You may never have heard of Paul Meier, but perhaps I can convince you he is one of the 2oth century’s greatest heroes. As I write from my “shelter in place” from the violent COVID-19 pandemic, I’m seeing news stories about potential treatments. The hope they convey for a treatment to end the death, suffering, uncertainty, and fear is overwhelming. Meier is the reason why we aren’t routinely administering potential treatments to people who are ill or dying from the coronavirus and he’s the reason we live in a society with so many safe and effective medical treatments. He’s primarily responsible for the idea and arguably the implementation of the Randomized Clinical Trial and the important reason we won’t have a vaccine available to the public, even one discovered today, until 2021 or 2022.
Meier was a biostatistician (and perhaps the first statistician to focus on medical research) who convinced the FDA to require a disciplined process of testing before medical treatments can be used on patients. Key to his method was randomization of patients in treatment and control groups. Prior to Meier researchers would compare outcomes of those who had received the treatment to those who had not, but without randomization these groups were not comparable. Randomization wasn’t an easy sell. Docs questioned how one could ethically withhold a treatment that could help a patient, as randomization demands?
â€œWhen I said â€˜randomizeâ€™ in breast cancer trials I was looked at with amazement by my clinical colleaguesâ€, Meier said in a 2004 interview published in the journal Clinical Trials. â€œâ€˜Randomize? We know this treatment is better than that oneâ€™, they said. I said â€˜Not reallyâ€¦â€™â€ (The Lancet, 2011).
Meier spent much of his life bringing about this change in practice and in so doing showed the importance of confounding factors when trying to infer a relationship between variables – the insight of applying randomization is a triumph of mathematical thinking. This change gave us a structured methodology that could be applied as a general recipe that maximized the reliability of the experimental results. When you consider the number of drugs and treatments that have received FDA approval (or not) under this standard, it’s not hard to imagine that Meier’s idea of randomization saved millions of lives by delivering safe reliable medical treatments. This is why it takes time to test proposed new vaccines and treatments and why there are no shortcuts.
Meier was a retired Professor in the Department of Biostatistics at Columbia when he died in 2011, a year before I started my first faculty position in the Department of Statistics there. I learned about his impact on FDA standards because his New York Times obituary was circulated through my department shortly after I started.