System with a Conventional Broth System

JOURNAL OF CLINICAL MICROBIOLOGY, Sept. 98, p. 55-5 95-7/8/955-6$./ Copyright C 98, American Society for Microbiology Vol. 6, No. Clinical Laboratory Comparison of a Slide Blood Culture System with a Conventional Broth System M. A. PFALLER, T. K. SIBLEY, L. M. WESTFALL, J. E. HOPPE-BAUER, M. A. KEATING, AND P. R. MURRAY* Clinical Microbiology Laboratory, Barnes Hospital, Saint Louis, Missouri 6 Received March 98/Accepted May 98 The recovery of bacteria and fungi from blood cultures was compared in conventional tryptic soy broth (TSB) bottles and in TSB bottles with an agarcoated slide attachment. A total of,66 sets of blood cultures, including 4 that were positive (5.5%), were evaluated. Significantly more gram-positive and gram-negative bacteria were recovered in the slide culture bottles than in conventional bottles (99 versus 5 isolates). Growth of gram-positive organisms and fungi was detected in the slide culture bottles 4 to 48 h earlier than in the TSB bottles. In addition, 76% of the isolates in the slide culture system were detected on the agar slide. In comparison, only 4% of the isolates in the TSB bottles were detected initially by blind subculturing. The incidences of contamination were.7% (7 cultures) for the slide culture system and.5% (9 cultures) for the TSB bottles. There have been numerous studies documenting the efficacy of biphasic blood cultures for the recovery of fungi from patients with suspected fungemia (, 6, 9). Although this technique was described originally for the isolation of Brucella spp. (4), very few investigators have used it for the recovery of bacteria from blood cultures (5, ). A modification of the standard biphasic blood culture system has been developed by Roche Diagnostics (Div. Hoffmann-La Roche, Inc.). The system consists of a routine blood culture bottle onto which can be attached a cylinder that contains an agar-coated slide with chocolate agar on one side and MacConkey and malt agars on the other side. The blood culture broth medium can then be subcultured onto the agar slide by tipping the bottle at the time of macroscopic examination. Bryan reported both better and faster recovery of organisms with this system than with a conventional system (). However, these results were inconclusive because different media and blood volumes were used in the two systems. Thus, to extend these observations we conducted a comparative study with similar broths and equal blood volumes and examined the efficacy of the slide culture system in the microbiology laboratory of a large U.S. teaching hospital. MATERIALS AND METHODS During this study, three blood culture broths were inoculated: the Roche bottle that contained 7 ml of tryptic soy broth (Roche Diagnostics) with.5% 55 sodium polyanetholesulfonate (slide culture bottle); a -ml bottle that contained tryptic soy broth with.5% sodium polyanetholesulfonate (TSB bottle; Difco); and a -ml bottle that contained Thiol broth with.5% sodium polyanetholesulfonate (Thiol bottle; Difco). Approximately ml of blood was collected aseptically by venipuncture from patients with suspected bacteremia or fungemia, and an equal volume (5 to 7 ml) was placed into each bottle. Upon receipt at the laboratory, the TSB bottle was vented chronically with a sterile cotton-plugged needle, and the Thiol bottle was not vented to promote recovery of anaerobes (). Both the TSB and Thiol bottles were examined visually twice during the first 4 h and daily thereafter for evidence of microbial growth. Gram stains were performed on any bottles with macroscopic evidence of growth. In addition, blind subcultures to chocolate agar plates were made from the TSB and Thiol bottles on days and of incubation. Before incubating the Roche slide culture bottles, the agar-coated slide system was attached to the bottle by removing the bottle cap, flaming the neck of the bottle, and screwing the slide unit securely in place. These initial manipulations were performed in a laminar flow safety cabinet to minimize contamination. All subsequent manipulations were performed at the bench. No attempt was made to keep the slide culture system anaerobic. The cultures were examined twice during the first 4 h and daily thereafter, as for the TSB and Thiol bottles. Subcultures of the slide culture bottles were performed immediately after placement of the slide unit and, subsequently, at the time of macroscopic examination by transiently inverting the bottle to allow the blood-containing medium to flow over the agar slide. Thus, the first subculture of the slide culture bottles was performed on the average 6 to Downloaded from http://jcm.asm.org/ on October 6, 8 by guest

