252

Correspondence

Eur. J. Clin. Microbiol.

Correspondence Do Selective Media Improve the Recovery of Group B Streptococci ? Group B streptococci (GBS) are often present in the adult genito-urinary tract, and their recovery from vaginal specimens of third trimester pregnant women is of epidemiological and clinical significance. Patients are asymptomatic, and for the diagnosis one has to rely essentially upon the adequacy of techniques used in the microbiology laboratory. Because there is luxuriant growth o f normal flora, the use of selective media has been suggested in order to enhance recovery of GBS from vaginal cultures ( 1 - 3 ) . However, Gray et al. (4) showed that some o f the recommended inhibitory agents (e.g. gentamicin) may actually inapede the growth o f GBS. In order to determine whether selective media improve or interfere with the recovery of GBS from vaginal cultures, we compared the rate o f recovery o f GBS using four selective and one non-selective media. Two hundred and forty-one vaginal specimens were collected from in-patients and out-patients

Table I: Comparison of five culture media for recovery of group B streptococci from 241 vaginal specimens. Medium

Inhibitors

Recovery rate

BHIA BHIA

8 mcg/ml gentamiein 15 mcg/ml nalidixic acid 7.5 mcg/ml nalidixic acid 17 U/ml polymixin 2.1 mcg/ml neomycin 8 mcg/ml gcntamicin 15 mcg/ml nalidixic acid 7.5 mcg/ml nalidixic acid 17 U/ml polymixin B 2.1 mcg/ml neomycin

36 (15 %) 33 (13.'/ %)

BHIA BHIB BHIB

BHIA = brain-heart infusion agar BHIB = brain-heart infusion broth

32 (12.3 %) 32 (12.3 %) "35 (14.5 %)

and transported to tire laboratory in a modified Stuart transport medium (culturette Marion Health and Safety Inc.). Upon arrival the applicator was place in 1 ml of sterile 0.85 % NaC1 and twirled for 3 0 - 6 0 sec to suspend the clinical material. The media (Table 1) were then immediately inoculated. All plates were incubated aerobically at 38 ~ tbr 1 8 - 2 4 h and examined /'or GBS. All broths were incubated at 37 ~ for 1 8 - 2 4 h and cultured on 5 % sheep blood BHIA. All colonies suspected being GBS were identified and serotyped (5). Results were statistically analysed by the chi-square test, with Yate's correction ]'or small numbers. Results are shown in Table 1. The overall recovery rate o f GBS was 16.1%, recovery being defined as a positive culture on any media. The rate varied from 12.3 % to 15 % depending on the medium used. There was no significant difference between the rate on selective and non-selective media (p > 0.05). With no single. media could GBS be detected in all colonized patients. Using non-selective BHIA, GBS were detected in 92 % of colonized patients, whereas the detection rates for selective media were 8 2 % and 85 % on solid media compared to 82 % and 90 % on liquid media. Only using a combination of non-selective BHIA and selective BHIB containing gentamicin and na[idixic acid could GBS have been detected in all colonized patients. Previous reports have established that the rate of GBS colonization of the genito-urinary tract in adults varies from 2.3 % to 29.8 % (6). Baker et al. (7) showed that the rate of vaginal colonization with GBS in non-pregnant college women was 18 %. Our results are consistent with those findings. However, contrary to the findings of other investigators ( 1 , 2 ) our rate of isolation was hardly altered by the use of selective media as compared to non-selective. One possible explanation for this is our choice of non-selective medium. We used BHIA as blood-agar base because clearly defined GBS beta-hemolytic zones were consistently obtained on it. Thus, GBS was always easily detected despite luxuriant growth o f normal flora on non-selective BHIA. Selective media cannot be used as a substitute for non-selective well-prepared streak plates for determining hemolysis. The superiority of non-selective BHIA combined with BHIB containing gentamicin and nalidixic acid

Correspondence

is questionable since high isolation rates were obtained with every media. Our findings show that selection o f the appropriate blood-agar base for the isolation of GBS from vaginal cultures can eliminate the necessity of using selective enrichment or selective primary media. M. N. A. M. M.

