It has been a couple of weeks since the landmark Oregon experiment paper came out, and the buzz around it has subsided.  So what now?  First, with passage of time, I think it is worth reflecting on what worked in Oregon.  Second, we should take a step back, and recognize that what Oregon really exposed is that health insurance is a small part of a much bigger story about health in general.  This bigger story is one we can’t continue to ignore.

So let’s talk quickly about what worked in Oregon.  Health insurance, when properly framed as insurance (i.e. protection against high, unpredictable costs) works because it protects people from financial catastrophe.  The notion that Americans go bankrupt because they get cancer is awful and inexcusable, and it should not happen. We are a better, more generous country than that.  We should ensure that everyone has access to insurance that protects against financial catastrophe.  Whether we want the government (i.e. Medicaid, Medicare) or private companies to administer that insurance is a debate worth having.  Insurance works for cars and homes, and the Oregon experiment makes it clear that insurance works in healthcare.  No surprise.

The far more interesting lesson from Oregon is that we should not oversell the value of health insurance to improving people’s health.  While health insurance improves access to healthcare services (modestly), its impact on health is surprisingly and disappointingly small.  There are two reasons why this is the case.  The first is that not having insurance doesn’t actually mean not having any access to healthcare.  We care for the uninsured and provide people life-saving treatments when they need it, irrespective of their ability to pay.  Sure – we then stick them with crazy bills and bankrupt them – but we generally do enough to help them stay alive.  Yes, there’s plenty of evidence that the uninsured forego needed healthcare services and the consequences of being uninsured are not just financial.  They have health consequences as well.  But, claims like 50,000 Americans die each year because of a lack of health insurance? The data from Oregon should make us a little more skeptical about claims like that.

So what really matters?  Right now, we are pouring $2.8 trillion into healthcare services while failing to deliver the basics.  To borrow a well-known phrase, our healthcare system is perfectly designed to produce the outcomes we get – and here’s what we get: mediocre care and lousy outcomes at high prices.  Great.

Let’s use cardiovascular disease as an example.  We know it kills more Americans than any other condition.  The CDC estimates that we spend about $500 billion on CV disease.  With that kind of spending, you’d think we would be really good at managing it.  When it comes to cardiovascular disease, management is relatively straightforward: there are four risk factors worth thinking about: hypertension, diabetes, high cholesterol, and smoking.  But guess what?  We’re really not that good at managing these conditions, and evidence suggests that health insurance has almost nothing to do with it.  Here’s the evidence:

1. Hypertension: Nearly 70 million adults (1 in 3) have it.  More than half of these Americans’ blood pressure is poorly controlled.  Rates of poor control are only marginally worse among the uninsured (58%) than among the insured (51%).
2. Diabetes: Nearly 26 million people have it. Rates of poor control?  You guessed it: about half, and the same between the uninsured (46%) and the insured (44%).
3. High cholesterol:  Again, about 70 million adults (1 in 3) have it.  Rates of control?  Even worse!  About 1/3 have their cholesterol under control.  The proportion with poor control is lower among the insured (60% versus 77%) than the uninsured, but even among the insured, frankly, cholesterol management is terrible.
4. Smoking: About 50 million people smoke.  None of them have it under adequate control (by definition).  Most of these people have health insurance.

Type of insurance really doesn’t matter. A landmark New England Journal of Medicine paper in 2003 found that the quality of care for privately insured Americans was about as bad as it was for those on government insurance or who were uninsured. On a global measure of how often patients get the right care, insurance really doesn’t make a big difference. See below:

*From: Asch SM, Kerr EA, Keesey J, Adams JL, Setodji CM, Malik S, et al. Who Is at Greatest Risk for Receiving Poor-Quality Health Care? New England Journal of Medicine. 2006;354(11):1147-56. PubMed PMID: 16540615.

This, of course, begs the question: how can we be spending so much money and not doing better on cardiovascular disease management?  How can this be?  The knee-jerk reaction that I hear over and over again is to blame the patient – they are not compliant with their medications.  They don’t follow up.  They don’t understand their condition.  But these are weak excuses for a healthcare system that only pays when a patient visits a doctor’s office or an ER or a hospital.  We have a supply driven healthcare system because of a failure of imagination – we only seem to know how to pay for visits and medications and tests and procedures.

If we’re going to get healthcare to improve health, we have to seriously rethink the way we pay for it.  I don’t mean adding a 1% incentive to a doctor’s reimbursement for measuring blood glucose.  That doesn’t do much and is usually just insulting.  I mean adding incentives to make providers focus on managing patients’ health.  The problem right now is that no one gets paid if they figure out how to get patients to take their medications regularly.  No one gets paid to communicate more effectively with their patients or get them to quit smoking.  We don’t financially reward providers who improve health.  In fact, we punish them: because as people get healthier, they will have fewer visits, decreasing provider revenue.

This is more than a diatribe against fee-for-service.  It’s a diatribe against paying for healthcare. We need to find a way to pay for health.  Yes, it sounds naïve, but we have to start thinking outside the box if we want transformative changes rather than iterative ones.  For instance, what if we paid for better blood pressure control?  Instead of getting paid to measure every patient’s blood pressure (as many pay-for-performance schemes do), what if we paid for lowering blood pressure among those with severe hypertension?  Yes, there are issues of case-mix adjustment, but those are solvable.  For each one of us, the things that would improve our health surely vary.  What if the payment system could take patient preference into account, paying for things that we each individually valued as important to our health and well-being?  None of this is easy.  But we surely haven’t built this insanely complex and dysfunctional payment system because it’s the easiest way to pay for healthcare.  We got here despite ourselves.

My lesson from the Oregon experiment is that our system pours hundreds of billions of dollars into stuff, but pays little attention to whether any of that stuff is improving people’s health.  Adding more people to the insurance rolls –pouring more money into a low value healthcare system – isn’t going to improve people’s health.  Will it help the uninsured financially?  Sure.  Is providing financial security to poor Americans a good thing to do?  Absolutely.  No American should be one car accident away from bankruptcy.  But until we improve the underlying functioning of the healthcare delivery system, we shouldn’t expect any intervention that improves access to more healthcare services to have a meaningful effect on people’s health.

Ashish Jha is an associate professor of health policy and management, Harvard School of Public Health.  He blogs at An Ounce of Evidence and can be found on Twitter @ashishkjha.