56 PFALLER ET AL. h before the first subcultures of the TSB and Thiol bottles. The agar surfaces of the slides were inspected daily for evidence of growth. No additional blind subcultures were performed. Gram stains were performed on bottles with macroscopic evidence of growth (turbidity, hemolysis, gas). If growth on the agar surface was observed concomitant with macroscopic evidence of growth in the broth, the initial mode of detection was considered to be growth on the agar. RESULTS During the months of this study,,66 sets of blood cultures were collected, of which 4 (5.5%) from 96 patients were positive. A total of 47 isolates were recovered from these cultures, including 79 in the slide culture bottles, 96 in the TSB bottles, and 89 in the Thiol bottles (Table ). Single isolates of Corvnebacterium sp., Propionibacterium sp., Staphylococcus epidermidis, Peptococcus sp., Bacillus sp., and Clostridium sp. were considered to be contaminants and were recovered in 9 cultures, for an overall contamination incidence of 4.% (9 of,66 total cultures). A total of 8 contaminants were recovered in 7 (.7%) of the slide culture bottles compared with 4 contaminants in 9 (.5%) of the TSB bottle cultures and 5 contaminants in 4 (.%) of the Thiol bottle cultures (Table ). These contaminants were excluded from the remaining analysis to define more precisely the performance of the slide culture bottles and TSB bottles. Therefore, the number of true-positive cultures (total positive cultures minus the contaminated cultures) was 6, for an overall true-positive incidence of.9%. A total of 99 isolates were detected in the slide culture bottles, and 5 were detected in the TSB bottles, with significantly more isolates of Enterobacteriaceae (P <.5) and Staphylococcus sp. (P <.5) recovered in the slide culture bottles (Table ). Streptococci were also recovered more frequently in the slide culture bottles, but this was not statistically significant. There was no significant difference in the overall recovery of fungi between the two systems. Of the 7 isolates detected in both bottles, 5% were detected initially in the slide culture bottles, 5% were detected initially in the TSB bottles, and 6% were detected simultaneously in both bottles (Table 4). Gram-positive bacteria and yeasts, but not gram-negative bacteria, were detected an average of to days earlier in the slide culture bottles. One advantage of the slide culture system is that the bottles can be subcultured each time they are examined. Thus, isolates may be detected more frequently by examination of the surfaces of the slides rather than by macroscopic examination of the broths. To evaluate this, we J. CLIN. MICROBIOL. compared the method by which growth was detected initially in the slide culture bottles and TSB bottles (Table 5). Of the 99 isolates detected in the slide culture bottles, 7 (76%) were detected on the agar slide, including 65% of gram-negative bacteria, 88% of gram-positive bacteria, and all 5 fungi. A total of 5 isolates were detected in the TSB bottles, with only (4%) detected initially by blind subculturing. DISCUSSION Previous evaluations of biphasic blood culture media have established their utility in the detection of fungi in the blood of patients with suspected fungemia (, 6, 9). Additionally, Hall and co-workers demonstrated that biphasic blood culture medium was comparable to conventional broth blood culture medium for the recovery of most bacteria and was slightly better for the recovery of Staphylococcus aiureius (5). The slide culture system, which is a modification of the biphasic system, has been evaluated previously in the United States by Keathley and Walker (Abstr. Annu. Meet. Am. Soc. Microbiol. 979, C7, p. ), who found that the recovery of organisms was similar for the slide and conventional systems, although contamination was greater in the slide culture bottles (e.g., 6.% versus.6%). However, they reported faster recovery of organisms with the slide culture bottles. Apart from a limited study reported in the literature (), there have been only two other recent evaluations of the slide culture system. Bryan found that the slide culture system, compared with an 8-ml Vacutainer blood culture system (Becton, Dickinson & Co., Rutherford, N.J.), was more sensitive in the detection of bacteremia and reduced the time required for specimen processing (). The disadvantages of the slide culture system were an increased contamination rate (4.%c versus.8%) and the inability to recover anaerobes routinely. However, this study was deficient in several ways. First, the number of isolates in each group of organisms was too small to allow meaningful comparison of the two systems for individual groups of organisms. Second, although it appears that the slide culture system was more sensitive than the Vacutainer blood culture system, this may be an artifact due to the larger volume of blood cultured in the slide culture bottle. Third, the method by which positive cultures were detected was not presented (e.g., growth observed on the slide or subculture plate versus macroscopic detection in the broth). A more recent study has been reported by Dubois and Fontaine (Abstr. Annu. Meet. Am. Soc. Microbiol. 98, C55, p. 4). Although their study was quite limited, the results were similar Downloaded from http://jcm.asm.org/ on October 6, 8 by guest

