Location, Location, Location

It's true: how much Medicare spends on you if you are old and sick depends on where you live. Researchers at the Dartmouth Institute for Health Policy and Clinical Practice have been saying this for years, but the skeptics insisted that maybe the people who had more tests and treatments were sicker, and maybe people just happened to be sicker in certain parts of the country than in others. Or maybe patients in some areas insisted on more aggressive treatment than patients elsewhere. Now, a new study carried out by the independent, well-respected Institute of Medicine suggests that the Dartmouth researchers were right all along. 

Not only are there parts of the country where both prices and “resource utilization,” or the intensity of tests and treatments, are higher than others, not only are these differences very large, but they have remained unchanged for the past 20 years. What’s going on here?

The new study finds that some of the geographic variation is related to the patients’ health status—people really are sicker in certain places than in others. But most of the difference between the high cost, high intensity regions and low spending regions is driven by the utilization of “post-acute services,” or care in places such as rehab centers after hospitalization. If post-acute care were the same everywhere, the variation in Medicare spending would fall by 73%. The rest of the difference between the high spending regions and the low spending areas is due to acute care, ie hospitalization. If hospitalization rates and hospital care were the same everywhere, the variation in Medicare spending would fall by 27%.

The authors of the IOM study suggest that the fix to the variation problem is to promote the “clinical and financial integration of health care delivery systems.” In other words, if hospitals and rehab centers and primary care doctors all shared responsibility for taking care of patients, if they had to share a fixed payment from Medicare, or so the argument goes, then each sector would behave efficiently so as to avoid penalizing both themselves and everyone else. This is the arrangement in the Accountable Care Organization model, currently being tested by Medicare in 32 sites across the country. 

Preliminary data suggest that the ACO approach may work to make the member organizations (hospitals, physician practices, etc.) more efficient. It may bring down costs. But concluding that ACOs are the way to control cost while maximizing quality, along with other forms of integration such as the “patient-centered medical homes” and “bundled payments,” begs the question of why there is geographic variation in the intensity of both acute care and post-acute care. If lack of integration is the culprit, is there less integration in some parts of the country than in others? Are rehab centers and hospitals doing more because they can get away with it, or for some other reasons?

Location clearly matters. But before you decide where to live based on your preference for a lot of medical interventions or few medical interventions (with little evidence of any consequence for your health either way), and before Medicare goes ahead and "incentivizes" integration within health care systems, we need some more answers.