Background: Modeling of multivariate longitudinal data provides a unique opportunity in studying the joint evolution of multiple response variables over time. This study was modeling of multivariate longitudinal data cell counts on human immunodeficiency virus infections at Gondar districts were used two different multivariate repeated measurement with a kronecker product covariance and random coefficient mixed models.
Methods: The study was based on data from 566 per four visits human immunodeficiency virus infections were enrolled in the first 4 visits of the 5-year secondary data with retrospective longitudinal study design.
Results: The results revealed that both models of human immunodeficiency virus infections cell counts values of CD4 and CD8 increase over time while hemoglobin decreases over time. Those models results reveals that a strong positive correlation between CD4 and CD8 cells, but the correlation between CD4 and hemoglobin as well as the correlation between CD8 and hemoglobin are not statistically significant at 5% level of significance.
Conclusion: According to this study in the early stage of human immunodeficiency virus infections, CD4 and CD8 steadily increase over time while the values of hemoglobin decrease over time. Consequently, the study suggests that concerned bodies should focus on awareness creation to increase cell counts of CD4 and CD8 over time while hemoglobin decrease over time for HIV infections at Gondar districts, Ethiopia.
HIV Infections; Cell counts; Multivariate longitudinal data; Kronecker product covariance; Random coefficients
Abbreviations: CD4 Count: T-Cell Test; CD8 Count: +T-Cell Test; HMG: Hemoglobin Levels; HIV: Human Immunodeficiency Virus; SAS: Statistical Analysis System; AIC: Akaike Information Critical; BIC: Bayesian Information Critical; -2LnL: Likelihood Ratio Test; CS: Compound Symmetry; AR(1): Autoregressive Order One; TOEP: Toeplitz; UN: Unstructured; ICC: Intra-Cluster Correlation; PHID: Patient Health Identification; Proc: Procedure.