Original Article

A Comparison of Two Versus One Blood Culture in the Diagnosis and Treatment of Coagulase-Negative Staphylococcus in the Neonatal Intensive Care Unit

¹Department of Pediatrics, Children's and Women's Health Care Center of British Columbia, University of British Columbia, Vancouver, Canada

²Department of Paediatrics, Royal Hampshire County Hospital, Winchester, UK

³Department of Paediatrics, Mayday Hospital, Croydon, UK

⁴Department of Pediatrics, Ben Gurion University of the Negev, Beersheba, Israel

Correspondence to: Simon Struthers, MRCP, Consultant Pediatrician, Pediatric Department, Royal Hampshire County Hospital, Romsey Road Winchester, Hants SO22 5DG, UK

Abstract

OBJECTIVES: This study compares two versus one blood culture in the diagnosis and treatment of coagulase-negative staphylococcus (CONS) in babies with suspected sepsis.

STUDY DESIGN: The study was performed at British Columbia's Children's Hospital Neonatal Intensive Care Unit between March 1999 to March 2000. One hundred pairs of cultures were drawn from two percutaneous sites from babies more than 48 hours old at the time of a sepsis screen. CONS cultured from both sites was regarded as evidence of infection. Positive culture from only one of the two sites was regarded as contaminant. The difference in rates of diagnosed CONS infection and reduction in antibiotic usage when using two versus one blood culture was calculated.

RESULTS: In 5% of babies, cultures from a second site did not substantiate the diagnosis of CONS when compared to the result from a single culture. The resultant reduction in antibiotic use was 8.2%.

CONCLUSIONS: Two blood cultures reduce the number of children diagnosed with CONS infection and reduce antibiotic usage. Journal of Perinatology (2002) 22, 547-549 doi:10.1038/sj.jp.7210792

INTRODUCTION

Coagulase-negative staphylococcus (CONS) is increasingly being identified as a major cause of nosocomial late-onset sepsis (defined as sepsis after 48 to 72 hours) in neonatal units of industrialized nations.^1,2,3 Late-onset infection rates of 11.5% to 32.4% have been reported in very low birth weight infants, with 55% of these being caused by CONS.⁴

CONS is a common skin commensal, especially in premature babies, and can contaminate blood culture samples.^3,5,6 A false-positive blood culture can lead to an unnecessarily long course of treatment for some babies. Treatment of suspected late-onset sepsis in neonates often includes vancomycin to cover CONS, which is often resistant to methicillin and other beta-lactam antibiotics.⁷ There is concern that the repeated use of vancomycin in this setting may encourage the emergence of resistant bacteria, such as vancomycin-resistant enterococci,^8,9,10 Staphylococcus aureus,¹¹ and CONS,¹² and that efforts should be made to reduce its use.^9,13,14 The Hospital Infection Control Practices Advisory Committee of the Center for Disease Control and Prevention (CDC) has recommended "prudent vancomycin use." They have also recommended that "vancomycin use should be discouraged in response to a single blood culture positive for CONS, if other blood cultures drawn in the same time frame are negative, i.e. if contamination is likely."⁹ This is based on the fact that CONS is more likely to be the etiological infective agent, rather than contaminant, if the same organism is isolated from blood cultures obtained from more than one site.^9,15

This study was designed to compare a policy of two blood cultures rather than one in the diagnosis and treatment of babies over 48 hours old in our neonatal intensive care unit (NICU) with suspected sepsis. Our hypothesis was that following this policy and the CDC guideline would reduce the number of babies diagnosed with CONS that in turn would reduce the number of babies receiving a prolonged course of antibiotics, especially vancomycin.

PATIENTS AND METHODS

The study was carried out in the Special Care Nursery of British Columbia's Children's Hospital in Vancouver. This unit admits about 700 neonates a year and is the quaternary referral center for all neonates born in British Columbia.

