Can Your Electric Bill Show Us How to Improve Medication Adherence?

Dose of Digital Mini White Paper

I’ve talked about medication adherence a few times on this blog including some thoughts on how to improve it (most recently: “The Only Way Pharma Can Improve Compliance: Fun“). [Quick disclaimer: I'm using "compliance" and "adherence" a bit interchangeably in this post. I know they're different, but I'm going to spare a big debate. Just go with it. Thanks.] It’s certainly not an easy challenge and one that no one has figured out yet. There are certainly incremental improvements that have been made over the years, but we have yet to discover a “magic bullet.” But perhaps there isn’t one. Instead, maybe there are a number of different “cures” with one working for one person and another working for someone else. If we believe this is the case, then the only way to find these “cures” is to try something. In fact, we’ve got to try a number of different things, see what works, and throw out what doesn’t.

One area where I think a big impact can be made is in game theory. That is, using the concepts and theories of this field of study to help improve adherence. I’m going to cover this concept in more detail in an upcoming post, but the short version goes like this (via Wikipedia): “Game theory is a branch of applied mathematics that is used in the social sciences, most notably in economics, as well as in biology (most notably evolutionary biology and ecology), engineering, political science, international relations, computer science, and philosophy. Game theory attempts to mathematically capture behavior in strategic situations, in which an individual’s success in making choices depends on the choices of others.”

Classic Wikipedia definition, so here’s the English: Don’t worry about anything but the last part of the definition.  As you likely have observed, we humans don’t make perfectly predictable or rational decisions when it comes to health, but game theory could be used to figure out that these decisions might not be so irrational after all. There’s certainly application when it comes to medication adherence where even for life-threatening conditions, adherence rates remain startlingly low.

For those of you who read the last part of the game theory definition carefully, you might be a bit puzzled as to why this applies to adherence. It ends with: “…an individual’s success in making choices depends on the choices of others.” What do the choices of others have to do with your behavior especially when it comes to adherence? Again, more to come in future posts on this topic, but the short explanation is this: the decision as to whether or not to take your medication as prescribed is equally influenced by factors you develop for yourself as factors that others develop. In other words, the reasons you use to justify whether or not to take your medication come in equal parts from your beliefs and perceptions as they do from the people around you. You’re not the only one who determines whether or not you take your medication. Other influencers include spouses, children, doctors, nurses, pharma companies, friends, pharmacists, and many others (and in different orders for different people). Yes, ultimately, you have the final decision, since you’re the one putting the pill in your mouth, but you aren’t the only one with a say in what you do.

The question then is this: which behaviors of others have the biggest positive influence on a person’s adherence?

This is where the real gaming aspect comes into play. As humans, we are genetically programmed to want to be winners. There is no second place in evolution. For that reason, many of our behaviors are geared to ensure our genetic success. This is why find spouses and take care of our children. It’s a primordial drive in all of us. In most cases, we strive for our genetic success over others. This is far more pronounced in the animal kingdom where you can watch males square off (sometimes to the death) for the right to mate. We humans don’t exactly do this, but sometimes we do either literally or figuratively (see any reality TV show for examples).

Since we’re programmed to want to come in first, that means we’re similarly programmed to not want to come in last. This is one of the big motivators in game theory and the explanation behind the problem, Prisoner’s Dilemma.

The question is, can this same fundamental theory apply to medication adherence? I think it can. But first, you should know where this idea came from.

I recently read an article in USA Today, “Do you use more energy than your neighbors?”, that talked about how some electric companies were trying some innovative things to reduce their customers’ use of electricity (interesting…wonder if pharma companies would be interested in helping reduce their customers’ use of their drugs?). There are a few reasons why the companies are interested in doing this, but I won’t get into them here. It also turns out that these companies have tried a number of things, but finally hit on an idea that involved comparing customers’ electricity use to that of their neighbors. This is the basic idea of what it looks like:

(Image from: Treehugger)

