Authors: David R Patient and Neil M Orr, August 29 2014 - Every few years a national HIV study comes out and people start to speak with great concern about the prevalence and incidence rates increasing. There seems to be a lot of confusion around the difference between the two. The purpose of this article is to unpack the difference between Prevalence and Incidence and explore each one’s relevance.

Prevalence is simple: The total number of people infected with HIV, no matter when they got infected, as a percentage of the total population. For example, in 2012 the HIV prevalence for all people older in South Africa was 12.2% according to the Human Sciences Research Council (HSRC) Report. This works out to be about 6,400,000 people living with HIV out of a total population of about 53 million.

Prevalence rates can be worked out for specific groups, such as babies, youth, and adults of specific ages, and for men versus women. For example, for the population aged 15 to 49 - the more sexually active part of the population - the prevalence rate is 18.8%. The prevalence rate four years ago for this group was 16.9%. So there has been an 11% increase in HIV prevalence in this age group over the last four years. What does this mean? Are things getting much worse? Is treatment and prevention failing? What is going on?

Incidence is a very specific thing: How many people got infected in the past 12 months only. In other words, this is a specific slice of the prevalence numbers, and only concerns new infections.

Imagine for a moment that HIV prevalence is a large swimming pool. At the one end of the pool is where new infections (incidence) enter and cause the water level to rise. Then there is the rest of the pool - the deeper and bigger part of the pool - of people that were infected longer than 12 months ago. Added all together, this is prevalence. The only way that people can be subtracted from this pool of prevalence is when they die, which will cause the water level to drop. In other words, if 1 percent of the population become infected in the past 12 months, and the same number of people died from AIDS, the amount of people in the pool (prevalence) stays the same.

In the days before ART [antiretroviral therapy], the only things that affected the prevalence rates were new infections and death from AIDS. If they were the same, the prevalence rate (the total water in the pool) stayed the same. If more people got infected than died, the water level rose. If more people died from AIDS than those who got infected, the water level dropped.

ART medicines changed this situation because the number of people getting sick with AIDS dropped dramatically. Many people got better, stayed healthy, and far fewer people are dying from AIDS. So the water level in our prevalence ‘pool’ does not drop nearly as much as before. In theory – if everyone who needs ART gets it and also takes their medicines every day (adherence) – there should be no more deaths from AIDS. The water level should, technically, just get higher and higher as more people get infected incidence).

The problem is that prevalence rates don’t tell you anything about when people got infected, who is sick or healthy, or how many people are being treated with ART (antiretroviral treatment) medications or not. In theory, if the prevalence increases (and new infections stay the same), this is a good thing because it says that people who need ART are getting it, and are staying alive, not dying from AIDS.  However, this is not the case in reality: For a variety of reasons, some people don’t get tested and treated in time, and many people don’t take their medicines as they should. So people are still dying from AIDS, even though this could stop.

Nevertheless, the number of people getting ill and dying from AIDS is getting less because of ART. According to the HSRC 2012 report, 31.2% of people living with HIV are taking ART (about 2 million people on ART out of 6.4 million people living with HIV).

Only if there was not a single new infection in South Africa, and if every person who needs ART gets it and takes it properly, the prevalence rate would stay exactly the same each year. If some people are not getting ART they will probably die, and the prevalence rate will come down. So an increase in prevalence rate is NOT something bad by itself - it could mean that treatment is working well.

It takes quite a few years before a person living with HIV needs ART. So incidence has nothing to do with ART. Also, incidence is only about new infections in the past 12 months, which means this number is not affected by what we did (or did not) do more than a year ago. That is why for people and organisations working with HIV prevention, HIV incidence is very information because it basically tells us how well we are doing now (or recently) in terms of preventing new infections.

Overall, the HIV incidence in 2012 for all people ages 2 years and older was 1.07% (HSRC, 2012). In other words, 1 in 93 people became HIV-infected in 2012. Another way of seeing this is as follows: Of the 12.2 percent (prevalence, 2012) of the total population infected with HIV, about 1 percent were infected in the last 12 months (incidence), while the remaining 11 percent were infected more than a year ago.

For the age group 15 to 49, the overall incidence rate is 1.72%. In other words, 1 in 58 people aged 15 to 49 became infected with HIV in 2012. This is 1 in 83 men, and 1 in 44 women in this age group. Shockingly, it was found that females between the ages of 15 and 24 years (1 in 39) were almost 5 times more likely to be newly infected than men (1 in 182) in the same age group (2.54% versus 0.55% incidence, 2012).

Here’s the flip-side: In 2002 the HIV incidence in women aged 15 to 24 was 5.5% (1 in 18). In 2012 this had been cut to 2.1% (1 in 48), which is a 64% reduction in new infections in this age group. In other words, prevention efforts in this age group are working, although there is a long way to go to get it as low as possible. In contrast, the incidence rate for men 15 to 24 years has increased by 40% (0.6% to 1.0%). Something is not working in this area in terms of prevention.

When you look at the incidence rates versus prevalence rate, a different - more useful - picture emerges, particularly for prevention efforts. For example, for all people aged 15 to 49, the number of new infections has decreased by 14% from 2002 to 2012 (2.2% versus 1.9%), but there are increases in specific groups, which we need to target more effectively.

In brief: The total number of new people getting infected (incidence) is, on average, getting a little less each year. However, because more people are getting treated with ART less people are dying from AIDS, and so the total number of people living with HIV (prevalence) is still going up.

Click here to read the original blog on the Living with HIV website.