A Bayesian approach to acute infectious diarrhea in adults.

Acute infectious diarrhea is a yearly occurrence for most Americans, and is associated with 1 million hospitalizations and about 6000 deaths in the United States annually. Up to 80% of acute infectious diarrhea is caused by noroviruses, which produce a clinically mild illness with a predictable short course and good outcome that make laboratory testing and antimicrobial treatment unnecessary. Most diarrhea-causing bacteria and protozoa can cause a clinical illness "like norovirus"; when they do so in healthy adults neither specialized testing nor antimicrobials is required. The presence or absence of epidemiologic evidence (such as travel, hospitalization, antibiotic use, other exposures)and clinical evidence (such as diarrhea frequency and duration, severity of abdominal pain and fever, character of stool, presence of chronic illness or immune deficiency) can change the probability of "not norovirus" from as low as 8% to as high as 100%. Such probabilities guide the use of laboratory testing and antimicrobial therapy in patients who have acute infectious diarrhea.