Well, imagine that this was your electric bill. What would you think? Would this be enough to get you to make a change? It probably won’t make you run out and buy solar panels for your roof, but I bet it will be in the back of your mind when you leave a room and don’t turn off the lights. This bill would be pretty hard to ignore for most people and most would make some minor changes even if the changes were subconscious. Game theory at work…no one wants to be last. This is especially true when your identity is attached to last place. Of course, here the information is anonymous (i.e., no one knows which specific neighbors are included), but you know where you stand. While you don’t like to be in last, you also don’t want someone else to be in first place. The higher your use of electricity the easier it is for everyone else to exceed the average (mean) performance. After all, you’re the one skewing the average much higher. So, you reduce your usage to both move yourself up in the standings and also to make it harder for others to achieve what they want. This, of course, motivates others to perform even better. In cases like the electric bill where it isn’t a zero-sum game, one person winning doesn’t mean that everyone else loses, there are grades of “winning.”

This innovative idea was presented at a recent TED conference and you can see the presentation below if you want to find out a little more. However, for now, know that this program has reduced electricity use by around 2-3%. As USA Today put it: “While that may sound small, the savings add up. The Sacramento Municipal Utility District, which started sending the reports to 35,000 households in 2008, says the households saved enough energy in a year to power 800 homes for a year.” That’s meaningful. Over time, I’d expect that these numbers only improve as people do more to try to be among the most efficient.

Here’s the TED talk:

Now, let’s apply this same concept to medication adherence. Let’s say that we knew you were on a chronic medication and how often you took this medication. What if you got a report each month that showed how often you took your medication? What if this report not only had your adherence rate (actually compliance rate, I suppose), but also that of others taking the same medication? If you found out that your rates were well below the average, what impact might that have?

Of course, it’s not this simple…here’s why…

First, you need a way to track someone’s adherence. I’ve already told you that people don’t really use trackers, so we’d need a solution for that piece. Among big reasons why people don’t use these trackers is because they’re hard to use and there’s not a real motivation to use one. Plus, they aren’t fun. Now, if you knew you were in a competition against others and the only way to ensure that you got proper “credit” in this competition is to track your actions, would that motivate you to track your actions? The answer is “it depends,” but, to be sure, this is a proven, effective model to get people to track their actions. Look no further than foursquare. Make no mistake, foursquare is a game. Period. The only reason why people get so religious about checking in is to steal mayorships from others and maintain their own (and get badges). Take mayorships away from everyone and I guarantee that foursquare goes away overnight. So, perhaps by adding a little competition to the mix, you can get people to track their medication use. Alternatively, you can use pharmacy refill data as a surrogate for adherence.

Second issue…privacy. All of this, of course, would have to be opt-in…or would it? If all the data is de-identified and pooled, there isn’t a privacy issue. And if people are actively using some tracker, then you can ensure that there’s a proper opt-in. If you can show that people get a benefit from giving up a little privacy, you’ll also be successful.

A few more details we can consider: what if the report that people got wasn’t just the numbers, but also included tips for how to improve your adherence and also medication-specific stats on why taking the medication is important (e.g., reduced mortality rate by X%).

The question then becomes: is seeing your rates compared to others enough of a motivator to get you to change your medication adherence? It’s probably not for most people. While we don’t want to come in last place, we also choose not to be involved in some “contests.” Just having a higher rate probably isn’t enough bragging rights to make it worth the effort. So, let’s up the ante. What if the standings weren’t in compliance rates, but in years left to live instead? On average, we can show for some drugs that there’s an improvement in mortality or other outcomes by taking your medication as prescribed for as long as your doctors deems appropriate. So, instead of showing that you took your medication 45% of the time and the average was 65%, we show that your additional years of life gained from taking the medication was 1.2 and the average was 2.5. That’s a game worth winning.

This wouldn’t be an easy system to implement, but it could be done reasonably simply via smartphone applications. There’d need to be a little data crunching to get to the outcome data, but it could be done for a lot of medications (just look for the pharmacoeconomic data to get started). Stay tuned in future posts for more of my thoughts on how we can use basic game theory to improve all sort of things in healthcare. In the meantime, feel free to give me your thoughts and turn off some lights. You don’t want to be in last place, do you?

Tags: , , , , , , ,

20 Responses to “Can Your Electric Bill Show Us How to Improve Medication Adherence?”

  1. Craig May 5, 2010 at 5:31 pm #

    From your headline I thought you were going to go down the path of the Ambient Orb and Glowing Bottle Caps.

