Background: Blood culture is a critical test for diagnosing bloodstream infections. Frequent microbial contamination during sampling and testing leads to abuse of antimicrobial agents. We evaluated methods for reducing contamination and obtaining more reliable results.

Methods: We analyzed blood cultures obtained between 2009 and 2015. We established 6 quality indicators: true positive rate, contamination rate, blood sampling volume, number of sets of blood cultures, delayed transportation rate, and percentage of samples collected from the femoral region, with reference to the CLSI guideline M47-A, 2007. Education was provided for interns and nurses responsible for blood sampling and transportation of specimens, and data were analyzed monthly.

Results: At baseline, the true positive rate was 12.8%, and the contamination rate was 4.0%. During the intervention period, these were decreased to 10.9% and 1.9%, respectively. The percentage of samples smaller than 5 mL decreased from 29.7% to 2.7- 11.3%. The rate of one set of blood cultures being ordered was always ＜5%. The delayed transportation rate decreased from 35.6% to 5.5-7.7%. Finally, the percentage of samples collected from the femoral region decreased from 41.5% to 22.0-31.0%, because of which we did not attain our goal, 20.8%.

Conclusion: The results showed improvements in contamination rate, specimen volume, specimen transportation time, and the percentage of samples collected from the femoral region. The quality management of blood cultures in 2011 was comparatively poor, which led to increased contamination rate, large number of samples containing ＜5 mL of blood, and increased percentage of samples collected from the femoral region. Thus, quality improvement methods can produce more reliable results of blood cultures. (Ann Clin Microbiol 2015;18:88-93)