HIV/AIDS epidimiological modeling

Kesaobaka Modukanele
Blog Post for Tuesday November3, 2009

From reading this article, I now realize the importance of mathematical models in epidemiology. Such simulations are always a great way to recreate a real life scenario and make predictions without having to wait for population statistics. One thing that I found really interesting was the paper’s conclusion that in resource poor countries, “HIV epidemics are not amenable to control through treatment, regardless of the extent of ART roll-out, and must be integrated with prevention methods.” The models also went on to reveal , “In the absence of substantial behaviour change of treated patients through effective counseling, prevalence is likely to increase”. The amazing thing is that, a lot of times, a country’s success in combating HIV/AIDS is usually measured by the availability of ARV’s for people. This situation is evident in the case of Botswana. With its pioneering “Masa” program, Africa’s first national Anti-Retro-Viral program, Botswana has oftentimes been referred to as “Africa’s Success Story” or “Beacon of light”. However, these models show that these efforts are not enough. Success of combating HIV/AIDS would actually be more accurate if measured in terms of the success of prevention programs. However, one thing that makes it difficult to use prevention efforts as a measure of a country’s success in combating HIV/AIDS lies in the difficulty of measuring the rate of behavior change.

Another interesting aspect of these models was how they show that “in the absence of substantial behavior change of treated patients through effective counseling, prevalence is likely to increase.” This creates a very interesting scenario because countries with pioneering ARV treatment programs could attribute a continued increase in prevalence rate to the fact that with more treatment, the life span among HIV positive people increases and thus a continued prevalence rate could be a “good” thing, as it indicates increased lifespan. However, the models show that a continuing increase in prevalence rate could also be attributed to ‘absence of substantial behavior change,” which is actually bad. In this scenario, epidemiologists would have to differentiate between the causes of increases in prevalence rate as this could either be a good indicator in terms of increasing life span, or a bad indicator in terms of lack of behavior change.