Laverdiere* Sayegh Robert Lefebvre Domaradzki

Laboratoire de microbiologie, H6pital MaisonneuveRosemont, 5415 Boulevard L'Assomption, Montreal, Quebec, Canada, HIT 2M4.

References

1. Baker, C. J., Goroff, D. K., Alpert, S. L., Hayes, C., McCormack, W. M.: Comparison of bacteriological methods for the isolation of group B streptococcus from vaginal cultures. Journal of Clinical Microbiology 1976, 4:46 48. 2. Mason, E. O., Wong, P., Barrett, F. F.: Evaluation of four methods for detection of group B streptococcal colonization. Journal of Clinical Microbiology 1976, 4: 429-431. 3. Ferrieri, P., Blair, L.L.: Pharyngeal carriage of group B streptococci: detection by three methods. Journal of Clinical Microbiology 1977, 6: 136139. 4. Gray, B. M., Pass, M. A., Dillon, H.C.: Laboratory and field evaluation of selective media for isolation of group B streptococci. Journal of Clinical Microbiology 1979, 9: 466-470. 5. Swift, H. F., Wilson, A. T., Laneefield, R. C.: Typing group A hemolytic streptococci by M. precipitin reactions in capillary pipettes. Journal of Experimental Medicine 1943, 78:127 133. 6. Patterson, M.J., Hafeez, A. B.: Group B streptococci in human disease. Bacteriological Review 1976, 40:774 792. 7. Baker, C.J., Goroff, D.K., Alpert, S., Evrard, J.R., Rosmer, B., McCormaek, W.M.: Vaginal colonization with group B streptococcus: a study in college women. Journal of Infectious Diseases 1977, 135: 392-397.

Vol. 1, No. 4

253

Klebsieila pneumoniae Colonization of the Bowel Associated with the Use of Trimethoprim-Sulfamethoxazole Sir, The intestinal tract has been a major reservoir o f Klebsiella pneumoniae, although the duration o f colonization o f the bowel has rarely been studied ( 1 , 2 , 3). We studied the duration o f intestinal colonization by Klebsiella pneumoniae in patients with spinal cord injury during an outbreak o f bacteriuria due to Klebsiella pneumoniae resistant to gnetamicin and tobramycin (GT r Klebsiella pneumoniae). Twelve male patients with spinal cord injury and urinary tract infection with GT r Klebsiella pneumoniae were studied. Anal swabs were obtained weekly until discharge from hospital. Eight o f the 12 patients were voiding by reflex and ten patients were using external urinary collection devices at the time o f bacteriuria. The patients had been in hospital 22 to 155 days (mean 64) before they developed urinary tract infection with GT r Klebsiella pneumoniae and were observed for 16 to 99 days (mean 84 days) until their discharge from hospital. Klebsiella pneumoniae was isolated from the rectal swabs of all 12 patients at some time during the study. Fifty-four isolates o f Klebsiella pneumoniae were obtained from 99 anal swab cultures over the observation period; 38 o f the isolates were resistant to gentamicin and tobramycin. At the time o f the initial urinary tract infection, the duration of documented anal colonization (continuously positive cultures) with GT r Klebsiella pneurnoniae after the positive urine culture was up to 30 days in six of the 12 patients. Three patients developed four later episodes o f bowel colonization with GT r Klebsiella pneumoniae !asting 17 to 35 days. The results were similar to previous observations o f patients in hospital and other groups o f patients receiving antibiotics (4) and indicate that patients with spinal cord injury do not seem to be at particular risk o f prolonged colonization o f the bowel with GT r Klebsiella pneumoniae. Some patients had previously had bacteriuria which h a d been treated with antibiotics a n d many patients were maintained on trimethoprimsulfamethoxazole or nitrofurantoin as pro-