VOL. 6, 98 SLIDE BLOOD CULTURE SYSTEM 57 TABLE. Number and type of organisms isolated in slide culture, TSB, and Thiol bottles No. of isolates in: Total no. of Organism Slide TSB Thiol positive Escherichia coli Klebsiella pneumoniae K. oxytoca K. ozaenae Enterobacter aerogenes E. cloacae Serratia marcescens Citrobacter freundii Proteus mirabilis P. vulgaris Morganella morganii Pseudomonas aeruginosa P. maltophilia Acinetobacter sp. Neisseria gonorrhoeae Branhamella catarrhalis Moraxella sp. Capnocytophaga sp. Bacteroides fragilis B. thetaiotaomicron Staphylococcus epidermidis S. aureus Streptococcus Group A Group B Group D Enterococci Viridans S. pneumoniae Corynebacterium sp. Propionibacterium sp. Peptococcus sp. Bacillus sp. Clostridium sp. Fungi Candida albicans C. tropicalis C. parapsilosis Cryptococcus neoformans Aureobasidium sp. bottles bottles bottles cultures 5 5 68 4 86 5 8 4 8 5 7 4 6 6 8 59 6 9 5 4 6 9 6 6 S I 8 4 5 4 4 6 8 9 I 4 5 6 4 4 78 to those of Bryan. However, contamination in their slide culture system was as high as 4%. In the present study, comparison of the slide culture system with a conventional -ml TSB bottle system revealed several advantages. More isolates of Enterobacteriaceae, Staphylococcus sp., and Streptococcus sp. were detected in the slide culture system. Additionally, more isolates were detected initially on the agar slides in the slide culture system than were detected by subcultures from the TSB bottles. The slide culture system was particularly effective for the detection of gram-positive organisms and Candida albicans. A total of 4% of the cultures containing gram-positive organisms, and % of the cultures containing C. albicans were initially detected in the slide culture bottles (Table 4). One possible explanation for the enhanced detection of gram-positive organisms and yeasts is that visible turbidity in the broth medium is normally delayed because these organisms tend to clump and settle to the bottom of the culture bottle. Daily inversion of the slide culture bottles resuspends the aggregates and inoculates the slides. In general, the gram-positive organisms were detected 4 h earlier and Downloaded from http://jcm.asm.org/ on October 6, 8 by guest

58 PFALLER ET AL. J. CLIN. MICROBIOL. TABLE. Presumed contaminants in each bottle No. of contaminants recovered in: Organism Slide T Total no. of culture SB Thiol contaminants bottles bottles bottles Staphylococcus epidermidis 46 6 7 7 Corvnebacterium sp. 7 Propionibacterium sp. 5 7 Peptococcus sp. Bacillus sp. 7 8 Clostridium sp. yeasts were detected 4 to 48 h earlier in the slide culture system. These findings are compatible with the findings of other workers who reported improved recovery of fungi and bacteria in biphasic blood culture systems (, 5, 6, 9). Frequently, in the case of gram-negative isolates, the organisms were detected concurrently in both the slide culture system and the TSB bottles. Again, the advantage of the slide culture system was that growth was initially detected on the agar slides, whereas the initial means of detection in the TSB bottles was most frequently the observation of macroscopic growth in the broth (Table 5). Although isolated colonies could usually be picked from the agar slides for identification and susceptibility tests, confluent growth of some gram-negative bacteria was seen. These isolates had to be subcultured to identify mixed cultures. Previous evaluations of this system have emphasized its time-saving capability. Although we did not perform a formal time study, we agree that the slide culture system can reduce the technical processing time of blood cultures by eliminating the routine subculturing of one bottle. This time-saving capability should justify the somewhat greater cost per bottle of the slide culture system ($ to $4.4 per bottle plus slide unit) versus that of the conventional system ($.55 per bottle). In addition, the cost savings on media and supplies formerly used for blind subculturing should also be substantial. As mentioned by Bryan, the slide culture system is not readily adaptable for anaerobic culturing (). Although this is somewhat of a drawback, there are considerable data that suggest that routine anaerobic subculturing is unnecessary (7, 8) and that this should not be considered a major disadvantage of the system. We found few major problems with the slide culture system. One early problem was an apparent effect of the broth media on the staining characteristics of gram-negative organisms. On several occasions, gram-negative organisms seen in the broth of the slide culture system appeared to stain more faintly than the same organisms in the TSB bottles. Subsequent work with a new lot of medium did not reproduce this TABLE. Organisms recovered in slide culture bottles alone, in TSB bottles alone, or in both bottles No. of isolaites in: Organism Slide culture TSB bottles Both bottles alone alone bottles Enterobacteriaceae Pseludomonas sp. Other Staphylococcus epidermidis S. aureus Streptococcus sp. Other Fungi C. albicans C. tropicalis C. parapsilosis CI-vptococcus tneofor-itnatis Auireobasidiuin sp. 7 I S 4 7 () 8 ) () 4 6 8 6 7 Downloaded from http://jcm.asm.org/ on October 6, 8 by guest

