How to track Medication Adherence with Digital Health Technologies
A recent article published in JAMA caught my attention and I finally had the time to read and process it. The article, “Interventions to Improve medication Adherence: A Review” by Kini and Ho, was quite well done in my opinion, capturing the issues that surround medication adherence as of late. They did a review on PubMed articles published from 2000 - 2018, and found a total of 49 trials that met their criteria (which were acceptable I feel ). Perhaps my biggest critique of their methodology was the interventions to track adherence, bypassed some novel means of tracking adherence such as mobile apps and digital medicines, but that could also be due to the factor of most of those types of trials are either (a) newer or (b) feasibility studies and didn’t meet their criteria. Still, I feel would have been nice to touch upon.
In any event, one item that gave me some thought as they reviewed the 49 studies was the mechanism used to track adherence by most researchers. The focus was high on claims data and electronic pill bottles. Some other data points included blood concentrations of therapy, biomarkers (e.g., BP, LDL, INR). However, the outcomes focused on included either a change in fill rates (from claims data) or clinical outcomes associated with disease progression or viable biomarkers. What struck me was there wasn’t a lot of positives across the board, and even the authors commented that there was no ‘Gold Standard’ of what adherence was for patients.
Now, the authors discussed what this means for clinicians who address adherence in their patients. They suggested a multifactorial approach using the following to help encourage adherence in patients including (personal thoughts in italics):
Patient Education - Such in a patient with a newly prescribed agent to show clinical outcomes of adherence (e.g. lower BP = lower CVD)
Medication Regimen Management - Consider combination pills - Its a shame they didn’t address adherence packaging here which I think is something worth discussion such as multi-dose packages or strip packaging a la PillPack.
Clinical Pharmacist Consultation - Goes without saying in my mind, but I think we should still bill for this service.
Cognitive Behavioral Therapy - Motivational interviewing techniques they suggest using a psychologist (which based on the studies used was occurring) - But I would argue a health professional that is trained could also offer such a service.
Medication-Taking Reminders - Staff calls from a pharmacy that are personalized and not generic.
Incentives to Promote Adherence - Reward platforms or gamification.
Overall, I would say that their arguments are sound, but still locked into a methodology of thinking that I feel is being outpaced due to what we currently have available through technology and what is coming down the pipeline and was missed.
Where does digital health fall into medication adherence?
One thing that the authors missed was the use of digital medicines and alternative means of tracking/encouraging adherence. For instance (as seen in the figure above), we have a plethora of sensors that patients can now purchase to attach to certain pharmaceutical products to help remind them to take their medication and can also track when utilized. But, even bigger I feel is the inevitable combination of medication and sensors that will turn into a an all in one package. This isn’t going to happen overnight though, and I would say it will be another decade before they are virtually commonplace, but we are getting there.
So, the figure I made addresses where we can use technology to track adherence. The key thing I am pointing out here is tracking and not modifying adherence. I do not think we are there where technology alone can be a digital therapeutic (maybe with time though) that can change adherence alone. Rather, I support the use of digital technologies as an adjunct tool to track actionable data to be utilized by a health team to intervene on medication adherence with objective data points. Some of these measurements are softer than others, and they all have different logistical barriers or costs associated with them. I am considering in a future post outlining which intervention to choose in a scenario, though it would be all supposition at this point as we lack any studies or literature at this time to say what is appropriate or not. Rather, we function more based on what makes sense from a workflow and cost perspective at this time.
Feel free to leave any comments or suggestions!