Review
Skin antiseptics in venous puncture site disinfection for preventing blood culture contamination: A Bayesian network meta-analysis of randomized controlled trials

https://doi.org/10.1016/j.ijnurstu.2016.04.004Get rights and content

Abstract

Objective

To compare the efficacy of several antiseptics in decreasing the blood culture contamination rate.

Design

Network meta-analysis.

Data source

Electronic searches of PubMed and Embase were conducted up to November 2015. Only randomized controlled trials or quasi-randomized controlled trials were eligible. We applied no language restriction. A comprehensive review of articles in the reference lists was also accomplished for possible relevant studies.

Review methods

Relevant studies evaluating efficacy of different antiseptics in venous puncture site for decreasing the blood culture contamination rate were included. The data were extracted from the included randomized controlled trials by two authors independently. The risk of bias was evaluated using Detsky scale by two authors independently. We used WinBUGS1.43 software and statistic model described by Chaimani to perform this network meta-analysis. Then graphs of statistical results of WinBUGS1.43 software were generated using ‘networkplot’, ‘ifplot’, ‘netfunnel’ and ‘sucra’ procedure by STATA13.0. Odds ratio and 95% confidence intervals were assessed for dichotomous data. A probability of p less than 0.05 was considered to be statistically significant. Compared with ordinary meta-analyses, this network meta-analysis offered hierarchies for the efficacy of different antiseptics in decreasing the blood culture contamination rate.

Results

Seven randomized controlled trials involving 34,408 blood samples were eligible for the meta-analysis. No significant difference was found in blood culture contamination rate among different antiseptics. No significant difference was found between non-alcoholic antiseptics and alcoholic antiseptics, alcoholic chlorhexidine and povidone iodine, chlorhexidine and iodine compounds, povidone iodine and iodine tincture in this aspect, respectively.

Conclusions

Different antiseptics may not affect the blood culture contamination rate. Different intervals between the skin disinfection and the venous puncture, the different settings (emergency room, medical wards, and intensive care units) and the performance of the phlebotomy may affect the blood culture contamination rate.

Introduction

Blood cultures are one of the most significant tests performed for the diagnosis and treatment of bacteremic patients. However, false-positive blood culture (blood culture contamination) accounts for up to 50% of all positive blood cultures (Little et al., 1997), which lead to longer length of hospital stay, inappropriate antibiotic use, additional laboratory test and extra hospital charges (Walzman and Harper, 2001, Zwang and Albert, 2006). Blood culture contamination is defined as identified pathogens that are coming from patients’ skin, contaminated recipients, healthcare personnel's skin or during laboratory procedures rather than the patients’ blood (Caldeira et al., 2011).

Some practices are undertaken to decrease rates of blood culture contamination. A trial conducted by Kim et al. (2011) demonstrated that application of routine sterile glove could decrease blood culture contamination rates. Blood culture contamination rates have been shown to decrease when blood samples were collected by a trained phlebotomist (Washer et al., 2013).

Due to the fact that patients’ skin at the venous puncture site is the most likely source of blood culture contamination (Washer et al., 2013), many antiseptics have been used to disinfect the skin at the venous puncture site to prevent blood culture contamination by reducing the count of bacteria of the resident flora, including, iodine tincture, alcohol, alcoholic povidone iodine, aqueous chlorhexidine, alcoholic iodine tincture, alcoholic chlorhexidine and povidone iodine. Chlorhexidine gluconate has been recommended as a topical antiseptic for infants greater than 2 months of age and adults by Centers for Disease Control and Prevention (O’Grady et al., 2011). However, there is no consensus regarding which antiseptics are superior for decreasing blood culture contamination rate (Malani et al., 2007).

Thus we conducted this network meta-analysis to summarize the results of randomized controlled trials to compare the efficacy of several antiseptics in decreasing the blood culture contamination rate.

Section snippets

Search strategy

Our network meta-analysis was conducted according to the PRISMA Statement (Moher et al., 2010) to search all relevant studies. We searched Embase and PubMed (from their date of inception to November 29, 2015) using combination of the following terms in all fields: “chlorhexidine OR iodine OR alcohol” AND “skin” AND “blood culture” AND “random*”. Only randomized controlled trials or quasi-randomized controlled trials (RCTs) were eligible. We applied no language restriction. A comprehensive

Identification of relevant studies

We identified 60 studies by searching in PubMed and Embase. After removal of duplicate studies, we retrieved 45 articles. Only 9 studies were related RCTs. One RCT (Kim et al., 2010), which was a conference abstract, was excluded. One RCT (Schifman and Pindur, 1993), which evaluated skin antiseptics different from any other included studies, was excluded. Eventually, seven studies (Calfee and Farr, 2002, Little et al., 1999, Mimoz et al., 1999, Nuntnarumit and Sangsuksawang, 2013, Suwanpimolkul

Non-alcoholic antiseptics versus alcoholic antiseptics

Five studies with 19,079 blood samples in the non-alcoholic antiseptics group and 14,555 blood samples in the alcoholic antiseptics group were included. No significant difference was found in this aspect (RR: 1.41, 95% CI: [0.96, 2.07], P = 0.08, I2 = 82%, Supplementary Fig. 1). However, obvious heterogeneity was detected among these studies.

. Comparison of blood culture contamination rate.

Chlorhexidine versus iodine compounds

Five studies with 6862 blood samples in the chlorhexidine antiseptics group and 11,003 blood samples in the

Discussion

Prior studies (Little et al., 1999, Schifman et al., 1998, Weinbaum et al., 1997) found that rate of blood culture contamination concerning 10% povidone iodine was significantly higher than that of 2% iodine tincture, which was not supported by our results. King and Price (1963) demonstrated that even 1% iodine tincture had greater microbicidal activity than 10% povidone iodine against skin flora in forearm and hand. Previous meta-analysis (Caldeira et al., 2011) showed significant higher rate

Conclusion

Different antiseptics may not affect the blood culture contamination rate. Different intervals between the skin disinfection and the venous puncture, the different settings (emergency room, medical wards, and intensive care units) and the performance of the phlebotomy may affect the blood culture contamination rate. More large well-designed RCTs are needed to support the conclusion of this network meta-analysis.
Conflict of interest: None declared.
Funding: None declared.

References (32)

  • A. Chaimani

    Graphical tools for network meta-analysis in STATA

    PLoS ONE

    (2013)
  • A.K. Chapman

    Safety of chlorhexidine gluconate used for skin antisepsis in the preterm infant

    J. Perinatol.

    (2012)
  • J.J. Deeks

    Analysing data and undertaking meta-analyses

  • N. Kim

    Effect of routine sterile gloving on contamination rates in blood culture

    Ann. Intern. Med.

    (2011)
  • T.C. King et al.

    An evaluation of iodophors as skin antiseptics

    Surg. Gynecol. Obstet.

    (1963)
  • T. Kunisada

    Investigation on the efficacy of povidone-iodine against antiseptic-resistant species

    Dermatology

    (1997)
  • Cited by (9)

    View all citing articles on Scopus
    1

    These authors are co-first authors and contributed equally to this work.

    View full text