Background: Because blood culture is a common test at our institution, we hoped to save money and reduce the risk of needlestick injury by modifying our current technique for the preparation of culture specimens without increasing the risk of contamination.
Objective: To compare the contamination rates of blood culture specimens obtained by the conventional method of switching to another sterile needle before inoculation with those of specimens obtained by a method without switching.
Design: Cross-over study.
Setting: Department of Medicine of university hospital.
Participants: Nursing personnel working in seven acute care medical wards in Siriraj Hospital, Bangkok, Thailand.
Interventions: From March to June 1991, participating nurses prepared blood culture specimens by means of both switch-needle and no-switch techniques in a cross-over study. All blood culture specimens were submitted to microbiology laboratory to determine the growth of microorganisms.
Measurement: Growth in a culture was considered to be "true positive" or "contamination" by predetermined criteria.
Results: Total number of blood culture specimens was 1619; of these, 813 were prepared by switch-needle technique and 806 were prepared by no-switch technique. The contamination rates were 7.6% and 8.3% for switch-needle and no-switch techniques, respectively (p = 0.61).
Conclusions: The switch-needle technique may not be necessary for obtaining blood culture specimens unless the needle is obviously contaminated. The no-switch technique for the preparation of blood culture specimens is more convenient and less expensive; it also poses less risk of needlestick injury.