Methods: We undertook a factorial survey depicting hypothetical cases of CDI at the William S. Middleton Memorial Veterans Hospital, a tertiary-care facility affiliated with the University of Wisconsin. Resident physicians (n=20) in all years of training on rotation at the VA hospital and nurses (n=20) on medical-surgical wards comprised the study population. Clinical factors were chosen based on a literature review and preliminary survey of physicians and nurses on 36 CDI risk factors. Five factors selected for inclusion were: hospitalization, fever, stool odor, abdominal exam, and white blood cell count. Participants judged their probability of ordering a CDI diagnostic test and the probability that the patient has CDI on a scale of 0-100% for 20 factor combinations. The cues of a 65 year-old inpatient with two loose stools that day and history of clarithromycin 4 weeks previous were given in all vignettes. Responses were collected anonymously via convenience sampling and analyzed using ordinary linear regression.
Results: The most important factor in residents’ judgment of their likelihood of ordering a test and that the patient had CDI was elevated WBC count, with a 7.9% and 15.3% increased likelihood, respectively. Nurses’ most significant factor was abnormal stool odor, which increased their stated likelihood of testing by 19.5% and stated likelihood that the patient had CDI by 17.6%.
Conclusion: There were differences among resident physicians and nurses’ perceptions regarding the relative importance of factors likely to be associated with a positive test for CDI. For nurses, the presence of a malodorous stool significantly increased their likelihood of testing for CDI.
R. Wigton, None
E. Yanke, None
See more of: Poster Abstract Session