The trusted source for
healthcare information and
Drug concentrations as measure of ART adherence
'Adherence is far less than what people reported'
Adherence is one of the largest variables in the concentrations achieved of antiretrovirals (ARTs). So when HIV researchers are conducting clinical trials designed to determine efficacy of new ARTs, they cannot be certain how big a role adherence played in their results.
"If at the end of the study you don't get the results you were looking for, and it appears the drug did not work, then was it because the drug was not taken properly?" says Craig Hendrix, MD, director of the drug development unit and professor of medicine at Johns Hopkins University School of Medicine in Baltimore.
This has been a vexing question, particularly for microbicides trials, he notes.
"You gave the drug to animals and know they received the right concentration, and the animal model suggests the drug works," Hendrix says.
But it's trickier to interpret results when the drug is given to people, who might report having used it properly 95% of the time they are having sex, he adds.
"In some studies the objective measures show the adherence is far less than what people reported, and we don't know why that is," Hendrix says. "Recall is imperfect, and people want to see themselves as a certain kind of person."
Hendrix wanted to see if researchers could use drug concentrations as a measure of adherence in the context of clinical trials.
"The gold standard so far is the so-called MEMS Cap, which is a computer chip that goes into the top of the medicine bottle, and every time you unscrew it the cap registers the date and time the bottle is opened," he says. "That is great because that gives you information right at the time of the event."
Its drawback is that there is no way to be certain whether or not the person who opened the cap actually took the pill or took the correct dose.
Other adherence monitoring strategies include having a pharmacist check patients' bottles and count pills, having patients keep a diary, or having study subjects recall how many pills they took in the past week.
"Pill count is more objective, but it's subject to errors, as well," Hendrix says. "We want people to be good subjects, and it's hard to know what happens with all the medications."
The advantage to using drug concentration levels as a measure of adherence in either clinical studies or clinical practice is that this is the most direct way of measuring how much of the drug got into the body, he adds.
There are issues with inter-variability and intra-variability, he notes.
"I could give people pills for a month, draw their blood right before the next dose, and their blood could change 30% to 40%," he explains. "And if I took 10 people and drew their blood, I'd get 10 different concentrations, and they could vary several-fold because of variability between individuals."
There is a third source of variability, and this involves the assays, Hendrix says.
Hendrix and co-investigators are studying how drug concentration as an adherence measurement might work, and they have good, reproducible assays.
"I can take care of intra-variability if I observe a subject take a dose and see what the typical samples look like over time," Hendrix says. "We give them an observed dose, measure a few points in time, and doing that enough we can reasonably predict what the concentrations will be after a single dose."
If investigators know what a subject's concentration levels look like at two, four, and six hours after a dose, then they can predict what the concentration looks like in the steady state.
The next step is to do a paired analysis in which the subject's concentrations are compared between an observed dose and an unobserved dose.
The obstacle is this method costs money and resources and would be difficult to replicate in clinical care, Hendrix says.
"I'm hoping to show in a research setting that it's worth it," Hendrix says. "It's of tremendous value to rule out adherence as a reason for a drug's failure because you'd hate to throw out a drug that works perfectly fine if people take it."
The value of this approach might be less clear in clinical settings.
"We're a long way from having insurers pay for this as an adherence test," he says.