The Digital Apothecary

View Original

Medication Adherence is NOT a $300 Billion Problem

Medication adherence holds a soft spot for me, likely due to my professional background as a pharmacist. I cannot tell you how often medication adherence comes up daily when addressing patient care. Even in the home healthcare setting, we have multiple nurses that are dedicated to visiting and administering medications to patients (usually those with a psych history or SUD or related comorbidities that impede medication use). I love the idea of technology aiding in medication adherence, and if you follow my work, you know that I have been reviewing apps and devices for almost a decade now. Nonetheless, I cannot but help notice a trend that is, I believe, really cutting out pharmacists from the medication "adherence" disruption space.

Heres the problem: Medication adherence is not a $300 Billion problem, and that may shock you. After all, almost every company website I visit discusses how they will transform healthcare with their solution/product by focusing on their intervention will reduce the cost of adherence in the US. And I cannot help but think this narrative is killing the pharmacy profession because it opens up a very different tale to the actual problem that is favorable to only a few people.


Why the Current Medication Adherence Argument is Flawed

Before I get into the literature and assessment, let me explain my concern. For decades, the pharmacist profession has advocated that they should be the de facto medication expert and geared towards addressing medication issues (including adherence) on a clinical level amongst patients - while being renumerated for it. For many people outside of the direct healthcare space, this may sound like a good thing, and there has been buyin to help empower pharmacists to perform these services. That's the problem, though, then you need to dedicate staff and resources to do this, and a payor would then need to cover this expense, which could possibly increase cost.

But, if adherence can be solved with technology, and with the integration of augmented intelligence, etc., why need humans [Pharmacist] involved? It sounds like a no brainer and seems to be a quick and seamless solution. After all, that is why we see a push for chronic condition management coming out with companies like Livongo, Omada, and others. So why not just assign an app or device to a patient, and call it done.

Well, if you follow that logic, then it makes sense to start developing this stuff and trying to get it to market. So, that is why we have seen an explosion of companies, products, services, and now clinical studies finally to see if they work. And most of those studies have not been positive, to small or pilots/feasibility studies to be honest (though there are some good ones and I do have high hopes). Throwing out an app or device is not leading to clinical endpoints. And when I read the limitations and discussions of these papers, I can't help but sigh.


The Number Problem & Current Evidence

Where then did the $300 billion comments come from? I almost feel like its the '10,000 steps' a day issue that many have discussed have no scientific basis. Well, the $300 billion does have some merit, but not in adherence alone.

Looking through papers that reference the cost of nonadherence, I keep coming across a document from NEHI. It plainly states that there is a $700 Billion problem in medicine and that $290 Billion are due to adherence (which can be solved multiple ways). It seems simple enough and pretty eye-catching, and is usually just rounded up to $300 Billion because why not.

But there is more to this. You see, NEHI or the New England Health Institue has a full document on this topic in a research brief published in August 2009 called "Thinking Outside the PillBox - a system-wide approach to improving patient medication adherence for chronic disease." When you go the Appendix I: Estimated Cost of Poor Adherence on page 17-18, you can actually see their methods, and it paints a different story. A critical section I will highlight as follows:

"The study estimated the likelihood of a patient experiencing one or more drug-related problem (DRP) in the ambulatory care setting and the cost of the subsequent negative outcomes. Specifically, DRPs included untreated indication, improper drug selection, subtherapeutic dosage, failure to receive drugs, overdosage, adverse drug events, drug interactions, and drug use without indication. The study did not delineate poor adherence from other DRPs, so the estimate includes the overall impact of all DRPs. There are five possible negative outcomes in the Johnson and Bootman model that create additional costs to the system (the two that do not are death and no treatment): an additional physician visit, additional treatment, ED visit, hospital admission or LTC admission. We replicated the Johnson and Bootman method for determining the number of events by multiplying the cumulative conditional probabilities for each of the six outcomes by the 2008 number of total physician visits estimated by the CDC, which was 901,954,000. The results of this calculation are listed in the table."


The table is also fascinating, as the breakdown highlights the different areas that lead to the $290 Billion cost:

So there you have it. The cost is not nonadherence, rather, Drug-Related Problems. For many outside of healthcare, this is a very different language. What are DRPs? For us in pharmacy, it is probably one of the hallmarks of how we identify and related our interventions. We even have national competitions on where students from schools gather to see if they can identify DRPs correctly based on clinical scenarios (here is an example). DRPs vary, but some include:

  • Adverse reaction to therapy (identified through patient communication)

  • Drug choice problem

    • Inappropriate drug choice (not guideline appropriate or other reasons)

    • Therapy duplication (two drugs in the same class)

    • Contraindication (such as a disease problem or allergy)