Our usual practice, in any baby over 48 hours old with a single blood culture positive for CONS, was to treat with 7 days of antibiotic therapy, which included vancomycin. A blood culture still negative at 48 hours would result in therapy being stopped if the clinical setting allowed. Blood cultures from two percutaneous sites were not routinely taken. Between March 1999 and March 2000, a change of practice was initiated to reduce the use of vancomycin in the nursery. Based on points made in the CDC recommendations, physicians were requested to obtain a set of blood cultures (aerobic and anaerobic) percutaneously from two different sites when screening for sepsis or when blood cultures were otherwise clinically indicated, e.g., before a change in antibiotics. Cultures were usually taken within 15 minutes of each other. Cultures from central venous catheters (CVCs), percutaneous intravascular central catheters (PICCs), umbilical venous (UVCs), or umbilical arterial catheters (UACs) were not included in the study, although any baby who had two peripheral, percutaneous cultures taken may have had catheters in situ which would usually also be cultured. Babies had to be at least 48 hours old to be included in the study. The decision to screen for sepsis, or change antibiotic regimen, was clinically based. The physician documented indications for the blood cultures.

Blood cultures were taken from two separate sites after skin cleansing using the standard policy of the unit, comprising a two-phase cleansing with 70% isopropyl alcohol followed by chlorhexadine gluconate in 70% alcohol, left to act for 60 seconds. This combination in a previous evaluation had been shown to be effective in minimizing CONS skin contamination.⁶ One milliliter of blood was obtained from most peripheral sites, allowing 0.5 ml of blood to be inoculated in the anaerobic bottle (BD Bactec Standard Anaerobic/F) and 0.5 ml for the aerobic (Bactec Peds Plus/F*) bottle. Because of the difficulty of obtaining blood from some infants, 0.5 ml only was obtained from some sites. If only 0.5 ml was obtained, this was used for the aerobic bottle alone. The minimum amount of blood per culture bottle was 0.5 ml (This volume has been previously verified as adequate.¹⁶) A pair of cultures is therefore defined as an aerobic±anaerobic culture drawn sequentially from two separate peripheral sites from the same baby. The bottles were labeled with exact sampling time, to determine later which was taken first, and were processed using Becton Dickinson Bactec 9240 (Le Pont de Claix, France). The following data were extracted from the medical records: gestational age, birth weight, age at screen, ventilator requirement, physician's indication for drawing blood cultures (up to three reasons per pair of cultures), first and second blood culture results, presence of UACs, UVCs, CVCs, or PICCs. Data were obtained in all cases.

In accordance with CDC recommendations,⁹ only patients with positive CONS cultures from both sites were considered to have CONS bacteremia. Discordant results from cultures from two sites were regarded as evidence of contamination. CONS cultured from the first set of cultures but not the second indicated babies who would previously have been treated as having CONS infection but could be "saved" a long course of vancomycin by the negative result of the second culture. The reduction in vancomycin use (from 7 to 2 days) in these babies was calculated. The positive predictive value and sensitivity of the result from the first blood culture alone was calculated relative to the result using the combination of the two.

After March 1999, when the change of practice was initiated, there were still some babies who had only had a single blood culture taken during a sepsis screen. These babies could not be included in the study analysis, but their culture results from March 1999 to August 1999 were obtained for comparison.

RESULTS

One hundred pairs of cultures were taken from 69 babies of gestational age 24 to 40 (mean 28.9) weeks and birth weight of 450 to 4000 (mean 1283) g at age 3 to 190 days. Fifty-three babies were ventilated at the time, or soon after the cultures were taken. Fifty-one of the 100 were "complete" pairs, having sets of aerobic and anaerobic cultures from both sites. Thirty-four pairs had aerobic and anaerobic cultures from one site, but only aerobic cultures from the second site. Fifteen pairs had only aerobic cultures from both sites.

