Post-study interviews and computer questionnaires conducted with former participants in a trial of pre-exposure prophylaxis (PrEP) that reported zero effectiveness show that participants concealed their low adherence to the study pills because, despite reassurances from researchers, they feared they would be asked to leave the study.
The interviews reveal that some participants went to elaborate lengths to disguise their low adherence – underlining that, while they had doubts about the value of the intervention on offer, they greatly valued the medical care that the study offered.
The researchers comment that “The appropriate level of ancillary care to be provided in biomedical HIV prevention clinical trials has been a longstanding ethical discussion in the field. Individuals may continue to enrol in clinical trials as a means to obtain ancillary healthcare until these needs can be otherwise sufficiently met.”
The FEM-PrEP study was a randomised placebo-controlled trial of tenofovir/emtricitabine (Truvada) as PrEP conducted in 2120 women in South Africa, Kenya and Tanzania. It was stopped early in April 2011 when an interim analysis revealed it could not yield a positive result; there was no difference in the HIV infection rate between the Truvada and placebo arms.
Trial participants had consistently reported their adherence as high (95% said they had ‘always or usually’ taken their daily pill) and pill counts suggested adherence of 88%. But drug level monitoring told an entirely different story. When FEM-PrEP’s results were first presented, results were already indicating that less than 40% of women had taken a pill in the last two days and final figures showed that only 12% of participants consistently had tenofovir levels indicating daily adherence. Twenty-three per cent took Truvada rarely or not at all, while the remainder took it inconsistently and rather occasionally.
Studies after the trial revealed that the women did not take the PrEP pill largely because they thought it would cause side-effects, that it would be ineffective (especially as they knew it could be a placebo) or even that it could increase the risk of HIV. One analysis (Corneli 2014) showed that individual doubts were reinforced by negative attitudes held by other trial participants, the women’s partners, and wider disparagement of the trial in the community. HIV stigma also played a part too: some women were afraid to be seen with the pills and be thought to have HIV.
Another study, on the other hand (Wong), found that women greatly valued being in the study for reasons that were nothing to do with PrEP. The most common reasons for joining and participating in the trial were the monthly HIV test (32% of participants said this was an important reason for joining the trial), the support the trial gave them to stay negative (25%) and the monitoring offered for other health problems (22%). There was evidence that some participants felt that simply being in the trial would protect them from HIV.
This study – reporting adherence
These earlier qualitative studies did not show, however why so many participants felt they had to conceal their lack of adherence from researchers. So the researchers conducted 88 qualitative, in-person interviews with participants who had been in the Truvada arm of the study and 224 computer-assisted self-interviews with participants who had taken either Truvada or placebo.
During the computer-assisted interviews, 31% of the participants admitted that they had over-reported their adherence during the trial – still not as many as actually did under-report it, but far more than reported low adherence at the time.
The most common reason given for over-reporting adherence was that participants feared they would be excluded from the trial. This was despite repeated reassurances by researchers that they would not be penalised for not taking their pills. The researchers speculate that a standard clause in the consent form saying that participants could be excluded for protocol violations might have influenced some participants – though in the event none were excluded.
The second most frequently reported reason was simply that it was easier to say “Yes, I’ve been taking my pills” and the third most important was that they thought the trial staff would be disappointed in them.
There were differences between the South African and the Kenyan participants who were interviewed for this study. Participants in Kenya were considerably less likely to admit to over-reporting adherence in general (20% versus 43%). For those who did, the most frequently cited reason for doing so was wishing to avoid being reminded or counselled to take pills, rather than fear of trial termination. Not disappointing staff was also important to them.
The researchers comment that the general reluctance to admit over-reporting even in a confidential computer interview, the fear of disapproval it revealed, and the difference between Kenyan and South African replies may all reflect cultural attitudes to authority and avoidance of confrontation as much as they reflect the specific issues mentioned.
What did the participants do with their pills?
This is also reflected in the answers people gave when asked what they did with the pills they did not take. Eighty-three per cent of participants said they brought the pills back to the clinic, even though only 12% were actually brought back. The researchers comment that the participants may have thought that the question asked whether they had brought some of their pills back – and, as we see below, many took steps to bring back carefully counted amounts.
Thirty-five per cent of participants (again, fewer in the Kenyan group) admitted to discarding their pills. They mostly threw them in the toilet; fewer threw them in the trash or in the bush on the way from the clinic. This is partly because participants had been told the pills could be toxic and they did not want children to mistake them for sweets. By the time these interviews were done, participants knew whether they had taken Truvada or placebo, and more on Truvada said they had thrown pills away than those on placebo – maybe to reassure researchers that they had disposed of them responsibly.
Very few participants reported giving the pills to someone else – only 4%, roughly split between giving them to friends/partners with HIV and to people with other illnesses they thought the pills might cure. Only two people said they had sold them. More, in fact, were highly critical of someone who might do such a thing: one participant said “How can you sell to someone a pill that you do not know if it works?” A slightly higher proportion stockpiled their pills in case the trial produced a positive result or in the event of their becoming HIV positive. Fearing the future costs of HIV treatment was one consideration here. One participant said:
“Most were throwing them in the pit latrine. But some were storing them in their houses, thinking that after some years, the drugs will help them. So they will not be buying those drugs, they will just take them.”
During the 88 semi-structured interviews (SSIs), more women (63%) were willing to describe what they had done with their pills (63%). In addition, to get round lingering fears of interviewer disapproval, participants were also asked what they had heard that other study participants had done and 84% reported that. A quarter of participants in the SSIs admitted stockpiling pills; however giving their pills to others was still rarely reported, with only 5% of participants saying they did this. No-one in the SSIs said they had sold pills.
The SSIs revealed that some participants had gone to elaborate lengths to conceal low adherence. A sizeable minority (34%) volunteered the information that they would count their pills before discarding them and returning a certain number to the clinic so they agreed with their self-reported adherence. One said:
“Let’s say after my appointment I took pills two or three times. Then I count the days to the next month and sort out the pills, like ‘Okay, I was supposed to have taken this many pills, this much must be left over’. So that’s what I did.”
Again, the considerable effort taken over maintaining the appearance of high adherence underlines how keen the women were to remain in the study and get the medical benefits it offered.
As the researchers say, “The value of participant-dependent measures [of adherence] cannot be enhanced without consideration of the broader perceived or real environment in which the data are collected.”