Richard Smith’s guest blog post: The still small power of medicine

Medicine has made progress from Voltaire’s famous aphorism that “the art of medicine consists of amusing the patient while nature cures the disease,” but two articles just published in Cases Journal show that progress is slow and not all that it appears to be. Tom Jefferson and Enzo Grossi describe the agonising and currently insoluble problem of advising individuals based on evidence gathered in populations.

Let me try and illustrate with the following dialogue:

Inquisitive patient: If I take this drug will I be cured?

Complacent doctor: Yes.

Inquisitive patient: How do you know?

Complacent doctor: The drug company told me. They kindly sent me reprints from a prestigious medical journal.

Inquisitive patient: Did you read them critically?

Complacent doctor: Well no. In fact I didn’t read them at all. I don’t have time. But the drug rep who gave them to me was charming and convincing.

Inquisitive patient: What did he say?

Complacent doctor: He said that there was strong evidence from large randomised trials that the drug I’m prescribing you works.

Inquisitive patient: What’s a randomised trial?

Complacent doctor: Some say it was the most important medical discovery of the 20th century. In a randomised trial a large number of people randomly receive either the drug or a dummy pill. Researchers then see what happens to the patients. Nearly a thousand people were given the drug I’m prescribing for you and about the same number were given the placebo. In the treated group 50 died while in the untreated group it was 200. So the drug clearly works.

Inquisitive patient: What happened to the others?

Complacent doctor: They’re still alive.

Inquisitive patient: But are they cured?

Complacent doctor: Well, no. They still have the condition, but it’s clearly much better to have the treatment.

Inquisitive patient: OK, so I won’t be cured. I accept that, but are you sure the drug will benefit me?

Complacent doctor: Of course. Look at the difference between those who took the drug and those who didn’t. Those who didn’t had four times the chance of dying.

Inquisitive patient: It seems to me that there are four possible outcomes here. I might take the drug and still die. I might take the drug and not die but still have the condition. I might not take the drug and die. Or I might not take the drug and not die but still have the condition. How do I know which group I’ll be in?

Complacent doctor: Well, clearly you should take the drug because you quarter your chances of dying. That’s a big difference.

Inquisitive patient: But I might take the drug and still die or I might not take the drug and live. How do you know what will happen to me?

Complacent doctor: I don’t.

Inquisitive patient: And just because this drug worked in some people in this trial how do you know it will work in the future? Aren’t you making the mistake of Bertrand Russel’s “inductive turkey,” who assumed that because he got fed at 9 every morning he would always be fed—until Christmas Eve when he had his throat cut at 9 am?

Complacent doctor: I’m only a doctor not a philosopher.

Inquisitive patient: That’s your problem.

The articles in Cases Journal describe two ways to move on.

Tom describes the first method, which would be to create a giant reliable database of what happens to individual patients rather than to populations, but we would need it to be truly giant (ideally including every patient ever treated) and we would need a very sophisticated search engine. Cases Journal is working towards both ends.

The second method described by Enzo would be to gather huge amounts of variables on patients within large trials and then match the individual patient to patients in the trials using new methods of searching.

Both methods will need currently unimaginable computing power, new methods of searching, and – most difficult of all – a fundamental change in the working practices of doctors.
Richard Smith

Cases Journal


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: