Doctors enjoy mocking each other, mostly because we tend to be alpha-types who grew up being the smartest person in the class. I myself am guilty of this. It's occasionally brutal, but usually good natured, at least where I practice.
It was a good natured dismissal I received when I presented an idea to an endocrinologist the other day. (Endocrinologists specialize in hormone-related diseases, especially diabetes.) The endocrinologist works for a drug company that manufactures a long-acting insulin. This insulin is approved for once-a-day injections. That's an advantage over the older long-acting insulins, which are injected twice a day. Nobody likes shots, and if you can reduce them, that's a good strategy for a drug company.
I asked him about a not-uncommon practice I've seen, using this insulin twice a day in patients with difficult-to-control blood sugars.
He dismissed this idea. "No study has shown that it works better with twice-a-day dosing,” he said. “And you would think, after all these years, that there would be a study on it if it worked."
I didn't agree with him, because there are many reasons why there would be no data on a specific point of medical practice.
First, there are cognitive biases in medicine. Medicine faces uncertainties, and in any given decade, there are hypotheses that are considered plausible and some that are not. The hypotheses that are thought implausible will not receive research funding, and studies arising from these hypotheses will be published less often. After all, it's human beings who make the decisions on which studies to fund, and which studies to publish. Human beings aren't perfect or omniscient.
Other reasons for “no data” on a medical topic are economic. For example, drugs that can't be patented don't get studied as well. Vitamins are a good example. No drug company will pay millions of dollars for a study if they can't recoup their investment through a patent. And I don't have data on this, but I believe most research is paid for by drug companies.
For the long-acting insulin, research would be funded by the manufacturer. Manufacturers are generally interested in strengthening their market position. Paying for a study of twice-daily dosing would be a bad business decision, because if it worked, it would show that the drug was not superior to the old medicines. It would reduce use of this insulin, and profitability. The manufacturer would be working against its own interests.
I'm not saying that giving long-acting insulins twice a day works better in certain situations, though it seems to. Really I don't know.
But when we try to decide on a controversial medical practice, we need to understand the limitations of the medical literature as a whole.