VOL. 6, 98 SLIDE BLOOD CULTURE SYSTEM 59 TABLE 4. Time of detection of positive cultures No. of isolates first detected in: Organism Slide culture TSB Slide culture bottles bottles bottles and TSB bottles Enterobacteriaceae 4 9 8 Pseudomonas sp. Other Staphylococcus epidermidis 8 8 S. aureus 8 9 Streptococcus sp. 7 4 9 Other 5 Fungi (C. albicans) 7 phenomenon, which was apparently related to tural defects that have been reported previously the initial lot of broth medium. (). Our performance of the initial slide culture Previous investigators have found a higher system manipulations in a laminar flow safety incidence of contamination in this system com- cabinet, not considered by us to be necessary for pared with conventional broth systems (; J. W. processing, may also account for the lower Keathley and C. N. Walker, Abstr. Annu. Meet. contamination rate seen in our study. There is Am. Soc. Microbiol. 979, C7, p. ). The no indication in the previous studies as to whethoverall contamination incidence in this study er this precaution was taken. was 4.%, which is slightly higher than the In summary, the slide culture system offers conventionally acceptable incidence of to % several advantages over conventional broth (). The slide culture system had a higher con- blood culture systems. It provides enhanced tamination incidence than the TSB bottles (.7% detection of many organisms and produces isoversus.5%). However, it was significantly low- lated colonies with which to perform biochemier than that reported in the studies mentioned cal and susceptibility testing, 4 h earlier than above. We ascribe this to modifications of the conventional systems in many cases. These adsystem by the manufacturer rather than to any vantages are particularly true for the gramspecial processing procedure in our laboratory. positive cocci and fungi. Thus, the slide culture In support of this impression, we encountered system is a satisfactory alternative to convenvery few problems with leaking or other struc- tional broth blood culture systems and may TABLE 5. Initial method of detection in the slide culture and TSB bottles No. of isolates detected initially by: Slide culture bottles TSB bottles Organism Macroscopic Agar slide Macroscopic Subculturing examination examination examination Enterobacteriaceae 56 85 9 Pseudomonas sp. Other 8 Staphylococcus epidermidis 8 4 S. aureus 5 7 4 Streptococcus sp. 8 7 5 Other 6 6 Fungi Candida albicans 6 Other Candida spp. Cryptococcus neoformans Aureobasidium sp. Downloaded from http://jcm.asm.org/ on October 6, 8 by guest

5 PFALLER ET AL. result in decreased technical processing time as well. ACKNOWLEDGMENT This work was supported in part by training grant -T- A77 from the National Institute of Allergy and Infectious Diseases. LITERATURE CITED. Bartlett, R. C., P. D. Ellner, and J. A. Washington II. 974. Cumitech, Blood cultures. Coordinating ed.. J. C. Sherris. American Society for Microbiology, Washington, D.C.. Bryan, L. E. 98. Comparison of a slide blood culture system with a supplemented peptone broth culture method. J. Clin. Microbiol. 4:89-9.. Caplan, L. M., and W. G. Merz. 978. Evaluation of two commercially prepared biphasic media for recovery of fungi from blood. J. Clin. Microbiol. 8:469-47. 4. Castaneda, M. R. 947. A practical method for routine blood cultures in brucellosis. Proc. Soc. Exp. Biol. Med. 64:4-5. J. CLIN. MICROBIOL. 5. Hall, M., C. A. Mueske, D. M. Ilstrup, and J. A. Washington II. 979. Evaluation of a biphasic medium for blood cultures. J. Clin. Microbiol. :67-676. 6. Kiehn, T. E., C. Capitolo, J. B. Mayo, and D. Armstrong. 98. Comparative recovery of fungi from biphasic and conventional blood culture media. J. Clin. Microbiol. 4:68-68. 7. Murray, P. R., and J. E. Sondag. 978. Evaluation of routine subcultures of macroscopically negative blood cultures for detection of anaerobes. J. Clin. Microbiol. 8:47-4. 8. Paisley, J. W., J. E. Rosenblatt, M. Hall, and J. A. Washington II. 978. Evaluation of routine anaerobic subculture of blood cultures for detection of anaerobic bacteremia. J. Clin. Microbiol. 8:764-766. 9. Roberts, G. D., and J. A. Washington II. 975. Detection of fungi in blood cultures. J. Clin. Microbiol. :9-.. Scott, E. G. 95. A practical blood culture procedure. Am. J. Clin. Pathol. :9-94.. Steurenthaler, W. 979. Experience with a new blood culture system (BCB system < Roche >) in a pediatric clinic. Med. Welt. :7-76.. Washington, J. A., II. 975. Blood cultures: principles and techniques. Mayo Clin. Proc. 5:9-98. Downloaded from http://jcm.asm.org/ on October 6, 8 by guest