Aarva

Asterisk ·Future-gazing

In Praise of Observational Evidence

Published 2026-06-30 00:00:00+00:00

Can you prove a medicine works just by studying the people who already take it?

0:00 / 32:26 · Narrator Gacrux

Context

Before a new medical treatment reaches a pharmacy, it usually has to pass a randomized controlled trial. For decades, that kind of testing has been treated as the absolute gold standard of science. But an essay published on 30 June 2026 challenges that assumption. Running those trials takes massive funding, and in global public health, they are often impossible or completely unethical to pull off. The alternative is simply observing what people are already doing. Thanks to modern computing and huge datasets, looking closely at things that already happened might actually be the smarter way to do science.

Show notes

An examination of the limits of randomized controlled trials and a case for relying more on observational data in public health. While clinical trials are the standard for testing medical treatments, they are often too expensive or unethical to run on a large scale. By outlining modern statistical tools, the author explains how researchers can adjust for bias in existing records to evaluate health outcomes without conducting new experiments.

Read on Asterisk →