educational level, job status, way of HIV transmission, year of diagnosis, TB co-infected, clinical stage at the time of diagnosis. Quantile regression models were fitted to the data. ResultsThe study findings indicated that variables, such as older age, clinical stage, transmission way, and year of HIV diagnosis could play the role of risk factors for CD4 counts at the time of diagnosis in HIV patients. However, the relationship between different variables was not uniformly distributed in many different quantiles. ConclusionsThe results of this study will be helpful for public health policymakers to adopt targeted policies aimed at improving the survival of these patients.