The recent HIV incidence rate (or hazard rate for infection) is an important quantity for use in monitoring the HIV/AIDS epidemic, evaluating HIV prevention programs, and allocating HIV prevention resources. Direct measurement of HIV incidence is difficult and time consuming, while estimating HIV incidence via backcalculation (deconvolution) using AIDS incidence data and the (presumed known) HIV incubation time distribution yields little information about recent infection. We propose a method for estimating recent HIV incidence in a population via a single sample at a single point in time. The method relies on understanding the progression of certain markers such as CD4 counts following infection. The actual formulas derived for the point and interval estimates of HIV incidence are simple and easy to use while the sample sizes needed to implement our approach are reasonable. We present two applications of our approach, comparing our results to those obtained from more conventional methods where possible.
- Health care
- Probability, applications
- Statistics, applications
ASJC Scopus subject areas
- Computer Science Applications
- Management Science and Operations Research