The Utility of Blood Cultures in Non-Febrile Patients and Patients with Antibiotics Therapy in Internal Medicine Departments

Le résumé fourni par la source

Background: The injudicious use of blood cultures is associated with low cost-effectiveness and leads to unnecessary follow-up tests for false-positive results. In addition, false negatives can result in missed diagnoses, leading to delays in initiating appropriate treatment and potentially worsening patient outcomes. The timing of the blood culture tests related to the highest diagnostic yield is not fully elucidated. We hypothesized that a high proportion of the tests are done within non-optimal timing, resulting in a lower clinical yield. We specifically focused on the consequences of BC obtained in afebrile patients. Methods: We assessed 73,787 blood cultures taken between 2014 and 2020 in patients hospitalized with a suspected infection. Blood cultures were considered taken at optimal timing if the per rectum temperature was 38.3 °C or more and no prior antibiotics were given. Only the first culture per patient was assessed. The primary outcome was a true bacteremia defined by the clinically important pathogen. Results: Therefore, 25,616 blood cultures were obtained at optimal timing (34.7%), with true bacteremia found in 6.15% vs. 5.15% in cultures obtained at non-optimal timing. In a multivariable model, optimal timing adjusted for the variety of the clinical, demographic, and laboratory findings’ optimal timing was significantly associated with an increase in the odds of detecting true bacteremia (OR:1.23, 95% CI: 1.12–1.35). Conclusions: Nearly two-thirds of patients hospitalized due to a suspected infection did not have their blood cultures taken at the optimal time. Our findings underscore the importance of integrating clinical judgment, patient-specific risk factors, and evidence-based criteria when deciding to perform blood cultures, rather than relying solely on fever as an indicator.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.