Swedish researchers report that antioxidants make cancers worse in mice. It’s already known that the antioxidant beta-carotene exacerbates lung cancers in humans. Not exactly what you’d expect given the extravagant – and incessant – claims you hear made about the miraculous effects of antioxidants.

In fact, they are either useless or harmful, conclude the editors of the prestigious Annals of Internal Medicine: “Beta-carotene, vitamin E and possibly high doses of vitamin A supplements are harmful.” Moreover, “other antioxidants, folic acid and B vitamins, and multivitamin and mineral supplements are ineffective for preventing mortality or morbidity due to major chronic diseases.” So useless are the supplements, write the editors, that we should stop wasting time even studying them: “Further large prevention trials are no longer justified.”

Such revisionism is a constant in medicine. When I was a child, tonsillectomies were routine. We now know that, except for certain indications, this is grossly unnecessary surgery. Not quite as harmful as that once-venerable staple, bloodletting (which probably killed George Washington), but equally mindless.

After “first, do no harm,” medicine’s second great motto should be “above all, humility.” Even the tried-and-true may not be true. Take the average adult temperature. Everyone knows it’s 98.6 F. Except that when some enterprising researchers actually did the measurements – rather than rely on the original 19th-century German study – they found that it’s actually 98.2.

But if that’s how dicey biological “facts” can be, imagine how much more problematic are the handed-down verities about the workings of our staggeringly complex health care system. Take three recent cases:

Emergency room usage.

It’s long been assumed that insuring the uninsured would save huge amounts of money because they wouldn’t have to keep using the emergency room, which is very expensive. Indeed, that was one of the prime financial rationales underlying both Romneycare and Obamacare.

Well, in a randomized study, Oregon recently found that when the uninsured were put on Medicaid, they increased their ER usage by 40 percent.

Perhaps they still preferred the immediacy of the ER to waiting for an office appointment with a physician. Whatever the reason, this finding contradicted a widely shared assumption about health care behavior.

Medicaid’s effect on health.

Oregon allocated by lottery scarce Medicaid slots for the uninsured. Comparing those who got Medicaid to those who didn’t yielded the following stunning result, published in the New England Journal of Medicine: “Medicaid coverage generated no significant improvements in measured physical health outcomes in the first two years.”

To be sure, the Medicaid group was more psychologically and financially secure. Which is not unimportant (though for a $425 billion program, you might expect more bang for the buck). Nevertheless, once again, quite reasonable expectations are overturned by evidence.

Electronic records will save zillions.

That’s why the federal government is forcing doctors to convert to electronic health records (EHR), threatening penalties for those who don’t by the end of 2014. All in the name of digital efficiency, of course. Yet one of the earliest effects of the EHR mandate is to create a whole new category of previously unnecessary health workers. Scribes, as they are called, now trail the doctor, room to room, entering data.

Why? Because the EHR are so absurdly complex, detailed, tiresome and wasteful that if the doctor is to fill them out, he can barely talk to and examine the patient, let alone make eye contact – which is why you go to the doctor in the first place.

Doctors rave about the scribes, reports The New York Times, because otherwise they have to stay up nights endlessly checking off boxes. Like clerks. Except that these are physicians whose skills are being ridiculously wasted.

This is not to say that medical practice should stand still. It is to say that we should be a bit more circumspect about having central planners and their assumptions revolutionize by fiat the delicate ecosystem of American health care.

In the case of EHR, for example, doctors were voluntarily but gradually going digital anyway, learning through trial and error what best saves time and money. Instead, Washington threw $19 billion (2009 “stimulus” money) and a rigid mandate at the problem – and created a sprawling mess.

This is not to indict, but simply to advocate for caution grounded in humility. It’s not surprising that myths about the workings of the fabulously complex U.S. health care system continue to tantalize – and confound – policymakers. After all, Americans so believe in their vitamins/supplements that they swallow $28 billion worth every year.

© 2014, The Washington Post Writers Group