New Health Dialogue - logo

COST: We Made the Hospital Bed, So Lie In It

I want to expand on Joanne's excellent and thorough post yesterday on the new Dartmouth Atlas Project report. One sentence in the AP/WaPo story especially caught my eye: the supply of beds in a locality is a key driver of how many days patients spend in the hospital. In other words, the more beds you have, the more patients you will admit. According to the Dartmouth report, this doesn't affect the fairly clearcut necessary hospital stays, an elderly patient with a hip fracture for instance. But, when the decision to hospitalize a patient is "more discretionary—as is the case for patients with heart failure and most other medical conditions—admission rates are strongly correlated with the local supply of hospital beds." Hmm.


The degree of variation is remarkable: during the five-year period 2001-05, on average, patients with chronic illnesses living in the region using the least supply-sensitive care spent about 6.1 days in hospitals during their last six months of life, while those living in the region using the most supply-sensitive services spent an average of 21.9 days as inpatients during the last six months of their lives.

That means one patient could spend over 3½ times as many days in the hospital as another patient with the same condition who lives in a community with fewer beds. The AP/WaPo story explains why this is bad:

Dr. Elliott Fisher, who led the study, said that more days in the hospital did not necessarily lead to better outcomes. Those patients were usually seen by more specialists, and they spent more time in the intensive care unit, but they did not live longer, on average.

"We know that hospitals are dangerous places if you don't need to be there," Fisher said, referring to the risk of infection or, for elderly patients, falls.

And the cherry on top: these hospitalizations cost Medicare (and thus taxpayers) more money and drive up the cost of health insurance for everyone else.

Now I'm not saying that any specific number of days in the hospital is right for someone in need of inpatient care. What I am saying is that if there is little variation for hip fracture patients, it's possible that there shouldn't be that level of variation for other medical conditions. And if what Dr. Fisher said is correct, we should try to reduce hospital days for patients.

Okay, so, let's just reduce the oversupply of hospital beds in those areas, and let's maybe take the rest of the afternoon off. Are the Nats playing today? But wait, Jane Sarasohn-Kahn has a thoughtful post today about Why It's Impossible to Close a Hospital. She says:

Always remember that one worker's income is another one's cost. For some communities, the hospital is the local monopsony providing the lion's share of meaningful employment.

The chart on the right from the AHA study illustrates that in many states, hospitals provide at least 1 in 10 jobs: this is true for Maine, North Dakota, Pennsylvania, and nearly 1 in 10 for Massachusetts, Michigan, Missouri, Ohio and West Virginia, among others.

The microeconomy of the hospital is thus a major contributor to the states' and nation's macroeconomy.

When there's talking of closing hospitals, there's no doubt why it's so tough to do so. Financing hospitals, appropriately, has implications well beyond "the bed" and the individual patient.

With closing hospitals off the table, we must look at reorganizing them. Dr. Guy Clifton posted recently how we could start:

The U.S. has been in the midst of a hospital building boom, but the evidence is that it has focused on high-margin operations such as back surgery, orthopedic surgery, and heart surgery and not on the kinds of services needed by patients with emergency medical conditions, which are less profitable for hospitals.

It all comes down to payment policy. Primary, preventative, emergency, and uncompensated care are currently undervalued by our health care system. Reimbursement for other high-margin procedures must be reduced so that we can increase payment to these four high-value areas of care.