Incentives Matter

The $528 Billion Gamble: Can Lotteries and Nudges Solve Medication Adherence?

March 13, 202625:41Incentives Matter

This episode explores the pervasive and costly issue of medication non-adherence, highlighting its staggering financial and human impact. It discusses the BETTER-BP trial, which demonstrated that while incentives can significantly boost adherence, this behavioral improvement doesn't always translate into better health outcomes, revealing a central paradox. Listeners will gain insight into the deep psychological biases, such as present bias and optimism bias, alongside practical barriers that drive this widespread problem.

Key Takeaways

Detailed Report

Medication non-adherence is a colossal and costly problem, estimated to cost the U.S. healthcare system $528 billion annually and contribute to 125,000 deaths and 10% of hospitalizations each year. It's not merely forgetfulness; behavioral science reveals that up to 80% of non-adherence can be intentional, driven by psychological factors like present bias—where immediate inconveniences outweigh distant health benefits—and optimism bias, the belief that serious health risks won't apply to oneself. Practical barriers such as medication cost, complex regimens, and side effects further compound the issue.

The Behavioral Science Toolkit

To combat these deep-seated biases, behavioral scientists employ various interventions:

  • Financial Incentives: Lotteries are often preferred over fixed payments due to the psychological appeal of intermittent reinforcement, the potential for larger wins, and regret aversion (showing what participants *would have* won).
  • Technology: "Smart" pill bottles with cellular chips objectively track bottle openings, providing real-time adherence data and enabling automated reminders.
  • Commitment Devices: Strategies where individuals voluntarily commit to a future action, often leveraging loss aversion (e.g., pledging money to an undesirable cause if a health goal is missed).
  • Other Approaches: Social support networks, simplifying medication regimens, and fixed-dose combination pills also play a role.

The BETTER-BP Trial: Design and Participants

The BETTER-BP trial, led by Dr. John A. Dodson at NYU Langone Health, was a pragmatic randomized controlled trial designed to test the effectiveness of financial incentives on medication adherence for hypertension. It recruited 400 adults with diagnosed hypertension, poorly controlled blood pressure, and self-reported poor adherence.

The study population was particularly vulnerable, consisting primarily of women, Hispanic individuals, and those insured by Medicaid or uninsured, reflecting a demographic often facing significant healthcare barriers. Both the control and intervention groups received a cellular-enabled smart pill bottle to monitor medication use. The control group received no further incentives or feedback, while the intervention group received daily lottery entries for opening their pill bottle the previous day, with cash rewards ranging from $5 to $50. They also received daily SMS text messages informing them of wins and confirming adherence. This intervention lasted for six months, with follow-up at 12 months to assess lasting effects.

Surprising Results: Adherence vs. Outcomes

Behavioral Success

The financial incentives were remarkably effective at changing behavior. At the six-month mark, 71% of participants in the lottery group achieved "adequate adherence" (opening their pill bottle on at least 80% of days), compared to only 34% in the control group. This demonstrated a powerful capacity for behavioral economic interventions to overcome present bias and significantly boost adherence.

The Health Outcome Paradox

Despite this dramatic improvement in adherence, the central puzzle emerged: there was no statistically significant difference in blood pressure reduction between the two groups. The lottery group saw an average systolic blood pressure reduction of 6.7 mmHg, while the control group experienced a 5.8 mmHg reduction. The difference was not statistically significant (P=0.62), challenging the straightforward assumption that increased adherence directly translates to better health outcomes.

Fleeting Effects

Furthermore, the study revealed that the behavioral changes were not sustained. After the incentives were removed for six months, adherence rates in the former intervention group plummeted, becoming similar to the control group (31% vs. 26% meeting the adherence threshold). This indicates that the extrinsic motivation from the lottery did not translate into internalized motivation for long-term health.

Unpacking the Disconnect: Why the Paradox?

