Wednesday, February 20, 2008

Does Preventive Care Save Money? QALYs, New England Journal and what it means for population health

According to a very interesting article in the New England Journal of Medicine, candidates for U.S. President are naively assuming that increased health care quality will translate into bazillions of dollars of savings that will reduce health care costs and lead to lower health insurance premiums, longer healthier lives, increased productivity, economic expansion, diversion of resources to combat global warming, cleaner dirt, less heinousness and being excellent to each other. I share Bill and Ted’s style of optimism, so I was a more than a little disappointed in this whoa, wait-a-minute, not-so-fast reality check.

The Journal’s analysis was based on the obscure notion of QALYs. Like many readers of the disease management blog, I also struggle with the concept. So, I made obeisance at the altar of the great and powerful medical terminology and created my own cairn of explanation. If you can get past it, the disease management blog offers some additional thoughts on what this means for the field of population-based health care.

“Quality Adjusted Life Years” are abbreviated as “QALYs.” If you are in perfect health for one year, congratulations, you are the owner of one QALY. You are well on your way to achieving that Holy Grail of Long Life as defined by health services researchers: 100 QALYs.

Suppose, however, that your health is less than perfect for one year. Under the QALY system, the value of that decreased quality can be mathematically described as a fraction of one QALY. This makes sense: being disabled for a year is less valuable than being in good health for a year. Defining that fraction can be arbitrary but lower the health status, the lower the QALY.

This coin of the realm is usually calculated in terms of trade-offs: if you lived longer (more years) but at decreased health (disability), how many extra years would it take to equal the value of one year of perfect health? For example, if you became blind in exchange for living longer, how much longer life would you demand in exchange? It may be that it would require another ½ year. So, 1.5 years of being blind is just as “valuable” as 1 year of perfect health. The inverse (mathematically expressed as a reciprocal) may put it into better perspective (pun intended): being blind but in otherwise good health for a year is worth 1/1.5 or .66 QALYs. Being blind shaves off 1/3 of the quality of life.

As I understand it, QALYs are determined by conducting surveys and asking people to hypothetically “trade” longevity in exchange for decreased quality. It can appear silly, but the good thing about QALYs is that this system levels the playing field and allows researchers to compare and contrast the benefit of different treatments for different illnesses. For example, which results in more QALYs for men age 70?: prostate surgery for treatment of prostate cancer or performing bypass surgery for angina?

You can also toss in the cost. Returning to the blindness example above, if there were a treatment that could avoid the blindness that cost $100, that would work out to $100/.33 QALYs or $300 per additional QALY. It can be far more complicated than that (you have to carry the math out over a person’s life expectancy and factor in the cost of inflation for example) but hopefully you get the point.

Note that the calculation of QALYs implies that greater quality comes at a price. It is possible to have an intervention that simultaneously reduces cost and improves health. For example, flu shots decrease costs and improve health, so the QALY calculation breaks down. It’s “less” than $0 per QALY

Check out this example from JAMA from back in 2000. Eye screening for persons with diabetes resulted in 3 to 21 days (a fraction of a year) of sight. When calculated on a per QALY basis, the amount of money spent was calculated to range from just of $40,000 per additional QALY to over $200,000.

If may interest you to know that the cost of hemodialysis is about $50,000 per QALY gained. That has become a benchmark. If an intervention costs more than $50,000, that could be considered “expensive.” For an excellent discussion of the conundrum around this, check out this commentary.

According to the party poopers in the New England Journal of Medicine, if one uses dollars per QALY to scrutinize the candidates’ rhetoric, “the vast majority” of medical interventions generally regarded as quality translate into more dollars of cost for every QALY gained. This includes screening for diabetes and high blood pressure

What does this mean?

Unfortunately, it is too easy to just read the headline and conclude “prevention doesn’t work.” Trying to overcome that by putting the authors use of $/QALYs into perspective doesn’t really overcome the damage. That being said, here are some points that I think should be kept in mind:

  • $/QALY are not used by patients and their doctors in the real world of clinical practice, so using $/QALY to assess their decision making doesn’t add a lot of insight. In other words, researchers should use whatever metric is being used by patients in their decision making.
  • Ditto insurers. They live by the medical loss ratio and actuarial trend.
  • QALY models are subject to lots of assumptions, especially at the lower ends of the QALY scale. If a person moves from one level of burden from a chronic illness to another level, QALY’s may not adequately portray the trade off. It’s less useful in chronic illness.
  • Societal decision-making on the value of medical interventions is based on far more inputs than QALYs For example, breast cancer screening for all women age 40-80 years can cost $58,000/QALY. We still think it’s a bargain. The calculus is far more sublime.
  • The $ numerator is often standardized without accounting for the cost efficiencies that are possible through disease management programs that promote self-care or alternative levels of care (for example, nurses).
  • The QALY denominator fails to take into account the added quality from having a greater sense of control over disease as well real value that patients place on their relationships/community with the healthcare team.

No comments: