Research In Practice Blog
Intro
Blood culture collection is an essential component of clinical care in the pediatric intensive care unit (PICU), but practices vary regarding the preferred source of blood cultures, frequency of repeat cultures, and surveillance culture practices in asymptomatic patients. The overuse of blood cultures can lead to false-positive results, increased length of stay and health care costs, and contribute to unnecessary antibiotic administration. Understanding current practice is critical for setting targets for safe and feasible reductions in blood culture use.
The exploratory BrighT STAR: Leveraging Implementation Science for Blood Culture Reduction Approaches (LIBRA) Type 2 hybrid effectiveness-implementation trial examined indications and patient-level factors associated with blood culture use in the PICU in preparation for implementation of a diagnostic stewardship intervention. Researchers hope to establish best practices that safely reduce overuse. This study describes a multisite assessment of current blood culture practices in critically ill children.
Study Design and Methods
Nine PICUs from eight academic referral centers were enrolled in LIBRA. During the preintervention baseline period, each LIBRA site was asked to audit 120 randomly selected blood cultures collected from patients admitted to its PICU. Data on culture indications, sources, and patient characteristics (whether the patient was immunocompromised, was postoperative within 24 hours, was experiencing signs of sedative withdrawal, or had a central venous line in place) was captured. Sites were asked whether the audited culture was obtained in response to one of three indications: new symptoms, persistent symptoms, or for “asymptomatic” patients. Selecting an indication prompted branching logic that asked whether there was a documented suspicion for sepsis at the time of culture and whether there was a suspected viral infection or alternate cause of symptoms.
Findings
All enrolled sites performed culture audits, totaling 1,068 cultures across nice PICUs. Overall, 646 (71%) cultures were obtained in response to new symptoms, 206 (22%) in response to persistent symptoms, and 64 (7%) in asymptomatic patients. All site cohorts included asymptomatic patients, but three sites accounted for the majority (84%) of all asymptomatic cultures.
Most audited culture results were negative (95%), with only 59 (6%) positive for presence of bacteria. Of 59 positive culture results, 39% were obtained to follow up a previous positive culture result; 57% were drawn for symptoms, either new (47%) or persistent (10%); 3% were in asymptomatic patients. Among 28 positive culture results drawn for new symptoms, 14 had documented concern for sepsis at the time of culture order; among the six positive culture results drawn for persistent symptoms, two had a documented concern for sepsis. Three key findings emerged from this retrospective multicenter audit:
- PICU clinicians frequently obtain cultures in patients with an alternative explanation for fever.
- There is substantial variability among sites in the use of surveillance cultures.
- All cultures obtained for persistent symptoms following a previous negative blood culture were ultimately negative.
Clinical Implications
Findings of a 6% positivity rate with 30% to 75% false positives align with previous information on the overall yield of blood cultures and provide updated data about the incidence of bacteremia in critically ill children. In addition, we found that PICU clinicians frequently obtain blood cultures in patients with an alternative explanation for fever. This suggests that factors unrelated to the patient, such as unit practice style or reflexive testing decisions, may be driving the decision to culture, something we noted in our earlier qualitative work.
The data confirm earlier findings of clinician-reported variability in surveillance blood culture practices. Importantly, recent literature supports decreased use of surveillance cultures in the PICU for stable patients on ECMO, and our results highlight that further work is needed to disseminate such practice change.
All cultures in the cohort that had been obtained for persistent symptoms and also had a preceding negative blood culture result, ultimately had negative results. This finding may reflect the low utility of repeated testing for stable patients who have already been demonstrated not to have bacteremia and represents another key target for stewardship efforts.
The results raise important questions about the utility of surveillance cultures and the value of repeat cultures in patients who have already had a negative culture result. Blood cultures will remain a critically important test for a “can’t miss” diagnosis. However, given the consequences of false positive results, and the frequent suspicion for alternative diagnoses, efforts to safely reduce culture overuse with diagnostic stewardship approaches are warranted. The diagnostic stewardship program informed by this study includes specific practice recommendations to reduce unnecessary surveillance cultures.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding organizations. The funding organizations did not participate in this work.