Several factors may explain the disconnect between adherence and health outcomes:

  • Measurement Error: Smart pill bottles track bottle openings, not actual pill ingestion. Participants might have opened bottles to qualify for the lottery without taking the medication.
  • Confounding Variables: Blood pressure is influenced by numerous factors beyond a single medication, such as diet, exercise, and adherence to other prescribed drugs, none of which were comprehensively tracked in the trial.
  • Hawthorne Effect: The act of being monitored, even in the control group, might have subtly influenced participants' behaviors (e.g., healthier lifestyle choices), thereby narrowing the difference in blood pressure outcomes between groups.
  • Statistical Power: It's possible a small, real effect on blood pressure existed but the study was not powered sufficiently to detect it as statistically significant.

Real-World Implications and Ethical Considerations

The BETTER-BP trial highlights the complexities of implementing behavioral interventions in real-world healthcare settings. There's a "black box" between data from smart devices and the nuanced realities of a patient's life. The "single-focus fallacy" often seen in trials—targeting one behavior in isolation—may overlook the broader context of patients juggling multiple chronic conditions and socioeconomic challenges like chronic stress, housing, and food insecurity, which can profoundly impact health outcomes.

This study also sparks a critical ethical debate around financial incentives in healthcare:

  • The Pragmatic Argument: Proponents argue that if incentives are effective and cost-efficient, they are a valuable public health tool, especially given the immense cost of non-adherence. Patients often view them positively.
  • The Ethical Counter-Argument: Concerns arise about coercion and undue inducement, particularly for vulnerable populations where financial offers might be difficult to refuse, potentially undermining voluntary choice. There's also the risk of undermining intrinsic motivation, turning health into a transaction rather than a personal commitment, and questions of fairness in rewarding behaviors others do willingly.

Ultimately, the design of incentive programs matters. Positive rewards are generally preferred over penalties, and respect for patient autonomy, along with proportionality in incentive size, are crucial for ethical implementation. The BETTER-BP trial, while ethically defensible in its design, still forces a difficult conversation about whether such programs create lasting value or merely temporary behavioral compliance.