    • Lack of indication (drug present with no medical history indicative of need)

  • Dosing Problem (might be too often or to much or little [dose for a child or perhaps for a patient with poorly functioning kidneys])

  • Drug utilization issue (such as adherence, or other problems [the patient has poor vision and cant measure how much insulin to give themselves])

  • Drug Interactions (which could cause the drug not to work or cause other problems for the patient [can be with other drugs, food])

  • Failure of therapy (the drug isn't working, so try something else)

The problem is, when you start saying all of these issues that come up with medications, you kinda lose a person's attention. It really isn't easy to put a slide together saying, "Medication issues are a $300 Billion problem due to XYZ and my product can solve one/some of them" versus "Medication adherence is a $300 Billion problem, and my product can fix it." I don't expect investors or many people in business to know this, and quite frankly, most health professionals I suspect wouldn't either. 


It's not Medication Adherence We Should Focus on, It's Optimization

Interestingly, a study that came out in 2018 the Annals of Pharmacotherapy by Watanabe and colleagues titled "Cost of Prescription Drug–Related Morbidity and Mortality." When I initially saw the article, I thought it would be another piece highlighting issues with costs and nonadherence, but a section of their introduction immediately grabbed my attention:

"However, the cost associated with drug use reaches beyond the purchase of prescribed medications to encompass additional medical costs of morbidity and mortality resulting from medication regimens that are not optimized to effectively treat the indication resulting in a treatment failure (TF), where the resolvable medical problem is not adequately treated, a new medical problem (NMP), where a newly prescribed medication causes or contributes to an incident clinical symptom or syndrome, or both a TF and NMP. This cost has most recently been estimated as $290 billion equating to 13% of total annual US medical costs in 2008. Although widely misdescribed in the published literature and policy documents as the cost associated with “patient nonadherence to medications,” this estimate and the preceding estimates ($76.6 billion in 1995 and $177.4 billion in 2000) actually reflect medical resource utilization caused by TFs and NMPs that arising from nonoptimized medication use. Nonadherence to the indicated medication regimen is just one of multiple potential causal factors leading to a TF, resulting in downstream health services use."


Finally, a paper highlighting the real issue of medication nonoptimization. I won't bore you with their analysis and how they got to their estimate (it is worth a read if you want to get into this business), but suffice to say, their estimate now pegs the issue as a $500 billion problem.


So are Adherence Companies on the Wrong Track?

No, absolutely not, and I do not want that to be your takeaway. Instead, what I would want you to consider, if it hasn't come across, is that adherence is a part of the issue with the optimization of medications. Not knowing if a patient is taking their medications appropriately causes a lot of problems that can lead to many downstream costs. But, a device that tracks adherence won't let us know about adverse drug reactions, if a patient is dosed inappropriately if the therapy is indicated, or other DRPs. That takes some medical training to determine which pharmacists are more than capable of. Rather, I love these companies because they are creating tools that can help us push medical care beyond what we are generally capable of. We need these innovative tools and services.

So Why Pharmacists?

Ok, so if you've followed this, then you can see that DRPs are a huge problem, which includes adherence. Who best to manage all of this but a pharmacist? We know drugs, pharmacotherapy, guidelines, clinical practice, can understand disease treatment, etc. We are trained to help patients struggling with therapy, and can work in a clinical team to help other providers guide care. Adherence tools fall into our wheelhouse to help assess adherence. Let pharmacists monitor adherence and make assessments from wearables on DRPs and treatment. Physicians don't want to be bothered by this, we know that. But pharmacists would take it up gladly. But pay us for it. Involve us in the design aspects and consider how the pharmacist can be the go-to person to make your product or service a reality. Pharmacists have been providing Medication Therapy Management (MTM) and Comprehensive Medication Management (CMM) for years, but it has not been a profitable business and hasn’t taken off. One reason is that it is a logistical nightmare with data collection and is time-consuming. Spending 30-60minutes to do an MTM to be paid $40 isn’t worth it when most pharmacists are paid >$50/hr. BUT, now we have many new startups using augmented intelligence platforms and data analytics to cut that time down, which I think can really lead to more CMM/MTM services that can solve this $500 Billion problem.

I think that is why the vast majority of trials using digital health adherence tools are failing to meet clinical endpoints. Adherence, in itself, will not deliver better clinical care. The data that can be gleaned from that data analyzed and interpreted and implemented by a pharmacist, I think, would lead to the transformation everyone wants, and other startups are really opening that gate for us.

The bottom line is that the $300 Billion quip is just ignoring the real uncomfortable problem that is more complicated and nuanced than many want to address, and pharmacists should challenge this issue, otherwise, it just leaves us out of the conversation for quick and fast tech solutions that won’t cut it alone.