ASA Connect

Dear New Yorker Editor,

Your article "When Dying Patients Want Unproven Drugs," by Gideon Lewis-Kraus, included a great deal about what other people had to say about biostatisticians. Perhaps you might consider talking to one.

I remember a talk by Tom Fleming* where he told us that in the many FDA advisory committees he served on, he often heard patients testifying "this drug saved my life." But he never once heard from a patient who had died on a trial. They never show up to testify. "I want you to go out there and speak for the patients who cannot speak for themselves," he told us. And that, as we see it, is our job.

The math we use works something like that in options trading, or a poker game. How much information do you need to reach a decision? I once felt uneasy about comparing what I do to these things. Our stakes are, after all, not money or chips but human lives. But at an FDA conference some years ago a patient advocate compared her decision whether to enter a trial to staking a poker game, and I felt better. And yet it isn't the same. Knowing when to hold and when to fold, for stakes this high, is very, very hard.

We know that the advice we biostatisticians give – "Never count your money when you're sittin' at the table" – is very hard advice, advice far harder to stomach when one is gambling for ones life, especially for the many patients who we know have no time to wait,  "'til the dealin's done" or otherwise. We know there isn't time enough.  And yet, if we don't give that advice, we know that the many patients who today have no voice, not just those who died but also those yet to come, will never ever ever have a voice if we don't do our best to speak for them.

Jonathan Siegel

Jonathan Siegel is a pharmaceutical biostatistician who works on oncology clinical trials.

*Thomas Fleming is a professor of Biostatistics at the University of Washington.