    Ambient Devices makes an “orb” for California PG&E energy customers; it glows different colors based on energy rates to motivate “compliance” with energy conservation. Glowing a bad color? Turn down the A/C.
    http://www.pge.com/drorbs/

    The founder of Ambient is CEO at Vitality, maker of the GlowCap (aka the “Glorified Alarm Clock” from your Dec08 post).

    So then the GlowCap should change color based on a competitive assessment of your compliance vs. the others in your social network?

    • Jonathan Richman May 6, 2010 at 12:54 pm #

      I think this would make sense. I know that GlowCap already has some sort of glowing light that you plug in to remind you if you missed your dose, so adding this additional enhancement would make sense. It also seems logical to add this sort of information to the regular reports that they send out to users on their adherence rates. They’re probably in a good position to test my theory.

  2. Craig May 5, 2010 at 1:31 pm #

    From your headline I thought you were going to go down the path of the Ambient Orb and Glowing Bottle Caps.

    Ambient Devices makes an “orb” for California PG&E energy customers; it glows different colors based on energy rates to motivate “compliance” with energy conservation. Glowing a bad color? Turn down the A/C.
    http://www.pge.com/drorbs/

    The founder of Ambient is CEO at Vitality, maker of the GlowCap (aka the “Glorified Alarm Clock” from your Dec08 post).

    So then the GlowCap should change color based on a competitive assessment of your compliance vs. the others in your social network?

    • Jonathan Richman May 6, 2010 at 8:54 am #

      I think this would make sense. I know that GlowCap already has some sort of glowing light that you plug in to remind you if you missed your dose, so adding this additional enhancement would make sense. It also seems logical to add this sort of information to the regular reports that they send out to users on their adherence rates. They’re probably in a good position to test my theory.

  3. Amber Benson May 5, 2010 at 5:55 pm #

    Jon,

    Great post! I thought I was the only one who read game theory for fun. I think the folks over at VitalityRx are definitely thinking in this direction. Check out their minute-by-minute adherence behavior map: http://vimeo.com/6051194. Imagine being able to see outcomes in a similar fashion.

    I think the combination of device-reported and self-reported data could finally give us a complete picture of adherence behavior.

    Amber

    • Jonathan Richman May 6, 2010 at 12:55 pm #

      As Craig mentioned in his comment and I said in my reply to him, I think there’s a good opportunity here with Vitality to test this theory of mine and see what impact the “competition” might have on adherence. I’d be willing to bet on an increase.

  4. Amber Benson May 5, 2010 at 1:55 pm #

    Jon,

    Great post! I thought I was the only one who read game theory for fun. I think the folks over at VitalityRx are definitely thinking in this direction. Check out their minute-by-minute adherence behavior map: http://vimeo.com/6051194. Imagine being able to see outcomes in a similar fashion.

    I think the combination of device-reported and self-reported data could finally give us a complete picture of adherence behavior.

    Amber

    • Jonathan Richman May 6, 2010 at 8:55 am #

      As Craig mentioned in his comment and I said in my reply to him, I think there’s a good opportunity here with Vitality to test this theory of mine and see what impact the “competition” might have on adherence. I’d be willing to bet on an increase.

  5. Jaimie Brown May 6, 2010 at 12:50 pm #

    The main problem I have with adherence or compliance is that they are both reactive behaviours based on an action that has been decided upon by a third party. The ideal desired patient behaviours is ‘concordance’, whereby the patient has a defining role in their treatment by being involved in its choice right from the start.

    There is no easy way to achieve this, particularly in Europe, since patients are generally not payors and the right to accurate information on medication is denied most people on the grounds of cost control.

    Secondary to this is the fact that although game theory is based on the desire to win, human evoltion and our improving ability to control our environment has fuelled our belief that we are largely protected from bad things happening. Being initiated on a statin at the age of 45 because your physician tells you that you might otherwise fall ill in 20 years’ time is not especially compelling. I admire your idea of presenting data in the form of ‘years of life left’ although find it hard to think of any pharma company that would be allowed to promote such measures or regulatory body happy to approve it either. The bragging rights of being on the right hand side of the NDC would be phenomenal, though, no matter how perverse.

