Evaluation of a Plastic Nonvented Aerobic Blood Culture Bottle for Use with the BacT/ALERT Microbial Detection System

Abstract

The current BacT/ALERT SA (BTA SA) aerobic blood culture bottle is made from glass, does not require venting, and contains a liquid emulsion sensor (LES). Its performance has been shown to be equivalent to that of the vented standard aerobic culture bottle. A further-improved version of the BTA SA bottle, designated the BacT/ALERT plastic SA (BTA PSA) culture bottle, is made from clear plastic to prevent breakage, does not require venting, and contains a modified LES (LES 2) to reduce the possibility of false positives. The BTA PSA provides a practical alternative to the current glass version of this bottle. The plastic bottle is also comparable to the current glass bottle in transparency and growth performance and additionally minimizes the exposure to infectious agents due to glass bottle breakage.

Exposure to blood-borne pathogens is a common risk factor associated with the collection of blood cultures. The risk of needle sticks and the breakage of glass blood culture bottles represent significant sources for exposure and thus should be regarded as potential safety issues for health care personnel. Furthermore, in an effort to reduce costs, some institutions have assigned the responsibility for the collection of blood cultures to nursing personnel who may not be as skilled as trained phlebotomists or who may lack training in the proper safety techniques associated with the collection of blood cultures.

In an effort to improve workplace safety for personnel who are responsible for the collection of blood cultures, a plastic blood culture bottle has been developed (bioMérieux, Inc., Durham, N.C.) for the purpose of providing a safer product by minimizing the exposure to infectious agents associated with breakage of glass bottles. The plastic bottle has been designed for use in the BacT/ALERT microbial detection system and utilizes two materials in a multilayer configuration to provide transparency, gas impermeability, and durability to withstand terminal sterilization, features previously available in a glass bottle (A. S. Massey and E. B. Moser, Abstr. 101st Gen. Meet. Am. Soc. Microbiol., abstr. C-5, 2001). Additionally, the plastic bottle does not require venting, which reduces the risk of needle stick injury and eliminates a step in the processing of bottles. The plastic bottle also contains a liquid emulsion sensor and provides a larger headspace for the proper atmospheric conditions. Preliminary studies comparing the performance of the plastic and glass versions of the BacT/ALERT standard aerobic, standard anaerobic, and charcoal-containing media demonstrated equivalent performance and showed that the plastic bottle provides a practical alternative to the current glass bottle (Massey and Moser, Abstr. 101st Gen. Meet. Am. Soc. Microbiol.; J. Walsh, M. Chi, C. Ronsick, B. Robison, and T. Thorpe, Abstr. 101st Gen. Meet. Am Soc. Microbiol., abstr. C-4, 2001). The study described here was designed to compare and demonstrate that the performance of the BacT/ALERT plastic standard aerobic (PSA) system, when used with the BacT/ALERT microbial detection system, is at least equivalent to the vented glass BacT/ALERT standard aerobic (VA) culture bottle for recovery and speed of detection of microorganisms in suspected cases of bacteremia or fungemia. A previous study had shown that the performance of a nonvented aerobic glass bottle was equivalent to that of the standard aerobic glass bottle; the latter is no longer commercially available (5).

This study was conducted in two tertiary-care medical centers, the University of Louisville Hospital (ULH), Louisville, Ky., and the William Beaumont Hospital (WBH), Royal Oak, Mich. Blood samples were collected by the phlebotomy team and nursing personnel at ULH and by the phlebotomy team at WBH (2, 4, 5). An aliquot of ca. 20 ml was obtained from patients with clinically suspected bacteremia or fungemia. Equal aliquots of blood were aseptically inoculated into each bottle with the VA bottle being inoculated first. Blood volumes were measured in the laboratory by comparing each bottle to known volume standards. Only blood culture sets with differences in fill volumes between bottles of ⩽30% of the larger volume were included in the analysis. Bottle pairs that did not meet these criteria were excluded from the study but were processed to maximize recovery of microorganisms from each culture. All VA bottles were transiently vented to air for at least 30 s prior to placement in the BacT/ALERT incubator module; PSA bottles were loaded directly. Cultures were incubated for a total of 5 days, with terminal subcultures performed at random on 20% of the instrument-negative PSA bottles to accurately establish the false-negative rate. When a bottle signaled positive, broth from the vial was Gram stained and subcultured onto sheep blood and chocolate agars, respectively. All isolates were identified by standard microbiologic procedures (3). False-positive bottles (i.e., those which had a positive instrument signal but were Gram stain negative and subculture negative) were reincubated until growth occurred, whereupon the bottles were reflagged as suspected positive, or until the original 5-day incubation period had expired. All bottles were processed independently of the other bottles in a given set, that is, a negative bottle was not examined when the other bottle in the set was flagged as a suspected positive. If these bottles remained negative for 5 days, they were terminally subcultured. Instrument-negative bottles that grew an organism on terminal subculture were categorized as false negative. The clinical significance of recovered microorganisms was assessed by chart review using published criteria (6).

