Scholarlink

Properly implemented antiretroviral therapy (ART) program has significant impact on the lives of people living with HIV. This program is affected by various socioeconomic factors in Namibia that derails its effectiveness. This study applies linear mixed effects and generalized estimating equation (GEE) models to analyze factors that affect viral load in a cohort of 154 people living with HIV on ART in Karas Region, Namibia. The patients were followed up at the Luderitz Hospital from January 2015 to December 2017. The data were analyzed using R software version 4.4.1. The results from both models showed that male patients were associated with increased log viral load compared to females, while patient’s baseline weight was inversely related with log viral load. Overall, the directions of estimates were similar between the two models. But the linear mixed effects model produced estimates with smaller standard errors compared to the GEE model. In addition, based on the linear mixed effects model, it was shown that there was high variation of log viral load between patients as time passed. We recommend applying the linear mixed effects model when analyzing ART panel data.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Modeling Viral Load Using Mixed Effects Regression Models and Generalized Estimating Equations

  • 章节

  • 发表日期：

    2025年

查看全文
  • Lineekela Gabriel,
  • Tsirizani M. Kaombe,
  • Lawrence N. Kazembe

摘要

Properly implemented antiretroviral therapy (ART) program has significant impact on the lives of people living with HIV. This program is affected by various socioeconomic factors in Namibia that derails its effectiveness. This study applies linear mixed effects and generalized estimating equation (GEE) models to analyze factors that affect viral load in a cohort of 154 people living with HIV on ART in Karas Region, Namibia. The patients were followed up at the Luderitz Hospital from January 2015 to December 2017. The data were analyzed using R software version 4.4.1. The results from both models showed that male patients were associated with increased log viral load compared to females, while patient’s baseline weight was inversely related with log viral load. Overall, the directions of estimates were similar between the two models. But the linear mixed effects model produced estimates with smaller standard errors compared to the GEE model. In addition, based on the linear mixed effects model, it was shown that there was high variation of log viral load between patients as time passed. We recommend applying the linear mixed effects model when analyzing ART panel data.