Some ideas have to be revived more than once in order to stick, even when the way they work has become well-understood. In this regard, there may be no sadder or more instructive case than that of the young Austrian doctor Ignaz Semmelweis, who was appointed deputy head of obstetrics in Vienna’s General Hospital in 1846.
The hospital had two maternity clinics, which had markedly different death rates. In the first, about 10 percent of mothers died from puerperal fever, two and a half times as many as in the second. Women entering the hospital to give birth begged to be assigned to the second clinic because the first had such a horrific reputation; some even deliberately gave birth in the street and then pretended they had done so on the way to the hospital. Ignaz Semmelweis decided to find out why so many women in the first clinic were dying. He compared every difference between the clinics he could think of. He considered the equipment, the furnishings, overcrowding, the climate, even the religious practices of the staff. Nothing seemed like a promising culprit.
Then another doctor at the hospital, a friend of Semmelweis’s, died in 1847. He had been accidentally cut by a student’s scalpel while they were performing a postmortem examination. The autopsy of this unfortunate doctor showed pathologies resembling those in women who had died of fever in the first clinic. The first clinic was also a teaching clinic for medical students, while the second taught only midwives. Semmelweis made the great logical leap: Medical students were arriving in the first clinic to treat expectant mothers right after they had been cutting open bodies downstairs. They must, he reasoned, be carrying invisible “cadaverous particles” on their hands that then infected the mothers. At once, he ordered that after working with corpses, the students should wash their hands in chlorinated lime solution, rather than the normal soap and water. Chlorinated lime was known as a bleaching agent that cleaned thoroughly. Semmelweis found it the most effective way to get rid of the putrid smell of dead tissue; so, he reasoned, it might be destroying the cadaverous particles along with the smell. (As we would now say, it was a disinfectant.) The maternal death rate in the first clinic immediately dropped by 90 percent. A few months later, it was zero.
Semmelweis began to spread this message as widely as he could. And what did he get for his efforts? Hostility and ridicule. Europe’s greatest obstetricians called his view unscientific, lacking in evidence, and unsupported by any credible theory. No one knew why hand-washing prevented infection, so it was all too easy to decide that it didn’t. What was worse, Semmelweis’s idea implied that it was doctors themselves who were (unwittingly) killing their patients. Semmelweis became increasingly frustrated and angry, writing letters accusing his opponents of being murderers. Even his wife thought he was going mad. In 1865, at the age of 47, he was forcibly committed to an asylum. He died there two weeks later, most likely from an infection that took hold after he was beaten by guards while he was in a straightjacket. His vindication would come only posthumously, when Louis Pasteur proposed the germ theory of disease years later. Semmelweis had been basically right all along: His “cadaverous particles” were bacteria breeding in the dead bodies used for student autopsies.
Astonishingly, this message has had to be revived again even within the last decade. One of the pioneers of the new movement of “evidence-based medicine” is the pediatrician Don Berwick, who co-founded the Institute for Healthcare Improvement in Boston. Berwick noticed that thousands of intensive-care patients in American hospitals were dying every year from infections after having catheters inserted in their chests. In 2004, he found an obscure study suggesting that improved hygiene by hospital staff — more frequent and systematic hand-washing, combined with other practices such as using antiseptic wipes on patients — could cut the risk of such infections by more than 90 percent. Berwick announced loudly that 25,000 lives a year could be saved if these reforms were immediately adopted, and still he met with resistance and apathy: The protocols were adopted slowly and piecemeal rather than — as logic, economics, and simple humanity would seem to demand — overnight. But where they were adopted, the results were dramatic. In those American hospitals that signed up to his challenge and adopted his reforms, it was estimated that more than 100,000 patient deaths were prevented within 18 months.
But too many doctors, in the 2000s as in the 1840s, continued to act like members of a lofty priesthood, offended by the suggestion that they were not already working optimally. The obstetricians who ridiculed Semmelweis’s hand-washing complained that it wasn’t supported by any reliable theory. They were correct: It wasn’t. It was a black box, a concept from engineering that describes a mechanism we can’t see. With a black box, all we can see are the inputs and outputs, and we have to guess about what’s going on inside. And ideas can be like that, too. No one understood how hand-washing could reduce infection. What no one knew was that the right theory — transmission of disease by bacteria — was just around the corner.
Black boxes turn out to be common in the history of discovery. Things may be found to work well as medicines long before the reason they work is explained by molecular biology. The first modern steam engines were cobbled together by mechanics and inventors; a thorough scientific explanation of why they worked had to wait a century until the formulation of the laws of thermodynamics. Until then, they, too, were black boxes. Given what we have seen, it would be surprising if there were not other old, discarded ideas that are right even though we don’t understand how they could be. Just because we don’t know how something works doesn’t mean it doesn’t work.
Steven Poole’s Rethink: The Surprising History of New Ideas, published by Scribner, hits book stores this week.