I'm not sure if this particular MRI was necessary or not, but I do know that I see MRIs being ordered all the time now. Various radiology organizations disagree about which women should undergo breast MRIs, and different insurers have their own criteria as to whether or not they cover the test.

Most health care experts agree that MRI is one of several examples of expensive medical technology that is likely being overused. In 2007 Congressional Budget Office testimony, White House Budget Director Peter Orszag attributed "the bulk of growth in health care spending to the development and diffusion of new medical technology." Patients are subject to a never-ending series of unnecessary, expensive and highly sensitive tests, which often lead to additional procedures, which, in turn, drive up health care costs.

While the idea of health care rationing has a negative connotation, our country faces some tough decisions in the next few years, and if we truly expect to reign in health care costs, we will need to hold the line somewhere. Even more to the point, however, it's become increasingly clear that more expensive care doesn't necessarily translate into higher-quality care. Study after study has shown that even though we spend more per patient than any other country in the world, we are not necessarily healthier.

Over the years, doctors have learned that it's better to go overboard and do extensive and costly workups to rule out the possibility of a serious problem rather than missing something that leads to a bad outcome, and risk a medical malpractice lawsuit. (Look for more about the contribution of malpractice to the health care crisis in a subsequent installment.) There are also financial incentives for physicians to perform more procedures, regardless whether they improve outcomes, because many get paid by the number of visits and procedures they perform. For example, in some OB/GYN practices, a routine obstetrical patient gets six or seven ultrasounds to check up on fetus along the way. Other practices follow more established guidelines recommending two or three, with additional studies only performed when necessary.There is no evidence whatsoever to suggest more ultrasounds improve outcome. In fact, additional testing can lead to worse outcomes due to the inherent margin of error of the test. An unindicated test may inaccurately show a small baby, say, which can lead to an early induction of labor and, thus, a higher likelihood of a ceasarian section.

And we not only over-test, we also over-treat, in part because we lack research comparing the effectiveness of various treatments and drugs. The $787 billion economic stimulus bill signed by President Obama in February included $1.1 billion to fund comparative effectiveness studies, an idea backed by health care industry leaders at a White House meeting last week. As I said in a previous column, this type of research is sorely needed so physicians will have a better idea if a more costly treatment is actually more effective than a cheaper alternative. It will also control spiraling health care costs by limiting spending on unproven technology and drugs.

Before I started my own private practice, I worked as the Northern Virginia Chief of Service for Kaiser Permanente, where we tried to make decisions through evidence-based "best practice" guidelines that small groups of the doctors would research, then present to our group of 30 or so OB/GYN's for approval. We were paid a fixed salary, regardless of the cost of our care, so only if we felt an expensive technology was more effective did we include it in what we considered best practice.

The interesting thing was that providing care based on evidence -- not on what the patient demanded, or what in private practice would bring in more income -- worked. Our outcome data matched or exceeded that of private practices in the same community. Private physicians grumbled at one point because our physicians had a very low c-section rate, and they suggested that we may have been taking unnecessary risks to cut costs. The grumbling stopped when the chairman of our department presented independent data showing that not only was Kaiser's c-section rate very low, but its admission rate to the Neonatal Intensive Care Unit was also lower than that of the private practice doctors.

Is the patient part of the problem? Absolutely. We've grown accustomed to having access to every sophisticated test or procedure regardless of cost or effectiveness. In the past, if you were insured, you felt impervious to cost. That has certainly changed with the advent of high-deductible policies and HSA accounts.

My family is no exception. Come September, my wife will probably follow the radiologist's advice and go have another $6,000 MRI. Since we've now met our deductible, it will be "free." But if we had to pay out of pocket, she might have investigated whether it's truly medically necessary, and whether lower-cost options exist. Actually, she's learned that making changes to her diet like cutting caffeine consumption may reduce the size of the breast mass, which could avoid the need to go through the same series of tests next year, when our premiums will likely go up another 12 percent.

This is the fourth article in a 10-part series, 10 Reasons to Reform Health Care Now:

Part 1: Syrocketing costs are choking American businesses

Part 2: 87 million uninsured, and growing

Part 3: The economy is making the crisis worse