Determination of the optimum incubation period of continuously monitored blood cultures from patients with suspected endocarditis or fungaemia.

AIM: It is routine practice to prolong the incubation period of blood cultures from patients with suspected subacute bacterial endocarditis (SBE) or fungal infection. The protocol in this laboratory required 28 days incubation with weekly subcultures. Following the introduction of automated continuously monitored blood culture instruments, the duration of incubation for specimens from these categories of patients was reviewed. METHOD: In a retrospective study of blood culture specimens submitted from 1 July 1994 to 31 July 1998, the time from collection to a positive signal from the BacT/Alert automated blood culture system, in patients suspected of SBE or fungal infection, was assessed. RESULTS: From 355 patient episodes, 896 bottles were incubated for up to 28 days, during which time 116 bottles (40 patient episodes) signalled positive. Significant organisms from suspected endocarditis patients were isolated from 87 bottles (87%), with the time to detection ranging from 9 to 96 hours. The data collected from significant isolates obtained from clinically suspected fungaemias were extremely small, prompting a review of the total fungal isolates from all blood culture bottles incubated during the study period. Yeast isolates were obtained from 78 (0.08%) bottles with the detection time ranging from 15 to 144 hours. CONCLUSION: The practice in this laboratory now is to incubate blood cultures for up to 7 days when the clinical notes indicate the possibility of SBE or a fungal infection.

Group B streptococcus screening in pregnant women: a comparison of three media.

This study compared the relative isolation rate of Group B Streptococcus (GBS) from 663 low vaginal swabs, collected from antenatal patients, on three media: horse blood agar plus neomycin (75 mg/l) (Neo), Islam agar (Islam) and Islam agar plus nalidixic acid (15 mg/l) and colistin sulphate (10 mg/l) (Islam Plus). One hundred and forty-seven (22%) GBS were isolated. At 24 hours the isolation rate was highest using Neo, but within 72 hours there was little difference. The isolation rates for Neo, Islam and Islam Plus at 24 hours were 124 (18.7%), 103 (15.6%), 109 (16.4%) (P < 0.05); at 48 hours 125 (18.9%), 116 (17.5%), 121 (18.1%) (P > 0.1); and at 72 hours 125 (18.9%), 121 (18.3%) and 127 (19.1%) (P > 0.1), respectively. Twenty-two isolates were missed on Neo, 26 on Islam and 20 on Islam Plus. Of those missed on Islam agars, 12 failed to produce pigment and were only detected on Neo. The disadvantage of Neo is the need to perform additional tests to confirm the identity as GBS. In the present study 100 suspicious colonies were identified as Group D.

Effect on organism recovery rate from BacT/Alert blood cultures with reduced incubation period.

This retrospective study evaluated 15,377 sets of BacT/Alert blood cultures to determine incubation time for blood cultures. Ninety-six per cent (1476) of total isolates signalled positive within five days and 56 isolates turned positive in five to seven days. Of the 56 organisms recovered between five and seven days, 49 were considered contaminants and seven were considered clinically significant. On assessing the medical records of the patients with the seven clinically significant isolates, it was determined that the clinical outcome would not have changed if these isolates were missed. We conclude that a five day incubation protocol reduces the recovery of skin contaminants while not significantly decreasing the recovery of clinically significant organisms. The data suggest that the incubation time can be further reduced but this policy will depend on the individual institution and their patient population mix.

A comparative study of the BBL crystal enteric/nonfermenter identification system and the biomerieux API20E and API20NE identification systems after overnight incubation.

The BBL Crystal Enteric/Nonfermenter (Crystal, Becton Dickinson Microbiology Systems) is a new multi-test identification system for Gram negative rods requiring no oil overlay or addition of reagents. One hundred and three selected Gram negative rod isolates from routine clinical specimens were tested in parallel using the appropriate API20E or API20NE (BioMerieux) reading after overnight incubation. The isolates included in the study, and the number tested, were as follows: Acinetobacter sp, 8; Aeromonas hydrophila, 1; Citrobacter diversus, 1; Citrobacter freundii, 1; Escherichia coli, 13; Enterobacter aerogenes, 10; Enterobacter agglomerans, 2; Enterobacter cloacae, 6; Klebsiella oxytoca, 6; Klebsiella pneumoniae, 12; Morganella morganii, 2; Proteus mirabilis, 6; Pseudomonas aeruginosa, 21; Salmonella sp, 2; Salmonella typhi, 1; Serratia marcescens, 1; Shigella sonnei, 2; Shigella sp, 2; Vibrio metschnikovii, 1; Xanthomonas maltophilia, 3; and, Yersinia enterocolitica, 2. The API20E and API20NE systems combined identified 74.8% (77/103) and the Crystal 97.1% (100/103) of isolates. Twenty six isolates required repeat and/or additional tests for correct identification with the API systems compared to 3 with the Crystal. Crystal is more convenient than API for routine clinical use because it requires fewer repeat and/or extra tests, and is easier and less time consuming to use.