• “Automated hovering” in health care

    If you are interested in novel approaches to improve health behaviors within and outside the formal healthcare system, there’s no better group to watch than the University of Pennsylvania’s Center for Health Incentives and Behavioral Economics. If you are not familiar with their work, the July 5 New England Journal of Medicine perspectives piece, Automated Hovering in Health Care — Watching Over the 5000 Hours is worth a look.

    Authors David Asch, Ralph Muller, and Kevin Volpp start out with a simple but powerful observation about how medical care (and systems outside medical care) must change to better promote health:

    The dominant form of health care financing in the United States supports a reactive, visit-based model in which patients are seen when they become ill, typically during hospitalizations and at outpatient visits. That care model falls short not just because it is expensive and often fails to proactively improve health, but also because so much of health is explained by individual behaviors,1 most of which occur outside health care encounters. Indeed, even patients with chronic illness might spend only a few hours a year with a doctor or nurse, but they spend 5000 waking hours each year engaged in everything else — including deciding whether to take prescribed medications or follow other medical advice, deciding what to eat and drink and whether to smoke, and making other choices about activities that can profoundly affect their health.

    The authors also note three developments which offer promise that our health system can do better.

    Payment mechanisms that reward providers based on patient health outcomes rather than medical system interventions will hopefully encourage greater emphasis on supporting our well-being during the vast majority of our lives that aren’t spent in the hospital or the doctor’s office.

    The proliferation of new computer and communication technologies such as cell phones and wi-fi is also a huge opportunity. It’s insane that we make people trudge to the hospital or to a doctor’s office for basic biometric measurements that could easily be performed at home and electronically communicated. It’s more insane that some of the most critical tasks in patient monitoring are performed in the same reactive fashion.

    Deepening understanding of behavioral economics focuses attention on novel ways to improve health behaviors at home and at work. This group has run several interesting clinical trials, for example studies to explore what can and cannot be achieved  through the use of financial incentives to promote smoking cessation and weight loss.

    Behavioral economics may be especially helpful to address a basic challenge in chronic illness and care: Helping patients to properly take their medicine. Non-adherence due to confusion, carelessness, or distraction is a critical issue that leads to many avoidable bad health outcomes, as well as the need for costly avoidable rehospitalizations.

    The authors describe one study of patients taking warfarin that included an electronic pill dispenser. Patients participated in an automatic daily lottery, with a small chance of winning $100 and a larger chance of winning $10:

    Each day, patients were electronically notified if their number had come up — which it would do about 1 day in 5 — but were eligible for the prize only if they had taken their warfarin the previous day, as signaled by the dispenser. The system provided daily engagement, the chance of a prize, and a sense of anticipated regret: no one wants to receive news of winning only to be disqualified for nonadherence the previous day. The expected value of the lottery was less than $3 per day, but the system reduced the rate of incorrect doses from 22% to about 3% and reduced the rate of out-of-range international normalized ratios from 35% to 12%.

    I wouldn’t oversell the transformational potential of such clever behavioral interventions. Yet this kind of rigorous exploration is often valuable. In matters from personal diet to the reading of labels and the signing of consent forms, we should spend more time understanding how real people process information and how we all might be helped to make better decisions, and to improve the routine, mundane but critical behaviors that will lengthen or shorten our lives.

    • I’ve long thought this kind of stuff would be useful for diets and other lifestyle changes. If you need to lower your cholesterol/blood sugar/pressure, it would be better to have a stepped set of guidelines like start by eliminating this, and in a few weeks we retest.

      Instead, most people get advice and a pamphlet and the next review is a year later. If they took some steps that worked but gave up long before the year was over, they got no positive feedback. If they took some steps that weren’t enough, they don’t know they need to do more for another year. In both cases, they’re probably starting from scratch a year later.

      The problem is that with job-related health insurance there’s little incentive for the insurance companies to invest in people’s health. The average American changes jobs ever 5 years, and the average company changes insurance about as often, so the insurance companies have a very short horizon to see a payback on their investment.

    • I’d love to see you guys review Eric Topol’s book, “The Creative Destruction of Medicine.” It’s similar to the automated hovering article yet much more optimistic about technology’s transformative potential in the next 10 years.

    • We’re doing contortions to solve the problem that people don’t bear the cost of their own behavior. Is it wrong to say if your poor health is due to your behavior then you pay the cost? Is the ethical issue that it is impossible to discern with certainty whether an individual’s health is due to exercise of their free will?

      At stake are the opportunity costs — If we don’t treat those whose health is due to their behavior (unless they pay) what we would do with the savings. There are some very good uses.

      • how would you even start to sort out what is a person’s fault (or free will as you put it…)?
        who would do the deciding?
        what is the collateral damage of such a policy (“sorry kid, your mom smoked and we’re not going to cover treatment for her pneumonia… sorry off to the foster home for you…”)?

        Why not just do what we already do and try and hit people on the front end when they do dumb stuff (cigarette taxes)
        Or even better, mix carrots with sticks to encourage movement away from bad behavior…