September 8, 2010

Dr. John Snow uses NUMB3RS to stop an epidemic!

To me, the history of the 1854 cholera epidemic in London reads a bit like a NUMB3RS episode (a CBS television show that ran for six seasons, starring a mathematician who used his math to help the FBI solve cases).

Most people back in the mid-1800s thought that illness was caused by breathing “bad air” or miasma—breathing poisonous vapors of some sort. However, Snow was skeptical of that theory. When there was a deadly outbreak of cholera in London, he talked to residents and created a map showing the location of the cases. Through these means he was able to trace the outbreak to the public water pump on Broad Street in Soho.

Snow was most convinced that the pump was to blame when he discovered that people who lived nearby who did NOT use water from that pump also did not come down with cholera.

Snow examined a sample of the water through a microscope, and ran some chemical tests as well, and he was not able to conclusively prove its danger. However, he still used statistics and his map to convince the local council to disable the water pump by removing its handle on this date in 1854.

And the outbreak stopped.

Later, it was discovered that this public well had been dug only three feet from an old cesspit, and it was being polluted by sewage. Snow was able to use statistics to show that a waterworks company was using water from sewage-polluted portions of the Thames River for use in homes—and those homes suffered from more cases of cholera than other homes. Because of Snow's work, our understanding of disease prevention and public health took a giant leap.

Yeah for numbers!

Learn more...

• Here is a virtual tour of a modern water treatment plant. 

• Here is a game in which you can be an epidemiologist (one who studies disease and epidemics). Look for the words "Role-Playing Games" at the left side, and click "GO." By the way, the sound didn't work for me...but I didn't need it, either... 

• Part of numbers and statistics is probability. Try using this dice-roll simulation. Roll the dice 5 time, then 50, 500, and 5000. How does the number of rolls affect the results?