Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Clustering of longitudinal viral loads in the Western Cape
Introduction: Routine viral load (VL) monitoring is important for assessing the effectiveness of ART in South Africa. There is little information however, on how the longitudinal VL patterns change for subgroups of persons living with HIV (PLHIV) who have experienced at least one elevated VL. We inv...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
Department of Public Health and Family Medicine
2020
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | Introduction: Routine viral load (VL) monitoring is important for assessing the effectiveness of ART in South Africa. There is little information however, on how the longitudinal VL patterns change for subgroups of persons living with HIV (PLHIV) who have experienced at least one elevated VL. We investigated the possible longitudinal VL patterns that may exist among this unique population. Methods: This mini-dissertation offers three components; a research protocol (Section A), a literature review (Section B) and a journal ready manuscript (Section C). We examined HIV VL data for the Western Cape from 2008 to 2018, taken from the National Health Laboratory Services (NHLS). Using< 1000 copies/mL as a threshold for viral suppression, we identified 109092 individuals who had at least one instance of an elevated VL. A nonparametric (KML-Shape) and a model-based (LCMM) clustering technique were used to identify latent subgroups of longitudinal VL trajectories among these individuals. Results: Both the KML-Shape and LCMM clustering techniques identified five latent viral load trajectory subgroups. KML-Shape found majority of individuals' trajectories belonged to clusters that had a decreasing longitudinal VL trend (76.6% of individuals), while LCMM found a smaller proportion of individuals' trajectories belonged to clusters that had a decreasing longitudinal trend (52.5% of individuals). Most of the trajectory subgroups identified had long periods of low-level viremia. Conclusion: Although majority of individuals belonged to clusters that had downward trends, further research is needed to better understand factors contributing to membership of clusters that did not have a downward longitudinal trend. Understanding these factors may help in the development of targeted HIV prevention programs for these individuals. |
|---|