91 - Implementing Criteria to Reduce Blood Cultures Ordering: A Pre- and Post-Intervention Retrospective Study in a Critical Access Hospital

Background: Blood culture utilization has been performed widely. Typically, clinicians order blood cultures in patients whom bacteremia is suspected. Our previous study showed that 35% of blood cultures performed in May 2019 could have been prevented since they did not meet the certain criteria. This study sought to examine the outcomes after education intervention by implementing criteria of blood culture ordering whether it could reduce unnecessary blood cultures.  

Methods: Electronic medical records of adult patients who had blood cultures done during pre-and post-study period were reviewed. Demographic data, clinical presentation, vital signs, location, quantities and sites of blood cultures were obtained. The measurement of qSOFA, SIRS and severe sepsis criteria were collected on the presentation. There were some clinical prediction rules for blood stream infection described in the previous studies. For this study, we use the criteria of at least 2 SIRS and/or at least one of the qSOFA criteria or severe sepsis to be a minimum indication for ordering blood cultures. The follow-up study was done after 6 weeks of educational intervention with implementation of criteria. Chi-square was used to compare the differences between two groups.

Results: There were a total of 165 patients included in our study (112 in pre- and 53 in post-intervention group). There were a total of 18 patients with positive blood cultures (12/112;10.71% in pre-intervention gr. vs 6/53;11.32% in post-intervention gr., p=0.91). Six out of 18 (33%) were deemed to be contaminated (3/12;25% vs 3/6;50%, p=0.29). Gram positive cocci were the most common organisms of the true positive blood cultures (10/12;83%). Of 165 patients, 78 (47%) had at least one of qSOFA (47/112;41% vs 31/53;58%, p=0.05), 18 (11%) had met severe sepsis criteria (9/112;8% vs 9/53;17%, p=0.09). There were 47 (28%) patients who had less than 2 criteria of SIRS and did not meet either criteria of qSOFA or severe sepsis (39/112; 35% vs 8/53; 15%, chi 6.87, p< 0.01). There was no true bacteremia in this group of patients.

Conclusion: Our study found that implementation of criteria for blood cultures successfully reduces the unnecessary blood cultures orders approximately 20% without missing true bacteremia in suspected patients.