To determine differences between the two aerobic bottles (PSA and VA) with respect to clinical sensitivity, McNemar's test for paired samples was applied to the positive yield data generated for compliant sets (1). A probability value of less than 0.05 would indicate a significant difference in the clinical sensitivities of the two bottles. The signed-rank test was used to determine the differences between the PSA and VA bottles with respect to time to detection (1). Differences were analyzed using a separate signed-rank test for each of the bacterial species. A probability value of less than 0.05 would indicate a significant difference in the time to detection between bottles. False-positive and false-negative rates were determined as the ratio of the number of false positives or false negatives to the total number of cultures tested.

Results for recovery and time to detection were compared for adequately filled bottle pairs, as defined above.

A total of 3,797 paired, PSA and VA samples were received for culture, of which 3,552 were adequately filled pairs. There were 407 positive cultures detected in one or both bottles. The PSA bottles detected 324 (9.1%) positives, 8 (0.23%) false positives, and no false negatives, while the VA bottles detected 291 (8.2%) positives, 4 (0.11%) false positives, and 1 (0.03%) false negative. The one false negative detected in the VA grew a Corynebacterium sp. on subculture. The corresponding PSA was positive for coagulase-negative staphylococci and Corynebacterium spp.

Of the 205 (201 monomicrobic and 4 polymicrobic) positive cultures with clinically significant organisms, 146 (71.2%) were detected by both bottles, 35 (17.1%) were detected by the PSA bottle only, and 24 (11.7%) were detected by the VA bottle only. The comparative yields of clinically significant bacteria and fungi from the two aerobic culture bottles are summarized in Table 1. There was no statistically significant difference between bottles in recovery of clinically significant organisms. The overall higher yield of microorganisms in the plastic bottle is most likely due to the increased headspace in the bottle (80 ml versus 73 ml in the vented bottle). The content of the plastic bottle headspace is 20% carbon dioxide in oxygen versus the same concentration of carbon dioxide in ambient air for the vented bottle. The latter also has a slightly higher vacuum.

TABLE 1.

Comparative yields of clinically significant microorganisms in BacT/ALERT PSA and VA blood culture bottles^a

Microorganism	No. of isolates detected by:			Pd value
	Both bottles	PSA bottles only	VA bottles only
Gram-positive cocci
Staphylococcus aureus	46	5	7	NS
Methicillin-resistant S. aureus	4	0	0	NS
CoNSb	28	17	7	NS
Enterococcus faecalis	7	0	3	NS
Enterococcus faecium	0	2	0	NS
Streptococcus pneumoniae	13	0	1	NS
Beta-hemolytic streptococcic	4	1	1	NS
Gram-negative bacilli
Acinetobacter baumanii	1	0	0	NS
Citrobacter koseri	1	0	0	NS
Enterobacter aerogenes	1	0	0	NS
Escherichia coli	19	3	2	NS
Flavobacterium spp.	0	1	0	NS
Haemophilus influenzae	0	1	0	NS
Klebsiella oxytoca	2	0	0	NS
Klebsiella pneumoniae	3	0	1	NS
Proteus vulgaris	1	0	0	NS
Pseudomonas aeruginosa	6	1	0	NS
Serratia marcescens	1	0	0	NS
Anaerobic bacterium	0	0	1	NS
Clostridium perfringens
Fungi
Candida albicans	4	3	1	NS
Candida glabrata	3	1	1	NS
Candida parapsilosis	1	0	0	NS
Candida tropicalis	1	0	0	NS
Yeast	0	0	1	NS
All microorganisms	146	35	26	NS

^aOrganisms were recovered from 201 monomicrobic cultures and 4 polymicrobic cultures; organism totals exceed the numbers of cultures because some cultures contained more than one organism.

