Analyzing Longitudinally Collected Viral Load Measurements in Youth With Perinatally Acquired HIV Infection: Problems...

Area of Science:

• Biostatistics
• Virology
• Pediatric Infectious Diseases

Background:

• Viral load (VL) is a key marker for HIV progression and treatment.
• HIV viral load summaries predict outcomes like neurocognitive disorders.
• Area Under the Curve (AUC) is a common but potentially limited summary of historical viral burden.

Purpose of the Study:

• To compare the traditional Area Under the Curve (AUC) with alternative historical viral load summaries (HVS).
• To evaluate how different HVS relate to cognitive outcomes in youth with perinatally acquired HIV.
• To address potential bias from incomplete VL data in observational studies.

Main Methods:

• Compared AUC with HVS: time-averaged AUC, duration of viremia, time with suppressed VL, peak VL, and age at peak VL.
• Utilized data from the Pediatric HIV/AIDS Cohort Study Adolescent Master Protocol.
• Employed an imputation strategy to handle incomplete VL measurements and conducted a simulation study.

Main Results:

• The choice of HVS and the timing of VL measurements significantly affect associations with cognitive function (full-scale IQ).
• Incomplete VL data can introduce selection bias into analyses.
• An imputation strategy can enable unbiased estimation of HVS with correctly specified models.

Conclusions:

• Alternative HVS may offer more nuanced insights than AUC for HIV research.
• Careful consideration of HVS and measurement timing is essential for interpreting VL data.
• Imputation methods are vital for mitigating bias and improving the reliability of findings from incomplete HIV datasets.