A comparison between Bactec® aerobic resin and hypertonic blood culture media

APMIS 96: 720-722, 1988

A comparison between BactecB aerobic resin and hypertonic blood culture media LENNART SJOBERG, HANS FREDLUND and ANN-SOFIE DUBERG Department of Clinical Microbiology, Orebro Medical Center Hospital, Orebro, Sweden

Sjoberg, L., Fredlund, H. & Duberg, A.-S. A comparison between Bactecw resin blood culture media. APMIS 96: 720-722, 1988. The presence of antimicrobial agents in patients' blood is thought to represent an important source of false-negative blood cultures. This has led to the incorporation of agents with inhibitory effects on antimicrobial drugs into culture medium. In the present study, Bactec'E aerobic resin-containing blood culture medium was compared with Bactecw hypertonic blood culture medium. 504 patients receiving cytostatic and/or antibiotic treatment were studied. Sensitivity calculations on detection of bacteremia in these patients gave 0.9 1 for the resin medium and 0.79 for the hypertonic blood culture system and showed a significant difference (p=O.O 16). In addition, the resin-containing system more rapidly detected positive cultures than the hypertonic system. Key words: Blood culture; radiometric; resin; antimicrobial agents. Lennart Sjoberg, Department of Clinical Microbiology, Orebro Medical Center Hospital, S-70 1 85 Orebro, Sweden.

Early and rapid detection of microorganisms causing bacteremia has high priority in clinical microbiology laboratories. The presence of antimicrobial agents, i.e. antibiotics and anti-tumor drugs, in patients is thought to represent an important source of false-negative blood cultures or to delay detection of possible bacteremia (4).This has led to the incorporation of different agents, such as resin ( 1 , 2, 5 , 6, 7, 9), with inhibitory effects on antimicrobial drugs into the culture medium. An increased sensitivity for Bactecm (Becton Dickinson, Johnston Laboratories, Towson, Maryland, USA) resin-containing medium No 16 compared with Bactecm isotonic media No 6 in detecting bacteremia is described (1, 4, 6). The detection rate was more rapid ( 1) but the resin-containing medium seemed most effective in detecting Gram-positive organisms (6, 9). Comparison between Bactecm hypertonic blood culture medium No 8 and Bactecm resin-containing media showed

Received April 13, 1988. Accepted May 2, 1988.

720

less difference in sensitivity and deserves further studies (9). In a pilot study at our laboratory regarding detection of bacteremia in antibiotictreated patients, a higher sensitivity for Bactecm resin-containing medium was found (0.83) compared with BactecB hypertonic medium (0.67). This study consisted of 242 pairs of blood-culture bottles from 130 patients, but the pairs were not always documented as being from the same blood sample. For these reasons, the present comparison study of the two aerobic blood culture media regarding sensitivity and rapidity in detection of bacteremia in patients treated with antimicrobial drugs was performed. MATERIALS AND METHODS At Orebro Medical Center Hospital during 1985/86, 56 patients treated with cytostatic drugs for leukemia, many of whom also received simultaneous antibiotic treatment at the Department of Hematology, and 448 patients treated with antibiotics at the Department of Infectious Diseases were studied. Patients were admitted to the study when septicemia was suspected and they were already being treated with antimicrobial drugs

SJOBERG el a[.

(cytostatic and/or antibiotic drugs). Venous blood was collected aseptically into a syringe, and the blood was immediately distributed in equal volumes (4-5 ml) into BactecQ blood culture medium No 7 (anaerobic isotonic), No 8 (aerobic hypertonic), No 16 (aerobic resin) and No 17 (anaerobic resin) and examined, using a radiometric method with BactecB instrument 460. In most instances the patients' blood was cultured three times. The bottles were incubated at 37 'C, and the aerobic bottels were continuously shaken during the first three days. The aerobic bottles were measured twice and the anaerobic bottles once daily for the first three days; thereafter all bottles were measured daily for a seven-day period. Gram-staining and subculturing of the blood-culture bottle was performed when the instrument gave a positive indication. Microorganisms were identified by standard methods.

.- L

01

I

2

3

b

5

6

7

B

Detection d%

Fig. 1. A comparison between two Bactecm aerobic blood culture systems No 8 (hypertonic) and No 16 (resin-containing) in 53 patients with a positive culture in at least one of the two systems. Cumyo= cumulative percent.

RESULTS During the study period, 1 156 pairs of blood culture bottles from 504 patients were assayed. Probable clinically significant organisms were isolated from 9 patients (16%) at the Department of Hematology, and from 44 patients (1OYo) at the Department of Infectious Diseases. Propionebacterium, coagulase-negative staphylococci or diphteroids, when found in only one bottle from the same patient, were regarded as contaminants. From the assayed pairs of blood-culture bottles, probable clinically significant microorganisms were isolated in 54 pairs by both systems; 8 in the hypertonic and 2 1 in the resin-containing aerobic systems only. The sensitivity of the hypertonic and the resin-containing media was 0.75 and 0.90 respectively when compared to the "true-positive" number of bacteremia. This number was estimated as the number where significant growth could be detected in any of the systems (Table 1). Only four anaerobic strains were isolated in the study and therefore no conclusions regarding anaerobes could be made. Bacteremia of probable clinically significant TABLE I . Growth Comparison of Bactecs Aerobic Blood Culture Media No 8 and 16 Hypertonic (No 8)

Resin (No 16)

