People Want To Know About Algorithms—But Not Too Much

ONE FINE SUMMER morning, you show up for a scheduled appointment for your regular checkup at the highly rated local health clinic. You walk in feeling fine. You raise your arm and say, “Doc, it hurts when I do this.” The doctor responds, “Then don’t do that.” You both chuckle. But several tests later, your mood changes. Your internist, friendly old Dr. Culverton Smith, has some bad news.

“I’m afraid you’re at high risk for Tapanuli fever,” Dr. Smith says. “It’s inevitably fatal. The good news is that there’s a medication that can help prevent it. I recommend you start taking it immediately. It may make you light-headed, irritable, and occasionally nauseous. But that’s a small price to pay to avoid the fever.”

You are taken aback. “I’ve never heard of this disease,” you protest as he scribbles out a prescription. “How do you know I’m at risk for it?”

Dr. Smith shows you a computer printout filled with incomprehensible numbers and terms. But he points to a line near the bottom of the form that reads: “TAPANULI FEVER: 17.88.” “That risk factor’s way too high,” he explains. “We want to get it down below 5. The pills should help.”READ MORE ON: WIRED