Overdosed America (28 page)

You would think that there would be a mechanism in place to ensure that our medical care was based upon a solid foundation of medical research. The United States was actually the first
country to implement just such a program
, known as health technology assessment, in 1975. Since then, most of the other industrialized nations have established formal mechanisms to determine the optimal use of new medical technologies and to protect their citizens from unproven or wasteful innovations. In accomplishing this public service, however, these programs may compromise the entrepreneurial potential of medical innovation, and therein lies the rub.

As the influence of technology assessment programs has grown abroad, the United States has virtually dismantled its own. A 1999 article in JAMA explains that this occurred, “despite the fact that strong arguments have been made for such an entity for more than 20 years,
most recently by a federal commission in 1994
.” Why? The medical-device industry and several doctors’ organizations opposed government control over the research and evaluation of new technologies. Specifically, the American Society of Cataract Surgery, the American Board of Ophthalmologists, and the North American Spine Society were
not pleased with government-sponsored guidelines
(issued by the Agency for Health Care Policy and Research, AHCPR) defining appropriate criteria for cataract and low-back-pain surgery. The final battle took place in 1995 over spinal fusion surgery. Spine surgeons took umbrage at the government’s proposed limitation on treating herniated disks in the low back this way. The surgeons’ protests resonated with the antigovernment environment in Congress at the time—not only were the guidelines discredited, but the entire AHCPR was almost zero-funded for the following year.

A December 2003 article in the
New York Times
explained the controversy about back surgery
. This excellent piece of investigative medical journalism points out that although the spinal fusion procedure is more complex and more expensive, and the recovery time is longer, there has never been proof of its superiority over laminectomy. Why the push to do the more complex procedure before it has been proven to be superior to the simpler procedure? Part of the answer, according to the
Times
article, is that doctors are paid approximately $4000 per spinal fusion procedure versus $1000 per laminectomy, and hospitals are paid $16,000 versus $7000. There is another part of the answer as well: Since the AHCPR’s guidelines were overturned, the number of spinal fusion operations being done in the United States has tripled, and the amount spent on materials for the procedure has increased fivefold, from $500 million to $2.5 billion each year. The
New York Times
reported that a former sales rep for Medtronic (the largest manufacturer of hardware for spinal fusion surgery) said in a lawsuit against his former employer that he had been told to do “whatever it takes” to sell more supplies for spinal fusion surgery. In his lawsuit, he charged that “whatever it takes” included “sham” consulting contracts and first-class trips to Hawaii. Medtronic settled the lawsuit and denied any wrongdoing. However, it did disclose, in September 2003, that it was being investigated by the Department of Justice for
allegedly making illegal kickbacks
, representing improper inducements under the Federal Antikickback Statute.

Two other former Medtronic employees told the
Times
that surgeons were routinely enticed to use Medtronic hardware with offers that included expensive trips, nights on the town that cost up to $1000, and sometimes even visits to the local strip club.

Is this how you would want the decision to be made about which operation your surgeon is recommending for you or a family member? (Perhaps when good data become available we will find that there is a role for spinal fusion sugery in the treatment of back pain. It is, however, quite telling that we still don’t know—at this late date.)

The term
“supply-sensitive care”
is applied to the kinds of medical services that are most vulnerable to getting pushed into use by the providers’ financial interests rather than pulled into service by the health needs of the surrounding community. Without a formal mechanism of health technology assessment, new medical services can be brought into use without strong scientific evidence of benefit. And without limits on spending, new services can be brought into use without evidence that they provide more health value than the services they would be replacing. The absence of both of these constraints on the growth of medical technology allows the U.S. health care system to be uniquely shaped by financial incentives.

Four features are shared by the medical services that are most vulnerable to overuse because of this supply-side push.

First, supply-sensitive services must be covered by insurance.
When insurance coverage shields patients from the real cost of their care, they are unlikely to question whether the health value of a test or procedure justifies its cost. If, for example, heart attack patients had to pay the extra $10,000 that American-style invasive post–heart attack care now costs, many would demand access to the kind of information presented above. My guess is that most patients with uncomplicated heart attacks, when presented with the best available evidence, would conclude that the likelihood of benefit from invasive cardiac testing and procedures is not great enough to justify the increased risk and cost, and they would opt for more conservative care. I also suspect that if the technological razzamatazz were no longer such a distraction, the kinds of commonsense interventions that lead to better health outcomes for most people would then move into the foreground, where they belong.

Second, supply-sensitive services must appear on the surface to be beneficial, preempting the need for proof.
How many women with advanced breast cancer, facing a poor prognosis, would turn down the opportunity to have a bone marrow transplant when the prevailing attitude is “It’s your only chance”? How many heart attack patients would require de-tailed proof of the benefit when their cardiologist says, “We should do a cardiac catheterization to make sure that none of your coronary arteries is about to become completely blocked and cause more damage to your heart”? How many new mothers would say no if their baby’s doctor recommended transfer to the intensive care nursery?

Third, the need for supply-sensitive services must be determined by the doctors who perform the service.
Even though doctors almost universally believe that their decisions are scientifically based, financial ramifications have a way of exerting subtle influence over their interpretation of scientific evidence. We know, for example, that
cardiologists who perform cardiac catheterization
and angioplasty are more likely to recommend these procedures than are other cardiologists and primary care doctors—though all claim to be guided by the best evidence available. In my experience, doctors rarely recommend procedures simply to make more money, but like most people, they like to use their special skills to help others; this creates a predisposition to want to use the latest tests, drugs, and procedures (not to mention defend themselves against the ever-present risk of a malpractice suit). As the saying goes, “When you have a hammer, the whole world looks like a nail.”

