Absent the preferred evidence, we have to work with what we've got. Dentists keep detailed records on their clients, including notes related to their medical situation so that they know e.g. "Alice has [condition] and should get this kind of care but not that.". Look at the state of evidence regarding flossing and gum health: I remember reading that the statement "flossing is good for your gums" is supported by exactly one study that's over 50 years old and followed a dozen people for two weeks, then had them self-report their flossing habits over that time. That's terrible, but my hygienist can accurately divine if I've been flossing or not just by looking at my gums for about 2 seconds, and I have less pain and taste less blood during cleanings if I have been. Some of dentistry certainly might have developed how it did because somebody thought it seemed like a good idea at the time and nobody's had the good sense to check the facts. But, absent studies, I don't have another way that's quite as robust of discerning that stuff from the dentistry that represents over a century of increasingly trained medical specialists noticing that their patients get better when they do X. And neither category represents a guarantee that a practice is good or bad for my health and wellbeing, nor does it identify short-term-gain-long-term-loss situations. So I have to treat everything my dentist says as at least slightly sus; but for a slightly sus claim about (e.g.) how taking care of my teeth and gums (which are connected to the blood supply) for a couple minutes every day can reduce my risk of heart problems later in life, the risk is large enough and the cost is small enough that I'm willing to brush, floss, and use that gum stimulator thinggie to get at the back surfaces where my wisdom teeth used to be every night just in case the claim is correct. "Unsupported by science" doesn't necessarily mean a practice is bad or a claim is wrong. It can also mean simply that nobody has