Pairs of bottles

Number of patients

growth growth non-growth non-growth

growth non-growth growth non-growth

54 8 21 1073

37 5 11 45 1

microorganisms were found in 53 patients; five were detected only in the hypertonic and 1 1 only in the resin-containing system. Sensitivity calculations gave 0.79 for the hypertonic and 0.91 for the resin-containing blood culture system among patients with significant growth of bacteria isolated from at least one bottle. The sensitivity and speed of detection of both systems was compared and is illustrated in Fig. 1. A modified chi-square test according to McNemar between resin/non-resin blood culture system showed a significantdifference(p=O.O16)in favour of the resin system. The resin system was also more rapid in detecting bacterial growth, a difference TABLE 2. A Comparison between the Present Resin Study and an Unselected Material from the County of Orebro during the same Period 1985/86 Resin study Number of patient isolates

53

Microorganisms: E. coli P. rnirabilis Enterobacteriaceae, others Ps. aeruginosa a-streptococcus Streptococcus, group D Staphylococcus aureus Staphylococcus, coagulase negative Candida sp Others

Percent 24 2 17 9 4

Unselected material 1344

13

Percent 24 3 9 3 6 5 18

19 4 0

11 2 19

8

72 1

BACTECe RESIN BLOOD CULTURE MEDIA

most clearly observed on day two (Fig. 1) when the hypertonic system had detected 4990 and the resin system 7490. The frequencies of isolated microorganisms are shown in Table 2 and are compared with an unselected material of bacteremias from 1985-1986. The frequency of patients with positive blood culture in the unselected material was 12.0% compared with 9.590 in the resin and 8.3% in the hypertonic groups.

tients treated with cytostatic drugs and/or antibiotics, the BactecB aerobic blood culture resin-containing system (No 16) is superior to the hypertonic system (No 8) in detecting bacteremia, both in regard to the number of bacteremic patients and in the rapidity of detection. These reasons indicate that the resin-containing blood culture bottle should supplant the hypertonic bottle, in patients treated with antimicrobial agents, despite the greater expense entailed.

DISCUSSION

This study was supported in part by a grant from Becton Dickinson, Stockholm, Sweden.

The difficulties of detecting bacteremia in cancer patients treated with cytostatic drugs are well documented. In patients with early presumptive septicemia, the etiological diagnosis is frequently not established (8). This problem also exists in patients who have already commenced antibiotic treatment prior to blood culturing (2, 7). Earlier studies indicate that resin-containing blood culture systems can increase the proportion of detected bacteremias in such patients (1,4, 6, 9, 10). In the present study, each blood sample was equally divided into each culture media for comparison. In addition to the higher sensitivity, the detection rate was faster in the resin-containing system. This earlier detection rate is of great importance, especially for granulocytopenic cancer patients on antitumor drugs (3). In a patient treated with an antimicrobial agent, it is more difficult to isolate the organism, and multiple cultures may be necessary (10). Among the 5 patients, where the microorganisms were isolated only in the hypertonic bottle, both resistent and sensitive isolates to received antibiotics were found. These isolates from the non-resin-containing media may simply be due to multiple culturing. Earlier studies (6, 9) suggest that the resin-containing medium is most effective in detecting Gram-positive organisms. This is not the case in the present study. There is, for example, an increase in frequencies of Enterobacteriaceae other than E. coli and of Pseudomonas aeruginosa (Table 2). The differences in the spectra of microorganisms isolated in this study compared to the unselected material from the same period (Table 2) are probably due to the criteria for patient selection and/or to the current epidemiological situation. In conclusion, this study indicates that in pa722

REFERENCES 1. Appelbaum, P. C., Beckwith, D. G., Dipersio, J. R.,

D-vke,J. W., Salventi, J. F. & Stone, L. L.: Enhanced detection of bacteremia with a new Bactec resin blood culture medium. J. Clin. Microbiol. I7:48-5 I , 1983. 2. Doern, G. V. & Gantz. N . M.: Detection of bacteremia in patients receiving antimicrobial therapy: An evaluation of the antimicrobial removal device and 16B medium. J. Clin. Microbiol. 18: 43-48, 1983. 3. EORTC international antimicrobial therapy project group: Three antibiotic regimens in the treatment of infection in febrile granulocytopenic patients with cancer. J. Infect. Dis. 137: 14-29, 1978. 4. Hopfer, R. L., Moore, D. G., Fainstein, V., Watkins, S. & Wenglar, M.: Routine use of Bactec 16B bottles to remove antibacterial and antitumor agents from blood cultures of cancer patients. J. Clin. Microbiol. 18: 759-764, 1983. 5 . Lundholm, M.: Evaluation of an antimicrobial device (ARD) for detection of septicemia. Scand. J. Infect. Dis. 18: 461-463, 1986. 6. McGuire, N. M., Kauffman. C. A , , Hertz, C. S. & Kovach, J. M.: Evaluation of the Bactec antimicrobial removal system for detection of bacteremia. J. Clin. Microbiol. 18: 449-451, 1983. 7. Peterson, L. R., Shantholtzer, C. J., Mohn, M . L. & Gerding, D. N.: Improved recovery of microorganisms from patients receiving antibiotics with the antimicrobial removal device. J. Clin. Pathol. 80: 692-696, 1983. 8. Solberg, C. 0.:Infections in the immunocompromised host. Scand. J. Inf. Dis. suppl. 124: 36-43, 1980. 9. Strand, C. L., Hickman, A , , Bryant, J. K. & Sutton, K.: Evaluation of the Bactec resin blood-culture medium. A. J. C. P. 80: 706-7 10, 1983. 10. Tilton, R. C.: The laboratory approach to the detection of bacteremia. Ann. Rev. Microbiol. 36: 467493, 1982.