Show Notes

Works Referenced

Glossary

Sources / References

Full Transcript

HostSo, imagine this: you've got a critical health issue, say high blood pressure, and your doctor prescribes medication. You *know* it's important, but taking that pill every day is just... hard. Now, what if someone offered you a chance to win cash prizes *daily* just for taking your meds? Sounds like a no-brainer, right?
ExpertYou'd think so! And a recent major study, the BETTER-BP trial, actually put this to the test. They gave people with hypertension a smart pill bottle and entered them into a daily lottery if they opened it.
HostAnd it worked! Like, *really* worked. They doubled medication adherence in the incentive group compared to the control group. More than twice as likely to take their pills consistently.
ExpertAbsolutely. A massive behavioral win by any measure. But here's the kicker, the part that makes you scratch your head: all that incredible improvement in behavior? It didn't translate into significantly better blood pressure control. Not at all.
HostWait, so people took their pills, but their blood pressure didn't get better? That's… that's completely counter-intuitive. That's the whole point, isn't it?
ExpertExactly. And that central paradox is what we’re diving into today. It's a $528 billion question, actually.
Host$528 billion? Wow. That's a staggering figure. What exactly does that number represent?
ExpertThat's the estimated annual cost of medication non-adherence to the U.S. healthcare system. We're talking about half a trillion dollars, purely due to people not taking their prescribed medications. It's not a niche problem; it's a colossal one.
HostHalf a trillion. That's more than the GDP of many countries. And it's not just money, right? There's a human cost, too.
ExpertAbsolutely. Beyond the financial drain, non-adherence is linked to an estimated 125,000 deaths every year in the US. It accounts for at least 10% of all hospitalizations. For conditions like hypertension, which the BETTER-BP trial focused on, poor adherence directly escalates the risk of strokes, heart attacks, and premature death. It’s a silent epidemic.
HostSo, why is this such a persistent problem? I mean, if the stakes are this high, why isn't everyone just… doing it? Is it really just people forgetting?
ExpertThat's the common assumption, but behavioral science tells a much richer story. While forgetfulness plays a role, it’s far from the whole picture. Up to 80% of non-adherence might actually be intentional, driven by a host of psychological factors.
HostIntentional? That's a strong word. What kind of psychological factors are we talking about?
ExpertThe biggest one, in my opinion, is present bias. It’s that tendency we all have to overvalue immediate rewards and undervalue future benefits. Think about taking a daily pill for hypertension. What’s the immediate experience? Maybe it’s a hassle to remember, perhaps there are mild side effects, or it’s a financial cost. Those are all very concrete, present-moment "costs."
HostAnd the benefit, I suppose, is something abstract and far off. Like, "avoiding a heart attack in 15 years."
ExpertPrecisely. That massive, life-saving benefit is distant, abstract, and probabilistic. It might happen, it might not. So, the small, immediate "cost" of taking the pill, or even just the effort of it, looms much larger in our decision-making than that huge, but far-off, potential "reward" of good health. It's why we'd rather eat that donut now than think about our cholesterol later.
HostThat makes so much sense. It's the same reason we procrastinate on taxes or skip the gym. The immediate discomfort outweighs the future gain.
ExpertExactly. And it’s compounded by optimism bias – the "it won't happen to me" mentality. If you have high blood pressure but feel perfectly fine today, it's incredibly hard to internalize the serious, long-term risks. You underestimate your personal vulnerability. "I'm sure *my* heart attack is decades away, if ever."
HostSo, a combination of "I want to feel good now" and "that bad stuff probably won't happen to me anyway." Are there other practical barriers too, beyond these cognitive biases?
ExpertOh, absolutely. Cost is a massive one. Millions of Americans simply can't afford their medications. Studies show a direct link between cost-related non-adherence and a higher risk of mortality. Then there's complexity: the more pills, the more times a day, the harder it is to stick to. And of course, side effects. If a pill makes you feel unwell *right now*, that's a very powerful immediate disincentive, reinforcing present bias.
HostSo, we have this colossal problem, driven by deep-seated human psychology. And this is where the BETTER-BP trial steps in, right? As a way to try and overcome some of these biases.
ExpertThat's right. The impetus for this whole discussion is Dr. John A. Dodson's work at NYU Langone Health. He presented these findings at a major American Heart Association conference, and they were published in a top journal. It really offers a fresh, data-driven look at whether behavioral economics can actually move the needle on this problem.
HostSo, before we get into the nuts and bolts of the trial and its surprising results, let's talk about the behavioral toolkit. What kinds of interventions do behavioral scientists typically use to try and solve this kind of non-adherence problem?
ExpertWell, if we're trying to counteract present bias, the goal is often to bring the future reward into the present, or at least make it more salient and immediate. Financial incentives, especially lotteries, are a classic example.
HostWhy a lottery instead of just, say, giving someone a dollar every day they take their pill?
