A medical reversal is different from replacing a good treatment with a better one. It happens when a practice adopted without solid evidence is later shown to be no better, or even worse, than what it replaced. The treatment did not stop working and no hidden harm suddenly appeared: it never worked, and doctors were wrong all along.
New treatments often take hold because there is a convincing explanation of why they should work, not proof that they do. Strangely, the treatments that make the most sense in theory are often the ones that fail when tested. The story is frequently told by someone who stands to profit when the treatment is used.
A test or treatment can succeed at what it directly measures and still fail to improve what patients care about. Cancer screening tests are good at finding cancers, yet the evidence that they prevent cancer deaths, or deaths from any cause, is often weak. Doing well on a stand-in measure is not the same as helping people live longer.
In a decade of articles in a leading medical journal that tested an established or recently adopted practice, 40 percent found a clear reversal. Only 38 percent confirmed that the new practice helped. That means a large share of standard care may be useless or harmful, which helps explain why spending keeps rising without making people healthier.
Once a practice spreads, nobody waits for one more piece of evidence, so it is very hard to remove. The fix is to raise the burden of proof, so that makers of drugs and devices must show in randomized controlled trials that an innovation clearly works before it is widely used. Approval standards, treatment guidelines and advertising often set that bar too low today.