Participant Characteristics for HIV Viral Load Predictors in Makonde District, Zimbabwe

530 participants (156 cases with high viral load and 374 controls) from a retrospective case-control study in Zimbabwe. The dataset, created by Laston Gonah and last updated in May 2026, contains clinical data used to identify predictors of virological failure in people living with HIV on antiretroviral therapy. A Naïve Bayes classification model was developed from this data to support risk prioritization for viral load testing.

Use Cases

• Training classification models to predict high viral load risk based on patient demographics and clinical history.
• Analyzing the association between factors like age, marital status, and weight loss with virological failure.
• Benchmarking the performance of simple machine learning algorithms like Naïve Bayes against logistic regression for clinical data.
• Supporting the design of triage tools to prioritize viral load testing in settings with limited laboratory capacity.

Strengths

• Dataset includes 530 patient records with a clear case-control structure (156 cases, 374 controls).
• Specific predictors like age group, marital status, and ART duration are identified with odds ratios and confidence intervals.
• Model performance metrics (78.2% sensitivity, 96.5% specificity) are provided for the developed Naïve Bayes algorithm.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Performance estimates are based on the development dataset only and require external validation.

Provenance

Source

Secondary clinical data from two public ART facilities in Makonde District, Zimbabwe.

Collection Method

Retrospective case-control study.

Time Range

null

Freshness

Last updated 2026-05-22 17:33:56; freshness should be verified.

Geography

Makonde District, Zimbabwe.