The Doctor Who Washed His Hands

In 1843, John Stuart Mill published A System of Logic. In it he described what he called the method of difference. If two cases are alike in every respect but one, and the outcome differs, the difference is the likely cause. It is a modest idea. It is also one of the most expensive ideas in history to ignore.

Four years later, in Vienna, a young Hungarian obstetrician named Ignaz Semmelweis was living inside Mill's method without, as far as we know, having read him. The Vienna General Hospital ran two maternity clinics side by side. Physicians and medical students staffed the First Clinic, and midwives staffed the Second. Women were admitted to them on alternating days. Yet mothers in the First Clinic died of puerperal, or "childbed," fever at a far higher rate. The difference was so well known that women begged to be sent to the Second, and some preferred to give birth in the street.

Semmelweis went looking for the one difference that mattered. He ruled out crowding, climate, and the priest's bell. Then, in 1847, his colleague Jakob Kolletschka cut his finger during an autopsy and died of an illness that looked very much like the fever killing the mothers. The doctors, Semmelweis realized, came to the wards straight from the dissecting room. The midwives did not.

He ordered that hands be washed in a solution of chlorinated lime before examinations. Deaths in the First Clinic fell dramatically, to roughly the level of the Second.

This should have been the end of the story. It was the beginning of a worse one. Semmelweis's conclusion implied that the physicians themselves had been carrying death to their patients, and the profession found that harder to accept than the data. He had no germ theory to explain why it worked, and he was slow to publish. When he did argue, he argued with growing fury. His contract was not renewed. He left Vienna and spent years largely dismissed. He died in 1865 in an asylum near Vienna, at 47. Within a few decades, the work of Pasteur and Lister made his conclusion orthodoxy.

There are two easy morals here, and both are incomplete. One says trust the data. But the data was available to everyone in that hospital for years. The other says the lone genius is always right. But many lone voices are simply wrong, and Semmelweis did his own cause real harm.

The harder moral is about institutions. An organization rejects evidence most fiercely when the evidence indicts the people who must accept it. No dashboard fixes that. What helps is a culture that separates the finding from the blame, so that "we were the cause" can be said out loud without ending a career. It also helps to ask, before any review, whether the answer you are afraid of is the one you have quietly ruled out.

Mill gave us the method. Vienna showed us the price of a room that will not apply it to itself.