Working Knowledge: Grow, Then Kill; April 2002; Scientific American Magazine; by Mark Fischetti; 2 Page(s)

You've had a fever for two days and have been coughing up phlegm, so you visit your doctor. He examines you, asks a few questions, suspects a bacterial infection and says, "I'm going to do a lab test." While you open your mouth and say, "Aaah," he rubs a long swab against the back of your throat to grab some infected cells. He then sends the swab to a lab at a local hospital or health department. If he had suspected other pathogens, he would have taken a sputum, blood, urine or stool sample and sent that, too. You go home, miserable, and wait for the results.

At the lab a technologist unpacks the sample at a bench beneath a hood that draws air away into a filter system. That way, pathogens don't become airborne in the room. From there, the specimen can undergo a variety of tests done in series, each step narrowing its possible identity until one specific pathogen is determined. The most common sequence in dealing with a bacterium is to grow more of it in a lab dish and then test it with carbohydrates and antibiotics to assess what it is and which drugs might kill it most effectively [see illustration at right]. Two days after your visit to the doctor, you get your test results.