ExpertGreat question. Lotteries are often preferred for several reasons. First, there's the psychological appeal of intermittent reinforcement – the unpredictability of a potential big win can be more engaging than a small, certain daily payment. Think about how addictive gambling can be. Second, there's regret aversion. Many lottery designs will tell participants what they *would have* won if they *had* been adherent. That feeling of missing out on a prize because you didn't take your pill is a powerful motivator.
HostOof, "you would have won $50 today if you'd just taken your medication." That stings.
ExpertExactly. It makes the cost of non-adherence immediate and tangible. And practically, from a program design perspective, a lottery can offer the *potential* for a high payout at a lower overall program cost than paying every participant every single day.
HostBeyond direct financial nudges, what else is in the toolkit? I've heard of smart pill bottles.
ExpertYes, technology is a huge part of this. "Smart" pill bottles or caps, like the ones used in the BETTER-BP trial, contain a cellular chip that records and transmits data every time the bottle is opened. This provides objective, real-time tracking of adherence, moving beyond unreliable self-reports. And of course, these can be paired with automated reminders via text or app.
HostAnd from a more academic, pure behavioral science angle, are there other concepts? Like commitment devices?
ExpertAbsolutely. Commitment devices are fascinating. These are strategies where individuals voluntarily "lock themselves" into a future course of action. The classic, slightly extreme example, is someone pledging to donate money to a political cause they despise if they fail to meet their health goal. It leverages loss aversion – the idea that we feel the pain of a loss more acutely than the pleasure of an equivalent gain. The thought of your money going to a cause you hate is a powerful incentive not to fail.
HostThat's brilliant, if a little terrifying. "I will take my blood pressure meds, or donate $100 to the Flat Earth Society."
ExpertPrecisely. You're creating an immediate, painful cost for not doing the desired behavior. Other approaches include social incentives and support – leveraging social networks, like having a patient report their adherence to a family member or a "buddy" for accountability. Or simply simplifying the regimen, making the desired behavior as easy as possible, like using fixed-dose combination pills or pill organizers.
HostSo, the BETTER-BP trial deployed some of these tools. Tell me about the actual interventions they used. Who were the participants, and what did each group get?
ExpertThe trial was a pragmatic randomized controlled trial, meaning it was designed to reflect real-world conditions, and it focused on a particularly vulnerable population. They recruited 400 adults with diagnosed hypertension, who were already prescribed blood pressure medication, had poorly controlled blood pressure, and self-reported poor adherence.
HostA population where interventions are really needed, then.
ExpertExactly. The majority were women, Hispanic, and insured by Medicaid or uninsured – a population that often faces significant barriers to healthcare. Both the control group and the intervention group received a cellular-enabled "smart" pill bottle.
HostSo everyone had their medication use monitored, which is already a step beyond standard care.
ExpertYes, that's important. The control group simply received that smart bottle and no further incentives or feedback. They were just passively monitored. The intervention group, however, also received the smart bottle, and for six months, they were entered into a daily lottery.
HostOkay, so this is where the cash incentives come in. How did the lottery work?
ExpertIf the participant opened their pill bottle the previous day, they were eligible for a daily drawing with cash rewards ranging from $5 to $50. And critically, they received daily SMS text messages. These messages not only informed them if they had won but also served as a real-time feedback loop, confirming their adherence or non-adherence from the day before.
HostSo it wasn't just a chance to win money; it was also a daily reminder and a confirmation of their actions. And this went on for six months. What happened after that?
ExpertThe study was designed to measure immediate effects at six months, while the incentives were active, and then any lasting, durable effects at 12 months, after the incentives had been stopped for the prior six months.
HostOkay, so we've set the stage. We have a massive problem, deep behavioral roots, and a smart intervention using lotteries and tech. Now, for the results. You said there was a huge behavioral win, but not a health outcome win. Let's start with the good news, the behavioral success.
ExpertThe financial incentives were incredibly effective at getting people to open those pill bottles. At the six-month mark, a remarkable 71% of participants in the lottery group achieved "adequate adherence," which was defined as opening their pill bottle on at least 80% of days.
HostAnd in the control group?
ExpertOnly 34% reached that same threshold. So, the incentive group was more than twice as likely to consistently take their medication compared to the control group. This is a powerful demonstration that a well-designed behavioral economic intervention *can* overcome present bias. Dr. Dodson himself stated, "Financial incentives clearly worked to change behavior during the study period."
HostDoubling adherence! That's a phenomenal behavioral change. You'd expect that to lead to, you know, lower blood pressure. So, tell me about this counter-intuitive finding. What did the blood pressure data show?
ExpertThis is the central puzzle. Despite that massive difference in adherence, there was no statistically significant difference in blood pressure reduction between the two groups. The lottery group saw an average systolic blood pressure reduction of 6.7 mmHg, which is good. But the control group also saw a reduction of 5.8 mmHg.