Fourth, supply-sensitive services must provide attractive enough financial opportunities to motivate hospitals and other facilities to invest in the capacity required to provide the service.
The increased capacity to do cardiac procedures and the growth of neonatal intensive care units are examples of “good investments” for hospitals.

In a health care system lacking effective health technology assessment and limits on spending, paying doctors and hospitals more for doing more, and disconnecting patient costs from health care value, supply-sensitive services are sure to be overused.

The most compelling data showing that more care is not necessarily better for health come from the ongoing studies performed by Dartmouth Medical School’s Center for Evaluative Clinical Sciences, headed by Drs. John Wennberg and Elliott Fisher. Their research focuses on the impact of regional variations in the use of medical services on health outcomes and costs. Intuitively, we would expect Medicare patients who receive more care in the higher-spending regions of the country to get better care and be healthier. But studies funded by the Robert Wood Johnson Foundation and the National Institute on Aging show that this is not the case. For example, without any discernible benefit in health outcomes,
more than twice as much
is spent taking care of Medicare patients in Manhattan than is spent on similar patients in Portland, Oregon. Over the 18-year life span of the average Medicare patient after turning 65, this difference amounted to $100,000 per person in 2000, without even taking into account differences in spending on prescription drugs.

Fisher and colleagues specifically studied Medicare patients admitted to the hospital with a first diagnosis of heart attack, broken hip, or cancer of the colon. The cost of caring for the patients
was 60 percent more in the highest-spending regions
than in the lowest-spending regions. The patients in the highest-spending regions spent more days in the hospital and the ICU, and had more visits with specialists, more diagnostic tests, and more minor, but not major, surgical procedures. The extra services provided to patients in the higher-spending regions did not, however, translate into better outcomes. The patients in the lower-spending regions had
better
access to care,
higher-quality
care, and
less
chance of dying over the five years of the study. The bottom line appears to be that once an adequate amount of care is being provided, as in the lowest-spending regions of the country for Medicare patients, more care is worse care. This seems to be particularly true for the kind of care that is pushed into service by supply-side pressure.

American medicine’s predilection for high-tech tests and procedures often leads to overuse of useful new technologies. The new body scans are a good example. In just the years 1999–2001, the
number of MRI scanners in the United States
increased by about 50 percent. Where are the data showing that the tremendous increase in the number of MRIs and CT scans being done improves clinical outcomes? There aren’t any to speak of. Though we don’t know the health consequence of these additional scanners, we do know their economic consequence. We can also be quite confident that the key consideration in the decision to put the vast majority of these new scanners into service was the bottom line.

Consider the use of CT scans for patients with suspected appendicitis. CT scans of the abdomen are routinely used for patients who arrive at the hospital with the usual symptoms of appendicitis: abdominal tenderness, fever, and vomiting. The scan, doctors presume, helps to confirm or rule out the diagnosis of appendicitis. Yet a large study of
patients in Washington state
, published in JAMA in 2001, revealed that the use of CT scans, abdominal ultrasounds, and laparoscopic appendectomy did nothing of the sort: neither the rate of unnecessary appendectomies (removal of what turned out to be a normal appendix) nor the rate of perforated appendixes—the most serious complication of delay in diagnosis—decreased after these newer technologies became available.

The saga of back pain and high-tech scans tells a similar story. Back pain is one of the most common complaints seen by primary care physicians. The vast majority of healthy people who come in with acute back pain recover completely. Many recover without any therapy, or with physical therapy, or occasionally with manipulation. Yet one of the most overused technologies that I see as a family physician is the MRI for patients with back pain. Though the MRI is an elegant technology that produces beautiful anatomical pictures, these beautiful pictures don’t necessarily translate into better clinical results.

A patient of mine, Mr. Oscar, was 69 years old when he developed low back pain that radiated into his left buttock. After four weeks of treatment with anti-inflammatory medication and back exercises, his orthopedic specialist ordered an MRI, which revealed a bulging disk between two of Mr. Oscar’s lower vertebrae. Mr. Oscar wanted to know what to do next. The answer wasn’t simple, because about
80 percent of people older than 50
without any back symptoms have at least one bulging disk in their lumbar spine on their MRIs, and two-thirds have more than one disk abnormality. So what was the chance that the abnormality seen on Mr. Oscar’s MRI really represented the cause of his discomfort? There was no way to know. The only real benefit of the MRI was to rule out serious unsuspected problems like tumors or fractures, though these were extremely unlikely, given Mr. Oscar’s history and the results of a physical exam. In short, the MRI wasn’t much help. His pain improved slowly over the next few weeks with physical therapy and exercises—which could have been prescribed without the MRI.

Mr. Paul’s back pain story is my favorite: An avid mountain biker in his mid-forties, Mr. Paul came to me because of moderate to severe back pain, radiating down one leg to his midcalf. His history and exam did not lead me to expect that his recovery would be unusually slow. I suggested that he rest his back (no biking), take an anti-inflammatory drug, and use ice on his back and heat for the muscle spasm in his leg. After three weeks, Mr. Paul’s pain had not improved, so I recommended physical therapy and manipulation. Mr. Paul waited patiently, but after he had experienced two or three months of pain and an inability to go biking, I referred him to a thoughtful and conservative neurosurgeon at Lahey Clinic. An MRI revealed a herniated disk that corresponded to the location of his symptoms. Because of the persistent pain, Mr. Paul opted to have surgery to relieve the pressure on his sciatic nerve root—the presumed cause of his pain. Given the duration and degree of his pain, I supported his decision. (I also would have supported a decision to wait.)