    • Jonathan Richman May 6, 2010 at 1:03 pm #

      I’m with you on the terminology, Jaimie, and tried to give a caveat that “adherence” certainly isn’t the ideal term, but we have to pick something. Concordance works in most situations, especially upfront when a treatment is being selected. However, after this point, I’m not sure it does. When you get home and decide whether or not to take your medication, you have to decide whether or not to “adhere” or “comply” with what you and your doctor reached “concordance” on. It’s a bit tricky and I don’t think there’s a perfect word.

      I agree that humans are bad at assessing risk especially when it applies to their own health. There’s no other way to account for the poor health practices most of us follow. For many people, whether or not they adhere (or concord) with treatment will be decided before the first pill. It’ll be based on whether or not they believe the benefits of the treatment outweigh the risks (and hassles) and if they can understand how to properly evaluate this information. If this isn’t done correctly, all bets are off.

      Regarding a pharma company presenting the years of life gained, I agree that it would be difficult, if not impossible, for them to do so. Perhaps they aren’t the best people to deliver this information. It probably would be much more credible coming from a neutral, third-party anyway. Maybe pharmacies, which have access to refill information already, would be a good way to get this information out. Or, even better, a company like Vitality and their Glo-Caps service (see above comments), which has no vested interest in the specific drug you’re taking.

  6. Jaimie Brown May 6, 2010 at 8:50 am #

    The main problem I have with adherence or compliance is that they are both reactive behaviours based on an action that has been decided upon by a third party. The ideal desired patient behaviours is ‘concordance’, whereby the patient has a defining role in their treatment by being involved in its choice right from the start.

    There is no easy way to achieve this, particularly in Europe, since patients are generally not payors and the right to accurate information on medication is denied most people on the grounds of cost control.

    Secondary to this is the fact that although game theory is based on the desire to win, human evoltion and our improving ability to control our environment has fuelled our belief that we are largely protected from bad things happening. Being initiated on a statin at the age of 45 because your physician tells you that you might otherwise fall ill in 20 years’ time is not especially compelling. I admire your idea of presenting data in the form of ‘years of life left’ although find it hard to think of any pharma company that would be allowed to promote such measures or regulatory body happy to approve it either. The bragging rights of being on the right hand side of the NDC would be phenomenal, though, no matter how perverse.

    • Jonathan Richman May 6, 2010 at 9:03 am #

      I’m with you on the terminology, Jaimie, and tried to give a caveat that “adherence” certainly isn’t the ideal term, but we have to pick something. Concordance works in most situations, especially upfront when a treatment is being selected. However, after this point, I’m not sure it does. When you get home and decide whether or not to take your medication, you have to decide whether or not to “adhere” or “comply” with what you and your doctor reached “concordance” on. It’s a bit tricky and I don’t think there’s a perfect word.

      I agree that humans are bad at assessing risk especially when it applies to their own health. There’s no other way to account for the poor health practices most of us follow. For many people, whether or not they adhere (or concord) with treatment will be decided before the first pill. It’ll be based on whether or not they believe the benefits of the treatment outweigh the risks (and hassles) and if they can understand how to properly evaluate this information. If this isn’t done correctly, all bets are off.

      Regarding a pharma company presenting the years of life gained, I agree that it would be difficult, if not impossible, for them to do so. Perhaps they aren’t the best people to deliver this information. It probably would be much more credible coming from a neutral, third-party anyway. Maybe pharmacies, which have access to refill information already, would be a good way to get this information out. Or, even better, a company like Vitality and their Glo-Caps service (see above comments), which has no vested interest in the specific drug you’re taking.

  7. Matt May 6, 2010 at 8:17 pm #

    Do you have any thoughts on rewarding patients for adherent behavior? Not simply with extended life expectency, but with monetary based rewards (cheaper drugs, discounts at pharmacy, etc.) These can help with behavior change to make an individual more adherent. As you said, most people decide before they take the first pill if they will be adherent. Do the benefits out-weight the risks (or hassles)? An obvious benefit is more years of life, though a little more difficult for the average person to see. How about we give them a better life expectancy by promoting medication compliance with more visible, immediate benefits? Benefits can increase (both immediate, and long term) as patients prove compliance.