HostAnd the difference between 6.7 and 5.8 was not statistically significant?
ExpertCorrect. The P-value was 0.62, which means it was not a statistically significant difference. It's essentially saying we can't confidently attribute that slight difference to the intervention. This finding directly challenges the simple assumption that *adherence leads directly to better health outcomes*.
HostThat's truly baffling. It's like baking a cake, meticulously following the recipe, and the cake just… doesn't rise. So, what happened after the incentives were removed? Did people keep up the good habit?
ExpertUnfortunately, no. The study also looked at what happened in the six months *after* the lottery ended. The adherence rate in the intervention group plummeted dramatically. In that post-incentive period, adherence was similar between the two groups, with only 31% in the former incentive group meeting the threshold, compared to 26% in the control group.
HostSo, the behavior wasn't internalized at all. Once the money stopped, the motivation stopped.
ExpertPrecisely. The extrinsic motivation from the lottery did not translate into intrinsic motivation for their own health. Once the external reward was removed, the behavior reverted almost to baseline. This is a common concern with financial incentive programs – can they create lasting change, or just temporary compliance?
HostOkay, so now we have the enigma: doubled adherence, no significant change in blood pressure, and the effect vanishes when the incentives stop. Let's unpack the "why." From an academic perspective, what are the potential explanations for this disconnect?
ExpertOne of the primary limitations, which the researchers readily acknowledged, is measurement error. The smart pill bottles only tracked when they were *opened*, not if a pill was actually *ingested*. It's entirely conceivable, and frankly, a common problem in adherence research, that participants opened the bottles to be eligible for the lottery but didn't always take the medication.
HostAh, the classic "gaming the system" scenario. They fulfilled the measurable behavior but not the actual health behavior.
ExpertExactly. It's a leap from proxy measures to true ingestion. Another major factor is confounding variables. Blood pressure is incredibly complex. It's influenced by so much more than just a single medication. The trial didn't track other lifestyle behaviors like diet and exercise.
HostSo, if someone was taking their pills but also eating a lot of sodium, or not exercising, that could easily negate the effect of the medication.
ExpertPrecisely. Changes, or lack thereof, in diet and physical activity could have significantly impacted blood pressure, potentially masking any positive effect from improved medication adherence. And the study only tracked adherence to *one* primary antihypertensive medication per participant, even though many patients with hypertension are on multiple drugs for various conditions. Adherence to those other medications wasn't monitored.
HostThat makes sense. It's not a single-variable equation in real life. What about the "Hawthorne Effect"? I always hear about that in studies.
ExpertThat's a great point. The Hawthorne Effect is when subjects in an experiment modify their behavior simply because they're being observed, not necessarily because of the specific intervention itself. Both the control group and the intervention group knew they were in a study and were being monitored with a special pill bottle.
HostSo, just being in the study, and being monitored, might have made *both* groups try harder in other ways, like eating better or exercising a bit more, which could have masked the differential effect of the lottery.
ExpertAbsolutely. That awareness alone could have prompted some behavior change in both groups, shrinking the difference in blood pressure outcomes. The control group's 34% adherence, while lower than the incentive group, is still higher than what you might see in a general population with no monitoring at all. And finally, statistical power. It's possible that a small, real effect on blood pressure *did* exist, but the study wasn't large enough to detect it as statistically significant.
HostSo, from a practitioner's standpoint, what does this tell us about the messy reality of trying to implement these interventions?
ExpertIt highlights several real-world complexities. Firstly, even with smart technology, there's still a "black box" between the data and the reality of a patient's life. We know the bottle was opened, but not necessarily *why* a pill might not have been taken, even if the bottle was opened. Was it side effects? Mistrust? The tech is better than self-report, but it’s not foolproof.
HostAnd the "single-focus fallacy" you mentioned earlier?
ExpertYes, clinical trials, by necessity, often focus on one intervention for one specific behavior. But in reality, patients with hypertension usually have multiple chronic conditions and are juggling complex medication schedules. An intervention that targets only one piece of this massive puzzle may have limited overall impact. You might optimize one variable, but others are pulling in different directions.
HostAnd the population context is crucial here too, isn't it? These were safety-net clinic patients.
ExpertAbsolutely. This study was conducted in a safety-net population with significant socioeconomic challenges. Factors like chronic stress, housing instability, food insecurity – these can all profoundly impact blood pressure and may be more powerful drivers of health outcomes than adherence to a single medication. It forces us to ask if we’re intervening at the right leverage point, or if there are much larger, systemic issues that dwarf the effect of a pill.
HostSo, this study, while incredibly well-designed, really throws a spotlight on the limitations of randomized controlled trials themselves, the "gold standard" of evidence.
