Dozens of studies show that for many bacterial infections, a short course of antibiotics, measured in days, performs as well as the traditional course, measured in weeks. Yet many physicians have the mistaken belief that a longer course of antibiotics forestalls resistant strains. His Web site (bradspellberg.com/shorter-is-better) tracks randomized controlled trials that show how short courses of antibiotics compare with longer courses. A study published in May found that six weeks of antibiotics for infections around prosthetic joints was less effective than 12 weeks. And although antibiotics tend to be overprescribed for childhood ear infections, a longer course is more effective for kids younger than two.