What do we get for our money?

Is there a useful way to compare what it costs to be served by different hospitals?  Well, the Medicare agency (CMS) is trying to do that, as a first step in rewarding more efficient hospitals and penalizing less efficient ones. 

A recent story on Kaiser Health News summarized the data.  CMS presented the numbers for “average hospital spending per patient,” how much the federal program spends for the average patient admitted at a specific hospital. This measure includes all payments to doctors, hospitals or other facilities for services provided to a patient during the three days before the hospital stay, during the stay, and during the 30 days after discharge from the hospital.  To create more accurate comparisons, Medicare adjusted its figures to take into account the health and diagnosis of patients and other factors.  The national median was $17,988 for the period from May 15, 2010, through Feb 14, 2011.

There are some huge outliers in other parts of the country.  Some places in Texas, California, and Pennsylvania had figures more than 50% above this national median.  In general, Massachusetts hospitals did much better than those, but were almost uniformly above the median.  For example, Worcester’s St. Vincent Hospital and UMass Memorial Medical Center stood respectively at $18,168 and $19,067, or 1% and 6% above the median.  The major academic medical centers in Boston had similar figures, just slightly above the median.  Interesting, there was not a difference between the large centers and smaller community hospitals like Harrington Memorial in Southbridge and Healthalliance in Leominster (3% and 2%, respectively).  Even hospitals that are acknowledged leaders in clinical resource management and efficiency, like Cooley Dickinson in Northampton, do not vary from the state pattern. 

The question before us is:  Do these numbers tell us anything?

I think the answer is that they give hints as to trends and potential problems, but not much more.  Certainly, if my hospital were graded as being 25% or 50% above the median, I would want to know why and what I could do about it.  But for the great bulk of hospitals hanging around the mid-point, there just isn’t enough information to know whether I have a problem or should be relieved.

KHN quotes Elliott Fisher, the respected researcher from Dartmouth, who questioned how anyone could use this information:

 "As a hospital administrator I would go, how does this help me?" he said. "We just don’t know whether a lot of specialists are running through the hospital doing everything they can to every patient who is horizontal, or whether they're discharging every patient to a rehab facility. Those are two very different causes of high costs."

If we want see if there is a relationship between cost and quality, the problem is confounded by the fact that when CMS compiles information about the quality of care delivered in hospitals, it does so in a manner that is likewise less than useful.  By the time CMS publishes quality data, it is two or three years old.  How it could possibly correlate cost data that is 1-2 years old with quality data that is 2-3 years old presents an interesting computational problem.

By the way, CMS obviously has both cost and quality data that is more current.  After all, the agency pays hospitals virtually every day of the year based on claims that are filed almost in real time.  Those claims, in turn, are based on actual patient records that contain important measures of quality.  And yet there is something about the administrative process behind these billing and clinical records that makes it impossible to release them on a current basis.

Recently, my friend e-Patient Dave decided to test the health care marketplace.  He had a basil cell carcinoma, a small skin cancer, and he wanted to shop around and get the best price for removing it.  He has an insurance policy with a large deductible, so this is money out of his own pocket.  So Dave issued a request for proposals and sent it to a bunch of hospitals and doctors in his area.

Last week, Dave reported on his findings and his approach.  Among other things, he discovered that there is a low-cost solution to his problem that appears to have similar clinical outcomes to the higher cost procedure.  He also found a doctor who was interested in engaging with him as a partner in the care decisions.  In the end, Dave expects to spend 1/3 to 1/6 of the alternative as a result of shopping around.

But most of us don’t have the time, inclination, or patience to go through Dave’s buying process.  And we don’t feel we need to because most of us also don’t directly incur the types of costs outlined in the CMS report.  But that doesn’t mean society is not paying the freight for our lack of choice.  I hope someday that we will have real-time cost and quality data so we can become more effective health care consumers.