ExpertIt really does. RCTs are fantastic for isolating variables, but they struggle with generalizability. The results from a highly controlled trial might not translate perfectly to a broader, more diverse population. And then there's the question of what we're actually measuring. This trial successfully measured a *process outcome* – adherence – but failed to show a significant change in the *health outcome* – blood pressure. This distinction is critical. We have to ask ourselves: are we intervening on the right thing? Just getting someone to take a pill might not be enough if other, more powerful factors are at play.
HostWhich brings us to the elephant in the room when we talk about financial incentives in healthcare: the ethics. Are we just bribing people to be healthy, and is that okay?
ExpertThis is where the debate gets really interesting and often heated. On one side, you have the pragmatic or utilitarian argument: if incentives can increase a desired health behavior and are cost-effective, they're a powerful tool for public health. Given that $528 billion annual cost of non-adherence, even a moderately expensive incentive program could be a sound investment, saving lives and money downstream. And crucially, studies often show that patients themselves are open to and view financial incentives positively.
HostSo, if it works and people are okay with it, why not do it? What's the counter-argument?
ExpertThe deontological or ethical argument raises concerns about coercion and undue inducement, especially for vulnerable populations. The worry is that an offer of money, particularly for low-income individuals, might be so significant relative to their resources that it becomes an "offer they can't refuse." It undermines their ability to make a truly voluntary choice. It's "coercion by carrot rather than stick."
HostThat's a powerful point. It's not a truly free choice if the financial pressure is overwhelming.
ExpertExactly. Another concern is undermining intrinsic motivation. Paying someone to do something can devalue the act itself. It sends a message that taking medication is something that requires compensation, rather than something done for one's own health and well-being. The BETTER-BP trial's post-incentive slump really illustrates this point. Once the money stopped, the motivation disappeared.
HostSo, it could turn health into a transaction instead of a personal commitment. And what about fairness?
ExpertFairness and justice are also key ethical considerations. Is it fair to pay some people for a behavior that others do willingly and out of self-interest? It could be seen as rewarding people for having poor health habits initially. And if these programs aren't designed very carefully, they could actually exacerbate health inequities, rather than solve them.
HostSo, it's not a simple yes or no. Is there a middle ground here? Does the design of the incentive matter for its ethical acceptability?
ExpertAbsolutely. The design matters immensely. There's a broad consensus, for instance, that positive incentives – rewards – are generally more ethically acceptable than penalties, like higher insurance premiums for non-adherence. Penalties are seen as more punitive, more coercive, and risk disproportionately harming the most vulnerable who might struggle to comply for reasons beyond their control.
HostSo, rewarding good behavior is better than punishing bad behavior.
ExpertYes. And respect for autonomy is paramount. Incentive programs should support, not override, a patient's right to make informed decisions about their own health. The goal should be to bridge the gap between their stated intention to be healthy and their daily actions, not to force them into a treatment they haven't freely agreed to. And finally, proportionality. The size of the incentive matters. It needs to be large enough to be motivating, but not so large as to be unduly coercive.
HostSo, the BETTER-BP trial, using a lottery, a positive reward, and targeting a behavior patients had already agreed to with their doctors, probably sits in a more ethically defensible space than some other programs.
ExpertI agree. But the results still force a difficult conversation about whether these programs create real, lasting value, or just a temporary, transactional change in behavior. And that brings us back to that central paradox.
HostThis has been a fascinating journey into the complex world of medication adherence. So, to wrap things up, what are the key insights we should take away from this conversation?
ExpertFirst, medication non-adherence is a massive, multi-billion dollar problem with profound human costs, driven by deep-seated behavioral biases like present bias and optimism bias. It's not just forgetfulness.
HostAnd second, behavioral interventions, like financial lotteries, can be incredibly powerful at changing specific behaviors, as demonstrated by the BETTER-BP trial's doubling of adherence.
ExpertBut third, and this is the crucial one, changing behavior doesn't always translate into the expected health outcomes. There's a complex interplay of measurement issues, confounding variables, and the inherent limitations of single-intervention trials in complex real-world health situations.
HostAnd fourth, the effects of extrinsic financial incentives can be fleeting. Once the reward is removed, the behavior often reverts to baseline, highlighting the challenge of fostering true, internalized behavior change.
ExpertFinally, the ethical debate around incentives is nuanced. While they can be powerful tools, we must carefully consider issues of coercion, intrinsic motivation, and fairness, especially when working with vulnerable populations. Design truly matters.
HostSo, if incentives can double adherence but don't significantly improve health outcomes, what does that imply about the underlying problem? Is it possible we're focusing on the wrong leverage point in healthcare?