BACKGROUND: The yield for positive stool cultures with non–C-diff species in hospitalized patients with diarrhea is very low (2.6%-6.4%) and is even lower after 3 days (0.6%). Thus, many hospital laboratories will not accept or will discard samples obtained 72 hours after admission (the 3-day rule).
POPULATION STUDIED: Derivation of the guideline was made on the data obtained from adults 18 years or older admitted to a tertiary-care hospital in Germany. The validation portions of the study were conducted with adults in secondary and tertiary health centers in Switzerland, Spain, and England.
STUDY DESIGN AND VALIDITY: Five phases looking both retrospectively and prospectively at 10 cohorts of patients made up the study. Hospital laboratories analyzed stool samples for the following organisms and species: Salmonella, Shigella, Campylobacter, Yersinia, Vibrio, Clostridium difficile, and Escherichia coli. To establish a baseline prevalence of infection, patient charts were initially reviewed retrospectively over 1 year, during which 2391 stool cultures were ordered on 1182 patients. In phase II, patient charts were surveyed prospectively over 7 months to identify characteristics that predicted positive cultures. Specifically, recent family history of diarrhea or travel outside Europe, nausea/vomiting, 8 or more bowel movements per day, temperature higher than 38° C, and presence of abdominal pain were evaluated.
OUTCOMES MEASURED: The investigators sought to identify and validate criteria predictive of which patients with non–C-diff diarrhea were most likely to have positive stool cultures after 3 days in the hospital. Monetary costs and technician time were also measured.
RESULTS: A total of 3416 stool cultures were obtained from 1735 patients in phases I and II. Of these, only 34 (1%) were found to have a bacterial pathogen other than C-diff. Most of these (29) were detected on the first sample obtained from a patient, and 20 were positive in the first 72 hours after admission. Analysis of the risk factors gathered in phase II showed a significant association between a positive culture and a recent family history of diarrhea or travel outside Europe. Criteria identified during phase III predictive of positive cultures obtained from patients with diarrhea 72 hours or more after admission included: age 65 years or older with preexisting comorbidity, human immunodeficency virus (HIV) infection or neutropenia, suspected nosocomial outbreak, and nondiarrheal manifestations of enteric infections (eg, acalculous cholecystitis, fever of unknown origin, polyarthritis). Applying these guidelines to the patients in phase II would have resulted in missing only 1.2% of the positive samples sent after 72 hours. Calculated savings were $10,300 annually as well as 356 technician hours. Prospective testing of the guidelines in one cohort with 65 positive cultures resulted in missing only 2 positive cultures from patients who did not need treatment. Validation in a separate cohort of 330 cultures found that 42% of the cultures could have been avoided while missing only one positive culture.
The modified 3-day rule reduces unnecessary stool cultures in patients who develop non–C-diff diarrhea subsequent to hospital admission. Specific criteria (onset of diarrhea after 72 hours in a patient who has at least one of the following risk factors: age 65 years or older and preexisting comorbidity, HIV infection or neutropenia, suspected nosocomial outbreak, or suspected nondiarrheal manifestations of enteric infection) can be used to improve the yield of positive stool cultures. This rule should be considered in preference to both reflexive ordering of stool cultures for diarrhea and automatic denial of stool cultures after patients have been in the hospital for 3 days.