Everyone smart knows homeopaths and witch doctors are quacks. And everyone knows “doctors are wrong sometimes”. But it seems like most “medical information” online, even from reputable sources, is effectively useless. Why so?

- No one uses numbers. Everyone knows “sugar is bad for you”. But how much sugar? Over what time period? How bad is “bad”? It’s like being told a house is “down highway 99″. How far down? Two? Seventeen? Three hundred and sixty-four? Good luck finding anything that way.

- No one uses probabilities. “Obesity may cause heart disease.” But how much of an effect does it have? What’s the chance of having a heart attack in the first place? 2%? 20%? 90%? Logically, you should be ten times as worried about something ten times as likely.

- Everyone over-emphasizes the extremely rare. Psychology shows humans have no intuitive concept of risks smaller than a few percent. When told something is “rare”, almost everyone interprets this as “maybe five or ten percent” – doesn’t usually happen, but still might once or twice. That’s the everyday meaning. But many medical risks are, not just “rare”, but statistically rare. Which often works out to, “you will never, ever, ever encounter this, unless you happen to be a medical professional”. The odds of a random person contracting many rare diseases are slimmer than them personally winning an Oscar.

- An awful lot is out of date. For one research project, I investigated the incidence of a group of diseases. More than half the sites I visited – even ones written by medical specialists – used data published over thirty years ago, which had been shown to be wrong over ten years ago. And it wasn’t a small mistake – they were off by a factor of twenty or thirty. Oops.

- Many recommendations are patently ridiculous. One site I read – again, maintained by reputable professionals – said to “call 911 immediately if you have any symptom of appendicitis”. Which sounds reasonable, until you read their list of symptoms. Among them were “loss of appetite”, “low fever”, “pain anywhere in the upper or lower abdomen or back”, “constipation” and “diarrhea with gas”. No one is ever going to rush to the ER because of indigestion. Nor should they – taken literally, this kind of nonsense would flood medicine with perfectly healthy people, making the sick harder to treat.

- Much is in the form of useless “lists”. How many times have you seen disclaimers like:

“Side effects may include headache, nausea, vomiting, death, dizziness, vaginal ejaculations, dysentery, cardiac arrhythmia, mild heart explosions, varicose veins, darkened stool, darkened soul, lycanthropy…. “

When you make a list, you’re implicitly claiming all the items are similar. The items on these “medical lists” almost never are. Some of them may be literally thousands of times more likely than others, or thousands of times more serious. It’s like getting in a boat, and being told “some possible risks are seasickness, chills, drowning, lightning strikes, shark attacks, death by sudden nuclear explosion… “

(Disclaimer: I work for a company which sells medical research, so I have reason to be biased. However, this essay was written by me on my own time, and is not any way endorsed or approved by my employer.)