Tuesday, March 17, 2009
New Analysis of Earlier Adherence Studies Quite Revealing
In response to my earlier post about adherence (Health Myths #1: Adherence & Red Beads), I received an email from one of the leading researchers in this field pointing me to the most current analysis which dramatically changes the accepted wisdom on medication adherence, while adding support to the point I was making.
Numerous surveys of adherence studies have noted that adherence rates during clinical trials vary widely—43% to 78% according to the Osterberg & Blaschke paper I refered to.
Well, some people decided to go back and reanalyze data from past studies, looking at the data far more carefully than earlier researchers. Essentially they decided to break "overall adherence" amongst a population into two components: how well did people execute their regimen while trying to adhere to their medication regimen, and when did they decide to discontinue the regimen. What they found was that most "non-adherence" was actually due to people simply discontinuing the regimen -- they were no longer trying to take their medication. Of those trying to be adherent, only about 10% were unsuccessful on any given day.
This new analysis can be found in Adherence to prescribed antihypertensive drug treatments: longitudinal study of electronically complied dosing histories by Bernard Vrijens, et al., published 14 May 2008. Well worth reading for anyone interested in this topic.
This paper supports, quite strongly, my key point—which is that we must help people be more successful with adherence, rather than assuming that they don't care. As the paper says "Patients who execute poorly need help integrating their daily dosing into their routine ... our data suggest the value of ... helping patients to integrate their dosing into daily routines." This is precisely what we have found as well. People have difficulty integrating their health activities into their busy lives. Therefore, Zume Life's efforts, from the start, have focused on developing a tool that makes this integration easier. Many fundamental design decisions stem from this focus, including helping users with the full spectrum of their health activities (not just pills, but all medications as well as biometrics, moods, symptoms, food, ...), and having a mobile device as a critical element of the overall solution.
The paper also notes that "ongoing information on the quality of execution [to the medication regimen] might signify impending early discontinuation, possibly allowing an opportunity to intervene early to prevent treatment interruption". Absolutely!! Such early-warning indicators can have great benefit. If a person is tracking their overall health, there could be many such indicators, based on a wide variety of factors the person is tracking. For example, at the start of a new anti-hypertensive medication treatment, a person may choose to record various symptoms that are known side-effects—perhaps "chest pain", "weakness/fatigue", "breathing difficulty", and "lightheadedness/dizziness". Just a reassuring note from a caregiver observing the person's progress, such as "I see you're not feeling well. This is quite normal. It'll pass in a few days. Keep up the good working on sticking to your regimen", could be a very helpful intervention that keeps the person from discontinuation. Similarly, observing poor execution of another health activity scheduled for the morning could guide the doctor to an evening schedule for the new hypertension medicine.
Finally, the paper notes "implications for practical clinical management" including drugs with more "forgiveness". Presumably there are tradeoffs involved in making drugs that are taken just once a day versus multiple times a day, and between drugs that are more or less forgiving. We can change the drugs to make adherence easier, or we can improve people's capability to adhere so that "better" drugs (from a medical perspective) can be used. I imagine that it would be better to increase people's capability, so that drugs can be more finely tuned.
Although such studies are almost always very limited in their focus, often studying just one medication rather than a person's entire health regimen, it is encouraging that they support a more nuanced view of non-adherence. Perhaps it will help accelerate dispelling the myths surrounding poor adherence.