Annie Bremmelgaard

Few microorganisms associated with bacterial vaginosis may constitute the pathologic core: a population-based microbiologic study among 3596 pregnant women

OBJECTIVE To evaluate the association between various microorganisms isolated from the genital tract in pregnant women with bacterial vaginosis. STUDY DESIGN A cross-sectional population-based study among pregnant women addressed at their first antenatal visit before 24 full gestational weeks from the referring area of the Department of Obstetrics and Gynecology at Odense University Hospital, Denmark, from November 1992 to February 1994. The main outcome measures were prevalence of various microorganisms and statistical estimates of interactions (crude, adjusted, and relative odds ratios) between the microorganisms isolated from the lower genital tract in pregnant women with and without clinical diagnosis of bacterial vaginosis. RESULTS Three thousand five hundred ninety-six (3596) pregnant women were asked to participate. Of the 3596 pregnant women 3174 (88.4%) agreed to participate before 24 full gestational weeks. After controlling for the presence of other microorganisms, strong associations between Gardnerella vaginalis, anaerobic bacteria, Mycoplasma hominis, and present bacterial vaginosis were found. Similarly Lactobacillus spp. were found to be associated with the absence of bacterial vaginosis. The combination of G. vaginalis and anaerobic bacteria and/or M. hominis was found in 59.6% of the cases with bacterial vaginosis and in 3.9% of the cases without bacterial vaginosis (odds ratio 36.4, 95% confidence interval 27.8 to 47.8). The crude odds ratio was found to be as high as 74.8 (95% confidence interval 32.3 to 174.1) when the combination of G. vaginalis, M. hominis, anaerobic bacteria, and no Lactobacillus spp. was associated with bacterial vaginosis. CONCLUSION There is a microbial foundation for bacterial vaginosis, and it is possibly due to an intermicrobial interaction in which the microorganisms G. vaginalis, anaerobic bacteria, and M. hominis are dominating, indicating that these constitute the pathologic core of bacterial vaginosis.

Determination of immunoglobulin A against Gardnerella vaginalis hemolysin, sialidase, and prolidase activities in vaginal fluid: implications for adverse pregnancy outcomes.

ABSTRACT A nested case-control study of low birth weight and preterm delivery was performed with singleton women. Immunoglobulin A (IgA) against the Gardnerella vaginalis hemolysin (anti-Gvh IgA) and sialidase and prolidase activities were determined in vaginal fluid at 17 weeks of gestation. Sialidase positivity and bacterial vaginosis with high prolidase activity were associated with 2- and 11-fold increased risks for low birth weight, respectively. No woman with bacterial vaginosis plus a strong anti-Gvh IgA response had an adverse outcome.

Susceptibility testing of Danish isolates of Capnocytophaga and CDC group DF-2 bacteria

Twelve Capnocytophaga and seven DF‐2 strains were tested for their susceptibility to 14 antimicrobial agents using an agar dilution and an agar diffusion method. Twenty‐three other antibiotics were evaluated using the diffusion test only. All strains were fully susceptible to penicillin, ampicillin, cefuroxime, cefotaxime, erythromycin, clindamycin, chloramphenicol, doxycycline, rifamycin and ofloxacin using both methods. Clindamycin, rifamycin and cefotaxime were most active. Using agar dilution some strains were susceptible to gentamicin, but agar diffusion showed total resistance. One Capnocytophaga strain was susceptible and another moderately susceptible to metronidazole, other strains were resistant. The agar diffusion test showed that both Capnocytophaga and DF‐2 were resistant to most other aminoglycosides, to fosfomycin, polymyxin and trimethoprim. All strains of both taxa were fully susceptible to piperacillin, cefoxitin, imipenem and fusidic acid and showed different susceptibilities to the other agents. Susceptibility testing by means of agar diffusion using an enriched chocolate agar and 5% CO2 atmosphere could be used to test Capnocytophaga and DF‐2 strains and gives sufficient accuracy for routine use, when revised inhibition zone breakpoints are employed.

