US on course to end its HIV epidemic – eventually

Fall in new infections only achieves half of 2010 target
This article is more than 8 years old. Click here for more recent articles on this topic

If current trends continue, the US may eventually end its HIV epidemic, a mathematical model recently published in AIDS & Behavior shows. In 2009, the average number of people each person with HIV would infect during their lifetime fell below one, and has now declined to 0.75, the model shows. This means the number of people with HIV will eventually start to shrink, as more in the ageing HIV-positive population start to die than get infected. For the moment, however, since mortality in people with HIV also continues to fall, the number of people with the virus in the US continues to grow slowly.

The reductions in new HIV infections and in the proportion of people with HIV who transmit each year are significant, but are only about half of the targets the US government set in 2010 when it published its National HIV/AIDS strategy. The writers of the paper comment that more funding for prevention services is necessary, alongside sustained funding for treatment, if the US is to meet its HIV reduction targets and achieve a quicker end to HIV as a significant epidemic.

Working out the figures

Robert Bonacci of Pennsylvania University and David Holtgrave of Johns Hopkins University based their model on published estimates of HIV prevalence, incidence and transmission rate from 2008 to 2012. They then projected these to 2015, assuming the same trends persist.

Incidence was defined as the number of reported new diagnoses minus the number of reported deaths due to any cause in people recorded as having HIV. The transmission rate was defined as the proportion of the total number of people living with HIV who were newly diagnosed in that year: this gives us an approximate figure for the average number of new HIV infections transmitted by each person with HIV that year – approximate, because it does not take account of people who die before they are ever diagnosed with HIV.

Glossary

mathematical models

A range of complex mathematical techniques which aim to simulate a sequence of likely future events, in order to estimate the impact of a health intervention or the spread of an infection.

virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

The ultimate measure of whether an epidemic will grow or shrink is the reproduction number or R0. This is the transmission rate multiplied by the length of time the average person remains infectious. In the case of HIV, in pre-treatment days this was their lifetime. Since the wide uptake of antiretroviral therapy (ART), HIV-positive people will spend large proportions of their lives non-infectious. However since there are no reliable estimates of the length of time the average person with HIV is infectious, the researchers took the current best estimate for the life expectancy of people with HIV after diagnosis (on average 29 years) as a conservative approximation. Note that this will tend to produce a higher R0 than is in fact the case.

The forecasts

Bonacci and Holtgrave had access to the estimates by the Centers for Disease Control (CDC) of HIV prevalence, new diagnoses and all-cause mortality from 2008 to 2012, and used these to extrapolate until 2015, which was the year by which the National HIV/AIDS strategy hoped to reach its targets. These targets were a fall in HIV incidence between 2010 and 2015 of 25% and a fall in the average number of infections transmitted by each person with HIV of 30%.

Between 2008 and 2012 the number of new diagnoses fell from 40,698 to 35,559. Deaths in people with HIV also fell, however, from 19,898 to 17,459 during the same period, meaning that the estimated number of people in the US living with HIV rose from 1.14 million to 1.22 million.

Because new diagnoses fell even while the number of people living with HIV rose, this meant the transmission rate fell too. In 2008 the average number of times each person with HIV transmitted their infection in a year was 0.035 – one transmission per 28 years. In 2010 it was 0.032 and in 2012 it was 0.029 – one transmission in 34 years.

The researchers then extrapolated these data to 2015, assuming continued trends in the diagnosis and mortality rates. They estimated that by the end of last year the number of new diagnoses would have fallen further to 33,200, and deaths to 16,100. This would mean an increase in the number of people living with HIV to 1.27 million. And that would imply a continued fall in the transmission rates from 0.029 to 0.0.026 or one transmission in 38 years.

This indicates a decline in HIV incidence of 11% instead of the target of 25%, and a decline in the transmission rate of 17% instead of 30%.

The researchers converted the annual transmission rate into the R0 number by multiplying by the latest figure for average life expectancy at the age of HIV diagnosis in the US, which is 28.86 years. This implies a fall in the R0 number from 1.028 in 2008 to 0.913 in 2010 and 0.754 in 2015.

They then did another analysis to take account of the fact that a proportion of undiagnosed people would transmit HIV, but die without ever being diagnosed and so would not appear in the HIV mortality figure. This would result in about 1900 more infections and 1900 more deaths in 2015 than in the unadjusted model, implying virtually the same HIV prevalence. But it would mean that since 2010 incidence has declined by 13% and the transmission rate by 19%, thus somewhat closer to the national strategy targets.

The transmission rate in this revised model would be somewhat higher, implying a rate of 0.0276 in 2015 and an R0 rate of 0.797.

Implications

This almost certainly overestimates the R0 rate because as the researchers point out, there is no reliable figure for how much of their life the average person with HIV spends virally undetectable and therefore non-infectious – especially as there is not even certainty about how many people are virally suppressed.

If this is the case, then the US may be getting close to the point where HIV prevalence starts to fall. This may not happen, however, if viral load is not suppressed in the groups most vulnerable to both acquiring and transmitting HIV, “namely young gay and other MSM, Black and Hispanic persons, and those living in the Southern US,” as the researchers note.

“The impact of the National HIV/AIDS Strategy appears to be important yet insufficient to achieve its key incidence and transmission targets however", they say.

"We must…[enumerate] the resources necessary to swiftly achieve its mission and maintain a laser-sharp focus on the needs of our most disproportionately affected and vulnerable communities.”

References

Bonacci RA and Holtgrave DR. Evaluating the impact of the US national HIV/AIDS strategy, 2010-2015. AIDS and Behavior, early online publication. DOI: 10.1007/s10461-016-1416-8. See abstract here. April 2016.