Medication non-adherence can be a tricky issue for healthcare providers, patients, and pharmacists to solve. In addition to cost, there are many other factors that influence whether someone takes their medication(s) as prescribed. As discussed in the literature, side effects, confusion about medications, and the general human tendency to forget may all play a role.
While medication is critical for many different conditions, lack of adherence after myocardial infarction has been shown to be particularly risky for readmission, additional myocardial infarctions, and death. Increasing financial pressure surrounding readmissions has generated renewed interest in the role of medication in readmissions for myocardial infarction. As much as 50% of medications are not taken as prescribed; potentially contributing to 125,000 deaths a year. Fortunately, there are many resources to improve adherence, some of which are listed here.
For researchers to sort out the effects of adherence and attempts to improve it, we need to have accurate measurements. Some measures involve asking patients directly, either specifically about each medication, generally about overall adherence, or by going through and counting pills/bottles directly. While these measures can work well for many purposes, they are limited in the population available. The proportion of days covered (PDC) is an alternative that is administratively measurable and widely used.
While electronic pill bottles have been around for quite a while, the technology has only recently become more affordable—signaling wider potential use. The wider use and integration of mobile apps that make behavioral nudges easier might also increase their effectiveness, though recent research has been mixed.
In a recently published Medical Care paper, Dr. Shivan J. Mehta and colleagues compare the claims-based (PDC) measure to adherence metrics obtained from electronic pill bottles for myocardial infarction patients. Adherence measured via claims was similar to adherence measured via electronic pill bottles, around 0.7. Multidrug adherence (adherence to statin, beta-blocker, and antiplatelet agent) was lower, around 0.45. Annual adherence showed a slightly different relationship, with PDC measuring higher levels than the electronic pill bottles.
Higher adherence was associated with increased time to first vascular readmission and lower all-cause readmission, though there were some differences based on medication type (statin only or multidrug) and measure. Both measurements of adherence had the same problems with censored and missing data for many patients. Additional work on which patients are missing data, and why, might help make the extent of bias concrete.
One of the main findings of this new study is that clinical outcome prediction was better when data from electronic pill bottles were included in models with PDC. We’ve covered other innovations in adherence on this blog, such as mobile apps, and comparing usage of medications across these additional platforms would likely increase our understanding of which barriers people face when taking medications and perhaps give some intuition about how they could be reduced.
Jess Williams