Effect of the Initial Specimen Diversion Technique on Blood Culture Contamination Rates

Abstract

The initial specimen diversion technique (ISDT) was first described by Patton and Schmitt (J. Clin. Microbiol. 48:4501–4503, 2010, doi:10.1128/JCM.00910-10). This study looked at the effect of implementation of the ISDT on blood culture contamination rates at our center. We found a reduction of 30.34% in potential blood culture contaminants.

TEXT

Blood cultures are essential for the diagnosis of bloodstream infections. A positive culture can lead to a definitive diagnosis and targeted antimicrobial therapy (1). A major limitation is the potential for false positives in culture which occur as the result of contamination. These false positives can result in increased health care costs related to antimicrobial use, extended hospital stays, and laboratory charges and to adverse patient events (1–4). These additional care costs have been reported to be between $1,000 and $8,720 per contaminated blood culture (2–4).

Different methods to decrease contamination rates have been published, and these methods include systematic skin preparation, disinfection of the tops of culture bottles, and use of a phlebotomy team (1). Patton and Schmitt evaluated the initial specimen diversion technique (ISDT), in which the first milliliter of the venipuncture sample is discarded (or used for another lab test) (5). This technique led to a significant reduction in contamination rates. ISDT is based on the hypothesis that the skin plug aspirated during venipuncture is a major source of contaminating bacteria (5).

In November 2011, the ISDT, as described by Patton and Schmitt, was implemented at the Queen Elizabeth II Health Care Centre with no other changes to practice for performing blood cultures. In our institution, blood cultures are taken by either the phlebotomy team or trained nurses. In brief, the steps pre- and post-ISDT implementation include use of alcohol swabs to disinfect the tops of blood culture bottles, use of sterile gloves, use of chlorhexidine swabs to disinfect the site prior to venipuncture, and use of a butterfly needle for blood collection. After ISDT implementation, the top of a blood collection tube (sodium citrate is used in our institution) is cleansed with an alcohol swab and used to collect the initial small amount of blood drawn through the butterfly needle. This tube is then discarded or used for other testing. Following that, the same needle is used to collect blood cultures. The bottles our institute uses are Bactec Plus Aerobic/F medium and Standard Anaerobic/F medium bottles (Becton, Dickinson Inc., NJ, USA). The minimal required amount of blood is 3 ml, and the optimal volumes are 8 to 10 ml for aerobic blood cultures and 5 to 7 ml for anaerobic blood cultures. Once the blood cultures are collected and transported to the lab, they are placed on our blood culture instrument (Bactec 9240) and incubated for a maximum of 5 days.

The purpose of this study was to evaluate the impact of the ISDT intervention on blood culture contamination rates. We performed a retrospective review of all our blood cultures for the periods from November 2010 to October 2011 (preintervention) and December 2011 to November 2012 (postintervention).

All isolated coagulase-negative staphylococci (CNS), Propionibacterium acnes strains, diphtheroids (primarily Corynebacterium species), and Bacillus species were used as markers for contamination and considered probable contaminants, while Staphylococcus aureus, Streptococcus pneumoniae, and the enterobacteriaceae were used as markers for true bacteremia.

Pearson's chi-square test was used to evaluate significance in compared groups; significance was defined as a P value of less than or equal to 0.05.

There was no significant change in either the number of processed blood cultures (13,776 versus 13,369) or the number of positives before and after the intervention. In the preintervention period, there were 1,522 (11.0%) positives compared to the postintervention period, during which there were 1,445 (10.8%) positives.

No statistically significant change was found in the numbers of organisms suggestive of true bacteremia (Fig. 1). However, the number of CNS fell from 407 (preintervention) to 288 (postintervention) (P < 0.0001) (Fig. 1). In total, there were 143 fewer potential contaminants, with a relative reduction of 30.34% (P = < 0.0001) (Fig. 1), and a total reduction of our estimated contamination rates from 3.4% to 2.4%.

FIG 1.

Numbers of pathogens and probable contaminants pre- and postintervention.

In our experience, ISDT was a simple-to-implement intervention associated with minimal incremental costs and minimal needed training, as use of a discard tube was not a new technique for the phlebotomists or nurses. Also, there were no significant reported difficulties with implementation. The intervention likely resulted in significant savings and improvements in patient care. Extrapolating from the literature, we might expect the 143 fewer potential contaminants to correspond to saving between $143,000 to $1,246,960 (based on additional health care costs ranging from $1,000 to $8,720 per contaminated blood culture) (2–4).

While CNS is the most common blood culture contaminant, multiple positive cultures from the same patient can indicate a true infection. A limitation of our study is that we did not conduct a chart review to identify the sites of obtained blood cultures (line versus peripheral), numbers of positive cultures, and clinical conditions that would allow us to confidently determine that the CNS were contaminants. However, even if we estimate that 50% of the CNS identified represented true infection, the reduction would still be significant (P < 0.01). Also, because the numbers of pathogens between the two periods were consistent and the noted potential contaminants had a noticeable reduction in their rates, we believe that the reduction of CNS identified is most likely due to a decrease in CNS as a contaminant.

We suspect our savings to be on the lower side, since in our laboratory, we limit the extent of workup and do not report susceptibility results on many suspect contaminants. Another limitation in our study is that we do not know the extent to which ISDT was implemented. Compliance by our phlebotomy team was likely near complete, but a small proportion of blood cultures are collected by others. If our results were achieved despite incomplete implementation with compliance, further benefits can be realized.

We concluded that implementation of ISDT in our institution has led to a significant decrease in blood culture contamination rates. This has likely translated into significant cost savings.