Why do best practices get reversed so often in medicine?


In yesterday’s column, I answered a reader’s question about why doctors seem to change their minds about the best treatments for medical problems. I said that we doctors keep changing our minds because we’re human. We sometimes believe things that seem reasonable and for which there is some evidence. But then we find out, as more and better research is done, that we were wrong.

I gave the example of a treatment that was widely touted in the 1960s for bleeding ulcers called gastric freezing. With further and better research, the treatment proved worthless.

In my professional lifetime, the standards for claiming that a treatment is effective have become higher. I can’t think of an example like gastric freezing in the past 30 years. Yet sometimes we are misled by the first evidence of a treatment’s benefits. The early returns don’t always accurately predict the final outcome.

In an article in the Mayo Clinic Proceedings, a research team reported on the decade from 2001-2010. They noted several reversals of what were believed to be current best practices. Here are some examples:

  • Antibiotic treatment of diabetic women with bacteria in their bladder;
  • Use of bed covers to reduce exposure to allergens in adults with asthma;
  • Antiviral treatment of vestibular neuritis, an inflammation of the nerve that plays a key role in balance and hearing;
  • Lowering body temperature during surgery for brain aneurysms;
  • Placement of stents in narrowed heart arteries, in people with stable symptoms;
  • Treating intensively with blood-sugar-lowering medicines for Type 2 diabetes;
  • Treatment with a medicine called aprotinin during cardiac surgery;
  • Treatment with anti-inflammatory drugs called COX2-inhibitors (such as Vioxx), particularly in patients with heart disease or risk factors for heart disease;
  • Placement of tubes through the eardrum (tympanostomy tubes) in children with recurrent ear infections;
  • Arthroscopic surgery for osteoarthritis of the knee;
  • Surgery called vertebroplasty for spine bone (vertebral) fractures.

In all of these cases, a reasonable theory had led to relatively small or short research studies that seemed to support the theory. However, larger and longer studies concluded that the treatment was of no value.

In other words, many doctors decide on treatments knowing that the evidence is not all in. We use our judgment to decide what is best, in the face of what we know is incomplete evidence. For example, I never prescribed Vioxx; I was worried about its side effects. And I did prescribe antibiotics to treat many ulcers for years before that practice was shown to be correct. I got those right. But I prescribed several vitamin and mineral supplements (and took them myself) for years before they were shown to have no value. I got those wrong.

Don’t misunderstand: The great majority of treatments we doctors prescribe are of proven value. But we don’t always get it right at first, and so we change our practice.