Related Experiment Videos
Modeling HIV infectivity: must sex acts be counted?
1School of Organization and Management, Yale University, New Haven, Connecticut 06520.
Journal of Acquired Immune Deficiency Syndromes
|January 1, 1990
Summary
This study models human immunodeficiency virus (HIV) transmission probability. While a constant infection probability per partner is supported, the Bernoulli model may not fit individual sexual contacts.
Area of Science:
- Epidemiology
- Biostatistics
- Infectious Disease Modeling
Background:
- HIV transmission models often simplify the process to independent Bernoulli trials.
- Understanding transmission probability is crucial for public health interventions.
Purpose of the Study:
- To derive and apply an approximate maximum likelihood estimator for HIV transmission probability.
- To compare simple Bernoulli models with nonparametric alternatives for HIV infectivity.
Main Methods:
- Developed an approximate maximum likelihood estimator for Bernoulli models of HIV transmission.
- Applied the estimator to two existing HIV transmission data sets.
- Constructed and analyzed nonparametric models of HIV infectivity for comparison.
Main Results:
- The Bernoulli model with constant infection probability per partner showed suitability.
- The Bernoulli model's appropriateness was questioned at the level of individual sexual contacts.
- Nonparametric models provided valuable comparisons to simple Bernoulli models.
Conclusions:
- HIV infectivity can be modeled as a Bernoulli process concerning partners.
- The simple Bernoulli model may be insufficient for modeling transmission at the sexual contact level.
- Further probabilistic research is needed to explain observed HIV transmission patterns.