HIV and ID Observations

In a previous post, we reviewed the various flavors of medication non-adherence, and concluded with this tantalizing line:

Next up:  An Adherence Intervention that Actually Works — But There’s a Catch

Well here it is, just published online in JAMA Internal Medicine.

Dr. Robert Gross (a long-time HIV adherence researcher from U Penn) and colleagues enrolled 180 patients with HIV RNA of 1000 or higher who were starting or changing their HIV regimen. They were randomized to usual care versus an adherence intervention called Managed Problem Solving, or “MAPS”. The study took place in three outpatient clinics in Philadelphia; study subjects could be treatment-naive or treatment-experienced. Participants were predominantly black, and 60% were men.

So what constituted MAPS? A trained “interventionist” — someone who went through 15 hours of study-specific education — met with the patient for 4 in-person sessions, some as long as 90 minutes; this was supplemented with 12 subsequent phone calls and then additional monthly calls for a year.

The good news is that the intervention worked: Those in the MAPS group had a 1.78 (95% CI, 1.07-2.96) times greater chance of being in a higher adherence category than those receiving usual care, and there was a strong trend toward a higher likelihood of virologic suppression as well. Of note, 36% of the MAPS group and 26% of those in usual care dropped out of the study, a rate substantially higher than in most clinical trials of antiretroviral therapy, strongly suggesting this was a difficult-to-treat patient population.

So what’s the catch? To say that the MAPS intervention would be difficult to institute in a non-study setting is a colossal understatement, as the vast majority of HIV clinics could never deploy the training, the personnel, or the time for this kind of extra care. While the study authors argue that this intervention would nonetheless meet criteria for cost-effectiveness, such analyses typically take a societal perspective, and would not make MAPS feasible in a clinic with a limited budget.

The solution? Editorialists argue for “real time” adherence monitoring using electronic pill containers connected to a central server through mobile networks. When in place, clinicians could detect individual lapses in pill-taking behavior, and intervene only for people who need it. (Presumably they could also use portable jet-packs to go visit those who need a house call.)

I do think this paper shows some progress in addressing the tricky problem of non-adherent patients (at least those who show up for care). It also demonstrates just how hard it will be to improve adherence, at least until we have some way of identifying ahead of time — or in real-time, as suggested by the editorial — people who are bad at taking their meds.

Until then, enjoy this nifty 2-minute video on the HIV cascade from the CDC — your taxpayer dollars at work!