Statistical Studies of Age - Specific HIV - Prevalence
Data
Helmut Knolle
Abstract
The infectivity function is a
function giving a measure of how infectious a given individual is t time units after becoming infected. Today, no feasible and ethically
acceptable study design is known, which would lead to estimates of
HIV-infection probability within
steady heterosexual partnerships, using standard statistical methodology. In
this paper a transmission model is used as a link between the infectivity
function and data sets which already exist or can be generated with standard
methods and moderate expenses. This model suggests that the distribution of
HIV-infections by age and sex depends on the infectivity function as well as on
age-dependent patterns of sexual partner choice. Application of the model
requires population-based data of age-specific HIV-incidences in men and women
of the general heterosexual population. At present, the only known data set
suitable for this purpose is a set of HIV-test results from a sample of 8690
Colombian women in pregnancy who attended prenatal care. The prevalence of HIV
was 0.33% in the group of 12-24 years, but only 0.16% in the group of 25-34
years. The model can explain this strange result. A data set of age-specific
HIV-prevalences in heterosexual Colombian men would be useful, but is not
known. Therefore, further research and data collecting is required in order to
arrive at well founded conclusions.
Key words:
AIDS, Infectivity, Health risk, Branching process, Threshold, Age distribution,
PDF (English)