Posted on August 5th, 2011 No comments
Prevention of HIV infection by pre-exposure prophylaxis (PrEP) is not sufficiently effective
PrEP is a prophylactic intervention where uninfected people take anti HIV medications before sexual intercourse to prevent becoming infected with HIV. The use of a vaginal gel containing anti HIV drugs has also been tested.
The results of several trials of PrEP have been reported in the past year, all but one hailed as huge successes, with reported efficacies of up to 90% among those adhering to the treatment regimen.
The efficacy of PrEP in preventing HIV infection was reported to be so great that this intervention has been trumpeted as signalling a revolution in HIV prevention. A new era has opened up we are told; PrEP is a “game changer”.
With such enthusiastic coverage it may come as a surprise that none of the reports explicitly told us what the actual efficacies of the interventions were in preventing HIV infection. Perhaps because they were so low, as I’ll explain. Maybe what’s even more startling is that this omission seems to have gone completely unnoticed, at least in the universally jubilant press reports and equally enthusiastic press releases from AIDS advocacy organizations.
How has this been possible?
The reason is that the results have been reported as reductions in relative risk only. This tells you nothing about actual risk reduction. What is reported is a percentage reduction in risk from a number that was never clearly stated. For example in the iPrEx trial of PrEP among men who have sex with men, the drug, Truvada, was reported to reduce the risk of infection by 44%. But 44% of what? We were not explicitly told, although it’s possible to calculate what it is.
In fact we can calculate that the absolute risk reduction conferred by Truvada is a measly 2.3%, a number nowhere to found in the trial report. The relative risk reduction may have been 44%, but this translates into only an actual 2.3% reduction in risk, as is shown below.
Reporting relative risk reduction only is the oldest trick in the book to exaggerate the effects of an intervention, used by salesmen, but apparently also by clinical researchers.
What makes the unquestioning acceptance of these reports of relative risk reductions achieved by PrEP even more remarkable is that there is a tremendous amount of material explaining the difference between relative and absolute risk reduction. Just type the words “relative risk absolute risk” into the Google search box.
Relative risk reduction tells you the percentage reduction in risk in the treated group compared to that in the group receiving placebo, or how much lower the risk with the intervention is relative to the risk to begin with.
If you are not clearly told what the risk is to begin with, then you can’t tell what the actual reduction in risk is when taking the intervention; all you know is how much lower it is than a number that’s not clearly presented to you.
Although not included in the iPrEx trial report there is information that allows one to calculate the absolute risk reduction conferred by Truvada. To do this we need to know what the risk of infection is to begin with.
This is the number of infections occurring in the placebo group over the time period of the study.
64 out of 1248 people in the placebo group were infected, which is 5.1%, or 0.051 in 1. (since then there have been additional infections reported at the Rome AIDS conference, reflecting an increase in the number of infections over a longer time period).
In the group receiving Truvada 2.8% of 1251 people were infected.
The absolute risk reduction conferred by Truvada is simply 5.1 minus 2.8 which is 2.3.
A 2.3% reduction in absolute risk conferred by Truvada is the more accurate measure of its efficacy. Hardly something to celebrate.
A 44% reduction in relative risk sounds much better, although far from spectacular, but unfortunately it’s a number that tells you nothing about actual risk reduction.
Relative risk reduction is calculated as follows:
It is the number of events in the treatment group subtracted from the number of events in the placebo group divided by the number of events in the placebo group.
On its own, relative risk reduction is not a helpful number.
Of much greater help to a person considering Truvada PrEP is knowledge of the actual risk while taking Truvada (over the period of the study, a median of 1.2 years).
That number is 2.8%.
Knowing the absolute risk reduction allows one to calculate another important measure. This is the number of people who need to be treated to prevent one infection (NNT). From information contained in the trial report 45 people need to be treated to prevent one infection. I did not notice this number in the trial report nor was the absolute risk reduction of 2.3% reported. NNT is a useful number as it allows one to estimate what it would cost to prevent a single infection with Truvada.
The cost of the drug is the least of it. A person taking Truvada PrEP needs to be monitored at regular intervals for toxicity and importantly, for infection, in order to avoid the inevitable emergence of resistant viruses as a result of sub optimal treatment.
If Prep is implemented on a large scale which some AIDS advocates seem to be calling for, but is unlikely to happen, then there may well be increases in new infections with viruses resistant to the drugs in Truvada in men who have sex with men, in IV drug users and in Africans.
PrEP is not a success, at least not with Truvada. However such a failure was transformed into a triumph, part of the explanation is the use of relative risk reduction numbers with care taken to remain silent on absolute risk reduction. Despite all the literature available to help people tell the difference between absolute and relative risk reduction, this evidently was a resource not used by those cheering along this ineffective intervention.