^bIncludes 21 Staphylococcus epidermidis isolates, 1 Staphylococcus hominis isolate, and 30 isolates that were not identified to the species level.

^cIncludes five Group B streptococci and one Group A beta-hemolytic streptococcus.

^dNS, not significant (P > 0.05).

The recovery of clinically significant microorganisms in each culture system with blood from patients with septic episodes is summarized in Table 2. A septic episode was defined as the initial isolation of a significant organism, the subsequent isolation of a different significant organism after 3 days, or the isolation of the same significant organism at a time interval of greater than 7 days after the last positive culture. Isolation of a different significant organism within 3 days of the last positive culture constituted a polymicrobic episode. Of the 145 (143 monomicrobic and 2 polymicrobic) septic episodes, representing 141 patients, 103 (71.0%) were detected in both bottles, 24 (16.6%) were detected in the PSA bottle only, and 18 (12.4%) were detected in the VA bottle only. There was no statistically significant difference between bottles in the recovery of clinically significant organisms from septic episodes.

TABLE 2.

Clinical isolates recovered from septic episodes of monomicrobic (143) and polymicrobic (2) bacteremia or fungemia detected by BacT/ALERT PSA and VA culture bottles^a

Microorganism	No. of isolates detected by:			Pd value
	Both types of bottle	PSA bottles only	VA bottles only
Gram-positive cocci
Staphylococcus aureus	29	4	5	NS
Methicillin-resistant S. aureus	3	0	0	NS
CoNSb	20	11	4	NS
Enterococcus faecalis	4	0	3	NS
Enterococcus faecium	0	2	0	NS
Streptococcus pneumoniae	10	0	1	NS
Beta-hemolytic streptococcic	3	1	1	NS
Gram-negative bacilli
Acinetobacter baumannii	1	0	0	NS
Citrobacter koseri	1	0	0	NS
Enterobacter aerogenes	1	0	0	NS
Escherichia coli	14	2	1	NS
Flavobacterium spp.	0	1	0	NS
Haemophilus influenzae	0	1	0	NS
Klebsiella oxytoca	1	0	0	NS
Klebsiella pneumoniae	3	0	1	NS
Proteus vulgaris	1	0	0	NS
Pseudomonas aeruginosa	4	1	0	NS
Serratia marcescens	1	0	0	NS
Anaerobic bacterium,	0	0	1	NS
Clostridium perfringens
Fungi
Candida albicans	2	1	1	NS
Candida glabrata	3	0	1	NS
Candida parapsilosis	1	0	0	NS
Candida tropicalis	1	0	0	NS
Yeast	0	0	1	NS
All microorganisms	103	24	20	NS

^aOrganism totals exceed the numbers of cultures because some cultures contained more than one organism.

^bIncludes 12 Staphylococcus epidermidis isolates, 1 Staphylococcus hominis isolate, and 22 isolates that were not identified to the species level.

^cIncludes four group B streptococci and one group A beta-hemolytic streptococcus.

^dNS, not significant (P > 0.05).

The comparative mean times to detection for the different organism groups are summarized in Table 3 and include only bottle pairs from which the same, single, clinically significant organism was recovered. Although the mean detection times were generally faster in the plastic bottles, ranging from 2 h faster for the gram-positive cocci and gram-negative bacilli to 5 h faster for fungi, there was no statistically significant difference between the bottles.

TABLE 3.

Comparison of mean times to detection of growth when the same clinically significant microorganism was isolated from both bottles^a

Microorganism group	No. of isolates	Detection time (h)
		PSA bottles	VA bottles	Pa
Gram-positive cocci	99	15.3	17.9	NS
Gram-negative bacilli	35	13.8	15.5	NS
Fungi	9	46.9	51.9	NS
Total	143	16.9	19.4	NS

^aNS, not significant (P > 0.05).

The performance of the new PSA bottle was shown to be comparable to that of the VA bottle for the recovery and speed of detection of microorganisms. After this study was completed, the VA bottle was replaced by bioMérieux with the BacT/ALERT SA culture bottle, a nonvent glass bottle. The glass SA bottle was shown to have performance equivalent to that of the VA bottle in a previous study (5). The combined safety factors of not having to introduce a venting unit, the use of sharps, the minimal exposure to infectious agents due to the elimination of glass breakage, and the comparable rates of recovery and speed to detection make the plastic bottle a suitable alternative to the glass culture bottle.