The Lion Story

Here’s a story on probabilities I`ve written for the FDA.

You have a first floor apartment at 75th Street and Fifth Avenue. While watching television one night, you hear what sounds like a lion’s roar. Well there aren’t any lions around there, so you say, “It sounded like a lion’s roar, but somebody must have knocked some garbage cans together, or a truck back-fired. It’s almost impossible for that to have been a lion.” You put it in probabilities and say, “It’s a million to one there’s no lion.” You use that low figure because it did sound like a lion.

Out of curiosity you go to your back window. It’s pretty dark, but you catch a glimpse of something large jumping over the fence. It looked too big for a dog, but it couldn’t be a lion. It must have been a dog. But with this extra piece of information you drop the odds from a million to one to a thousand to one. By now you’re curious and go into the back yard with a flashlight. You find tracks that would take one heck of a large dog to produce. Now, the combination of a roar, large figure and large tracks, makes you reassess the probabilities. You say, “Either I’m crazy, or there is a realistic probability that was a lion,” and you change the probability to a hundred to one against a lion.

Next morning you pick up the paper and see an article on the front page headed, “Circus Comes to Town.” Still thinking about the lion, you say, “By golly, the circus came to town,” and you drop the probability to ten to one against a lion. A subtitle says, “Lion Escapes,” and you change the probability from ten to one against a lion to a hundred to one there was a lion in your back yard. The next sentence says, “While the circus was going down Fifth Avenue, around 76th Street, a lion escaped.” Now it’s a million to one there was a lion in your back yard.

This shows how, starting from one end of a probability, fact after fact can change the probability, and it can swing almost 180 degrees in the other direction.

It’s the same with the drug phenytoin. When it was synthesized in 1908, the probability that the medicine would be useful for seventy or more symptoms and disorders might have appeared to be one in a billion. There had never been anything like that. But when the clinical evidence had been published by thousands of independent physicians (let me stress the word independent—they had no common interest other than to help others), in forty-eight countries, written in twenty different languages, and a host of basic mechanism studies support the clinical evidence, the odds that phenytoin is a broadly useful substance become astronomical And the notion that it`s only an anticonvulsant is absurd. But that’s how it’s listed with the FDA.

If a remarkable medicine is out of patent and sells for pennies, and no company deems it worth its while to go through the arduous and expensive procedures to get it approved by the FDA, should our public suffer?

The question is simple. Was the FDA established just to prevent the negative, or was it established for the health of the American public? It was certainly the latter. Action by them can correct a great catastrophe.