Indications listed by physicians for performing the blood cultures are as follows: apneas/bradycardias/desaturations (23%), pyrexia/temperature instability (18%), suspicious blood count/differential (15%), suspected necrotizing enterocolitis (11%), lethargy (10%), increased ventilatory requirement (6%), documentation of clearance of previous infection (6%), poor perfusion (3%), and others (8%). Eighteen babies had a CVC (11), a PICC (3), a UAC (2), or a UVC (2) in situ. One baby had both a PICC and UAC. Sixteen babies had CONS infection (with blood cultures from both sites growing CONS), and 10 babies grew CONS in cultures from only one of the two sites, indicating contamination. A total of 84 were therefore considered as negative for CONS infection. Five of 100 (5%) pairs of cultures grew CONS in the culture from the first site but did not grow CONS from the second. These babies would normally have been treated as having CONS infection if the first culture had been the sole test, but with the added benefit of the second culture, were treated as negative for CONS (contamination). The reduction in vancomycin use by "saving" these babies a 7-day course of vancomycin, and giving only 2 days of empiric therapy until the culture results were known, was 8.2% of the total vancomycin use.

CONS positive rate in the first set of cultures was 21% and was exactly the same in the second. Using a combination of two blood cultures, 16 had CONS infection compared with 21 (an increase of 31%) if just the first result was used. For a diagnosis of CONS, the result of the first culture alone showed a positive predictive value of 0.76 and a specificity of 0.94 when compared with the result from a combination of two blood cultures. In the babies who only had a single blood culture taken, between March to September 1999 (n=53), the rate of CONS-positive blood cultures was 19.0%.

Five babies who had a pure growth of organisms other than CONS in both sets of cultures (S. aureus (2), Escherichia coli (1), Enterococcus faecalis (2)). Three other babies grew an organism other than CONS in just one set of cultures (Serratia marcescens, E. coli, S. aureus).

DISCUSSION

Our results indicate that, in our nursery, using a combination of results from percutaneous blood cultures drawn from two separate sites can reduce the number of babies diagnosed with CONS infection. This enabled physicians to reduce the length of the vancomycin course in 5% of babies and reduce the total amount of vancomycin used by 8.2%. Using a single culture alone, rather than two, would have resulted in a 31% increase in the number of babies diagnosed with CONS infection. Of note, however, is that 20 babies needed to have blood cultures taken from two sites for every single reduction in the length of a vancomycin course and combining culture results from two sites did not change CONS diagnosis or treatment in 95% of sepsis screens.

During the study period, not all babies who had sepsis screens had blood cultures taken from two sites. This was due to many factors including inability to obtain enough blood from a second vein, the desire to minimize multiple percutaneous blood sampling attempts and occasionally due to the pressure of time in a busy unit. Systematic selection bias, however, is unlikely due to the similar rates of CONS-positive blood cultures in the study group and the babies who only had single cultures taken. This also makes it unlikely that physicians taking blood from babies in this study were more diligent in their skin preparation. The study included many babies who only had aerobic cultures taken from one or both sites. CONS, being an aerobic organism, should grow very well in this type of aerobic bottle and, for similar reasons to those stated above, it is often difficult to obtain enough blood for more than one blood culture bottle. Some babies had more than one pair of blood cultures taken, this is explained by the common occurrence in the NICU of some babies having many distinct episodes of suspected sepsis.

These results may not be generalizable to all nurseries. Patient populations, prevalence and incidence of late-onset CONS infection and the techniques of skin preparation, blood taking and culture analysis vary from nursery to nursery. Specifically, our skin preparation is more rigorous than some nurseries' and more thorough than some recommend.¹⁷ Given the study population and high prevalence of CONS in most level 3 nurseries, however, these results are likely to be relevant to many units.

Taking blood cultures from a second site requires valuable nurse and physician time, costs money, increases baby handling, inflicts pain, and further breaks the baby's fragile skin. However, reducing antibiotic usage may benefit the individual baby, potentially limits the emergence of resistant organisms, and may reduce overall costs.

