Some technologies are advertised directly to consumers before they've been evaluated by the kind of clinical trials that demonstrate whether they're appropriate for widespread use. The American Cancer Society recommends that most women have mammograms—the regular or digital kind—starting at age 40. Only women with a 20 percent or higher lifetime risk of breast cancer are advised to get an MRI also. And no public-health group recommends screening with ultrasound. But there's a perception even among women of average risk, says Connie Lehman, director of breast imaging at Seattle Cancer Care Alliance, that a high-tech test is superior to a mammogram. "If they're going to go in [to be screened], they want an ultrasound or an MRI," she says. Ads encouraging women to have more advanced tests have helped sow that perception, she says.
Avoiding lawsuits. Defensive medicine is another factor pushing the heavy use of technology. Obstetricians get sued at a particularly high rate, which is one of the reasons childbirth now involves so many aggressive procedures and tests, says Maureen Corry, executive director of Childbirth Connection, a nonprofit organization that promotes evidence-based maternity care. Just one example of an overused intervention not supported by evidence: constant electronic fetal monitoring during labor. "When you look at all the systematic reviews and all the randomized controlled trials, it has no benefit compared to intermittent monitoring," says Corry. And it has downsides: a lot of false alarms that lead to more unnecessary care, and a less comfortable labor, since it means women lie flat on their backs. Despite all that, says Corry, constant monitoring is standard practice.
"I understand some doctors may feel the need to order more tests and treatments to avoid being legally vulnerable," President Obama recently told members of the American Medical Association. He went on to say physicians must be supported in their choices not to order that test.
Reversing the trend toward excessive care, some economists argue, requires a fundamental realignment of our current fee-for-service system toward a system more like the Mayo Clinic's, which rewards the quality, not quantity, of care. Technology creep might be ameliorated by other changes, such as a higher bar for approval of devices and technology, says Dana Goldman, director of health economics at the Rand Corp. Currently, a device, for example, can reach the market if it's similar to an already-approved device, whether or not it's proven to produce better outcomes for patients. More practically, the current system could stand, but patients might pay more depending on whether treatments have been shown to significantly improve health, says Goldman. One sign there may be changes on the payment front: Medicare—whose coverage decisions are often followed by private insurers—recently said it would not pay for CT, or "virtual," colonoscopies, saying the evidence was inadequate. (Comparative-effectiveness research attempts to provide data on which to base these judgments.)
Hospitals and doctors can also question their own habits. "What we're really doing is asking people to reflect on their own practice and ask whether what they do makes sense," says Julie Bynum, a geriatrician at Dartmouth Medical School. At Massachusetts General Hospital in Boston, researchers studied whether a simple software program would help cut the number of outpatient CT and MRI scans and ultrasounds ordered. When a doctor ordered a test, she'd get a score reflecting the level of evidence supporting its use in that particular circumstance. Control still remained in the doctor's hands, says James Thrall, radiologist-in-chief at MGH, but the annual growth in CT scans declined by 11 percent.