    • Jonathan Richman May 6, 2010 at 8:46 pm #

      I’m actually not a big fan of incentives like this. One of the big reasons is because they generally don’t change the attitudes of people, but their behavior. Behavior changes when it comes to adherence are usually short-lived. However, if you can change someone’s attitude towards a treatment, you might have something more lasting.

      Check out the details in number 5 from my post “7 Biggest Mistakes in Digital Relationship Marketing”: http://www.doseofdigital.com/2009/12/7-biggest-mistakes-relationship-marketing/

      • Matt May 7, 2010 at 3:08 pm #

        I appreciate the honesty, and based on your previous post mentioned above, you are pretty passionate about it. Actually, I don’t disagree with you at all. The ultimate goal of any product or program should be to improve the consumers’ attitude towards it. But if there is a lack of awareness or importance, something usually needs to be done to grab the consumers’ intention and begin the conversation. Balancing incentives and proper education (as well as having a product or program worth using) are essential to long term behavior and attitude change.

        • Jonathan Richman May 7, 2010 at 4:54 pm #

          I’m not saying that incentives should be out the question. The concept of the point of minimum justification says that some level of incentive is all right and necessary, but not exceeding this minimum amount is critical. At the same time, we have to ensure that the incentives are meaningful to people. If they aren’t the incentives can have the exact opposite effect than what we want. For example, if you offer an asthmatic a book about asthma, they’ll likely not only say no thanks to your incentive program, but it potentially damages their confidence and interest about your medication as well. That can only hurt adherence.

          One other problem is taking incentives at a future date. If people are taking your medication for the incentives, as soon as you withdraw the incentive, they’ll stop taking it. That’s why the changing attitudes part is critical.

  8. Matt May 6, 2010 at 4:17 pm #

    Do you have any thoughts on rewarding patients for adherent behavior? Not simply with extended life expectency, but with monetary based rewards (cheaper drugs, discounts at pharmacy, etc.) These can help with behavior change to make an individual more adherent. As you said, most people decide before they take the first pill if they will be adherent. Do the benefits out-weight the risks (or hassles)? An obvious benefit is more years of life, though a little more difficult for the average person to see. How about we give them a better life expectancy by promoting medication compliance with more visible, immediate benefits? Benefits can increase (both immediate, and long term) as patients prove compliance.

    • Jonathan Richman May 6, 2010 at 4:46 pm #

      I’m actually not a big fan of incentives like this. One of the big reasons is because they generally don’t change the attitudes of people, but their behavior. Behavior changes when it comes to adherence are usually short-lived. However, if you can change someone’s attitude towards a treatment, you might have something more lasting.

      Check out the details in number 5 from my post “7 Biggest Mistakes in Digital Relationship Marketing”: http://www.doseofdigital.com/2009/12/7-biggest-mistakes-relationship-marketing/

      • Matt May 7, 2010 at 11:08 am #

        I appreciate the honesty, and based on your previous post mentioned above, you are pretty passionate about it. Actually, I don’t disagree with you at all. The ultimate goal of any product or program should be to improve the consumers’ attitude towards it. But if there is a lack of awareness or importance, something usually needs to be done to grab the consumers’ intention and begin the conversation. Balancing incentives and proper education (as well as having a product or program worth using) are essential to long term behavior and attitude change.

        • Jonathan Richman May 7, 2010 at 12:54 pm #

          I’m not saying that incentives should be out the question. The concept of the point of minimum justification says that some level of incentive is all right and necessary, but not exceeding this minimum amount is critical. At the same time, we have to ensure that the incentives are meaningful to people. If they aren’t the incentives can have the exact opposite effect than what we want. For example, if you offer an asthmatic a book about asthma, they’ll likely not only say no thanks to your incentive program, but it potentially damages their confidence and interest about your medication as well. That can only hurt adherence.

          One other problem is taking incentives at a future date. If people are taking your medication for the incentives, as soon as you withdraw the incentive, they’ll stop taking it. That’s why the changing attitudes part is critical.