BACKGROUND Contaminated blood cultures lead to diagnostic challenges and place a burden on healthcare services.

AIM To determine the impact of introducing a clinical skills test (CST) as part of the medical licensing examination and an institutional education programme on the contamination rates of blood cultures.

METHODS A prospective cohort study was conducted from 2009 through 2013 in all wards of a tertiary-care teaching hospital. We evaluated the effects of the CST, which was added to the National Medical Licensing Examination in Korea (KMLE) in 2010 and our institutional education programme, which began in 2013. The medical interns in charge of collection of blood for culture were divided in three groups with presence or absence of CST and the institutional education programme. The primary outcome was the percentage of blood cultures contaminated in each group, which were compared using the Poisson regression model. Participants' self-rated scores for the blood draw procedure were also analysed.

FINDINGS Although introduction of the CST in the KMLE failed to reduce blood culture contamination rate (1.36% vs 1.35%; P = 0.734), the institutional education programme significantly reduced the contamination rate (1.35% vs 1.00%; P < 0.0001). Most participants answered that they always followed each step correctly except for waiting the recommended contact time after applying the antiseptic.

CONCLUSION The educational intervention, not the introduction of CST in the KMLE, was effective in reducing overall contamination rates.