Evaluation of the Impact of Stool Bacterial Cultures on Clinical Management on Patients Admitted With a Primary Diagnosis of Diarrhea

Abstract

Introduction: Most cases of acute diarrhea are self-limited and due to viral or baterial infections. Stool cultures only test for the specific enteroinvasive bacteria, including Campylobacter jejuni, Salmonella species, Shigella species, Escherichia coli, which represent a small fraction of the infectious agents that can cause acute diarrhea. Previous studies have shown these organisms are positive in approximately 0.2 to 1.8% of cultures. We evaluated the effect of stool culture results on the management of inpatients with acute diarrhea. Methods: The database from the inpatient electronic medical records system at a university hospital was queried for stool bacterial cultures performed on inpatients from June 2010 through June 2014. Chart review was performed to assess the impact of stool culture results on clinical management and was limited to patients with an admitting diagnosis of acute diarrhea.Table 1: Stool Culture DistributionTable 2: Stool Culture DistributionFigure 1Figure 2Results: A total of 4576 stool cultures were evaluated within the study period. Of these stool cultures, 527 (11.5%) were positive, the most common organisms being: Candida sp. (n=282, 6.16%), Staphylococcus aureus (n=88, 1.92%), Campylobacter jejuni (n=25, 0.55%), Salmonella sp. (n=22, 0.48%), Shigella sp. (n=13, 0.28%). Escherchia coli 0157:H7 was not isolated in any cultures. Candida and S. aureus were considered contaminants and did not impact clinical management. This left 54 positive stool cultures for analysis, excluding duplicate cultures and those taken from patients without diarrhea. Seven cultures (15.5%) resulted in a change of the patient's management. Management changes included initiating antibiotics (n=4, 7.4%) and changing to narrower spectrum antibiotics (n=3, 5.6%). Results from 33 of the positive cultures (61%) were obtained after the patient was discharged with improved symptoms (with or without empiric antibiotic treatment). Results from the remaining 14 stool cultures (26%) were obtained during the patient's admission, but did not result in a change in management. Conclusion: In this large single center retrospective analysis, we demonstrate that bacterial stool cultures have a low rate of positivity, and positive culture results are of limited utility in guiding clinical management in patients admitted with acute diarrhea. Clinical judgment as to the benefits of culture results in guiding management should be considered prior to collecting them, and may enhance high value care and cost effectiveness.