Having reviewed our results and considered these overall harms and benefits, we have reverted to the practice of a single culture in our nursery. However, this study has shown that in NICUs with high rates of CONS-positive blood cultures, the policy of two cultures can be useful.

Acknowledgements

The authors thank D. Philippe Chessex for his helpful suggestions in reviewing the manuscript.

References

1 Gaynes RP, Edwards JR, Jarvis WR, Culver DH, Tolson JS, Martone WJ. Nosocomial infections among neonates in high-risk nurseries in the United States. National Nosocomial Infections Surveillance System. Pediatrics 1996; 98: (3 Pt 1) 357-61. MEDLINE

2 Isaacs D, Barfield C, Clothier T et al. Late-onset infections of infants in neonatal units. J Paediatr Child Health 1996; 32: 158-61. MEDLINE

3 Hall SL. Coagulase-negative staphylococcal infections in neonates. Pediatr Infect Dis J 1991; 10: 57-67. MEDLINE

4 Stoll BJ, Gordon T, Korones SB et al. Late-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996; 129: (1) 63-71. MEDLINE

5 St Geme JW III, Bell LM, Baumgart S, D'Angio CT, Harris CM. Distinguishing sepsis from blood culture contamination in young infants with blood cultures growing coagulase-negative staphylococci. Pediatrics 1990; 86: (2) 320-2. MEDLINE

6 Champagne S, Fussell S, Scheifele D. Evaluation of skin antisepsis prior to blood culture in neonates. Infect Control 1984; 5: (10) 489-91. MEDLINE

7 Baumgart MD, Hall S, Campos JM, Polin RA. Sepsis with coagulase-negative staphylococcus in critically ill newborns. Am J Dis Child 1983; 137: 461-3. MEDLINE

8 Uttley AH, George RC, Naidoo J. High level vancomycin resistant enterococci causing hospital infections. Epidemiol Infect 1989; 103: 173-81. MEDLINE

9 Hospital Infection Control Practices Advisory Committee (HICPAC). Recommendations for preventing the spread of vancomycin resistance: Recommendations of the Hospital Infection Control Practices Advisory Committee (HICPAC), CDC. Am J Infect Control 1995; 23: 87-94.

10 Leclerq R, Courvalin P. Resistance to glycopeptides in enterococci. Clin Infect Dis J 1997; 24: 545-56.

11 Hiramatsu K, Aritaka N, Hanaki H et al. Dissemination in Japanese hospitals of strains of Staphylococcus aureus heterogeneously resistant to vancomycin. Lancet 1997; 350: 1670-3. MEDLINE

12 Schwalbe RS, Stapleton JT, Gilligan PH. Emergence of vancomycin resistance in coagulase negative staphylococci. N Engl J Med 1987; 316: 927-31. MEDLINE

13 Sinkowitz RL, Keyserling H, Walker TJ, Holland J, Jarvis WR. Epidemiology of vancomycin usage at children's hospital, 1993 through 1995. Pediatr Infect Dis J 1997; 16: 485-9. MEDLINE

14 Isaacs D. Rationing antibiotic use in neonatal units. Arch Dis Child Fetal Neonatal Ed 2000; 82: F1-2. MEDLINE

15 Garner JS, Jarvis WR, Emori GT, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988; 16: (3) 128-40. MEDLINE

16 Jawaheer G, Neal TJ, Shaw NJ. Blood culture volume and detection of coagulase negative staphylococcal septicaemia in neonates. Arch Dis Child Fetal Neonatal Ed 1997; 76: (1) F57-8. MEDLINE

17 Practice guidelines for evaluating new fever in critically ill adult patients. Clin Infect Dis 1998; 26: 1042-59.

October/November 2002, Volume 22, Number 7, Pages 547-549

Table of contents    Previous  Article  Next    [PDF]