Analysis of repeated low-dose challenge studies.

Area of Science:

• Veterinary Medicine
• Immunology
• Biostatistics

Background:

• Preclinical evaluation of human immunodeficiency virus (HIV) vaccines uses challenge studies in non-human primates.
• Repeated low-dose challenge (RLC) studies are increasingly common to mimic natural HIV exposure.
• Current statistical methods for RLC studies may not guarantee a valid vaccine effect test.

Purpose of the Study:

• To detail appropriate statistical approaches for analyzing RLC studies in HIV vaccine development.
• To provide guidance on implementing valid statistical methods in SAS or R.

Main Methods:

• Review of statistical methods used in RLC studies.
• Comparison of standard methods (e.g., modified Fisher's exact test, log-rank test) with exact log-rank tests.
• Demonstration of implementation in SAS and R statistical software.

Main Results:

• Standard statistical methods may not be valid for RLC studies.
• Exact log-rank tests offer a valid approach but can be difficult to implement.
• The paper provides accessible implementations for valid statistical analyses.

Conclusions:

• Accurate statistical analysis is crucial for valid interpretation of RLC studies in HIV vaccine research.
• Researchers can utilize the provided methods for robust analysis of RLC study data.
• Easy implementation of valid statistical tests enhances the reliability of preclinical HIV vaccine evaluation.