A Novel Screening Method for the Detection of Microbial Contamination of Platelet Concentrates

Platelet concentrates, stored at room temperature, are nowadays the commonest cause of posttransfusion septicemia [l, 21. Reported contamination rates of platelet concentrates varies from 0.9 to about 6% [l-51. The transfer of microorganisms from the donors skin or blood to the sampling unit is the greatest risk [5 , 61. When the storage temperature of platelets was changed from 4°C to room temperature, several investigators became concerned about the possibility that bacteria might grow under these new storage conditions [3]. Screening of platelet concentrates by conventional culture methods is laborintensive and takes time. We performed a small experimental pilot study to elucidate the efficacy of a newly developed technology based on noninvasive detection of microbially produced C 0 2 in platelet bags (Holtech Medical Aps, Charlottenlund, Denmark). An adhesive label containing a pC02-sensitive color indicator was placed on the surface of the flexible bag (PL-1240@, Baxter Fenwal, La Chatre, France). As this platelet bag is semipermeable, the pCOz of the medium (human donor plasma) equilibrateswith thepC02in theindicator fluid. At normal pCOz ievels, the indicator fluid is blue, changing to green and yellow at elevated pCOz levels generated by microbial Fig. 1. Changes in colony count, pH and pCOz during incubation with E . coli.

Acquisition and Elimination of Bacterial Vaginosis During Pregnancy: A Danish Population-Based Study

Objectives: the aim was to examine factors associated with acquisition and elimination of bacterial vaginosis in pregnancy. Methods: a group of 229 pregnant women were randomly selected from a population-based prospective cohort study of 2927. They were examined at enrollment (mean gestational weeks 16w + 0d) and again in mid-third trimester (mean gestational age 32w + 3d). Measures: BV (Amsels clinical criteria), microbiological cultures of the genital tract and questionnaire data. Results: BV prevalence decreased from 17% in early second trimester to 14% in mid-third trimester due to a tenfold higher elimination rate (39%) than incidence rate (4%). Heavy smokers (> 10/d) in early pregnancy were at increased risk (5.3 [1.1–25]) for the acquisition of BV during pregnancy, as were women receiving public benefits (4.8 [1.0–22]), having a vaginal pH above 4.5 (6.3 [1.4–29]) or vaginal anaerobe bacteria (18 [2.7–122]) at enrollment. A previous use of combined oral contraceptives was preventive for the acquisition of BV (0.2 [0.03–0.96]). Elimination of BV in pregnancy tended to be associated with a heavy growth of Lactobacillus (3.2 [0.8–13]) at enrollment. Conclusions: acquisition of BV during pregnancy is rare and is associated with smoking, while the presence of anaerobe bacteria and a vaginal pH > 4.5 are interpreted as steps on a gradual change towards BV. In the same way heavy growth of Lactobacillus spp in early pregnancy may be an indicator of women on the way to eliminate BV.

Quality of the surveillance of surgical wound infections: A 10-year prospective study of 12,364 wounds

We did a survey of postoperative wound infection rates in our department over a 10-year period. It was based on the results of forms completed by surgeons in the department after each operation and when an infection was diagnosed (primary registration). These forms were sent to the electronic data processing department for registration of data in a database. Every third month, detailed reports of the data were discussed with the surgeons. The quality of this routine survey was evaluated as follows: 1) Every year, the years data in the surveys database were compared with those in the National Patient Register and, in case of discrepancies, the patients records were checked and missing information about the operation and infections were recorded in the surveys database (secondary registration). 2) In case of a primary registered infection, the patients record was carefully examined. Infections not fulfilling the criteria for an infection were called registration failures. The primary registration of operations reached an acceptable level (about 95%) after 5 years of the survey, but the recording of infections was unreliable throughout the study due to an unacceptably low registration rate (about 60%) and to a high rate of registration failures.

Orthopedic wound infections: 182 cases after 8913 operations during an 8-year survey

We performed a surveillance of postoperative wound infections at our department for a period of 8 years. The surveillance was based on forms completed by the surgeons at the department. Retrospectively, by scrutinizing the records of patients with recorded infections, the criteria for infections used by the surgeons were compared with the criteria recommended by the Centers for Disease Control, USA, for routine infection surveillance studies: 1) pus present in the wound, 2) bacteria isolated from the wound, 3) spontaneous wound breakdown or surgical intervention due to suspected infection; if culture is performed, it must be positive, 4) evidence of infection at reoperation or at radiologic or histopathologic examination (restricted to deep infections) or 5) diagnosis of infection by a surgeon. 182 infections were recorded; in 106 (58%) pus was present, drained either spontaneously or by surgical intervention, in 24 (13%) no pus was seen, but the culture was positive. In 116 cases, surgical intervention was performed or wounds spontaneously broke down; pus was found in 101 of these and culture of non-purulent material was positive in 13 and negative in 2. In 52 (29%) patients, only criterion 5 was fulfilled. In surveillance studies where the recordings are based on several observers, the subjective criterion 5 may, as in our study, give a falsely high number of infections, also preventing comparison between studies. We recommend that all suspected infections are cultured and that only criteria 1, 2 and 4 are used.

Evaluation of the E-test for susceptibility testing of the Bacteroides fragilis group

The susceptibilities of 100 clinical isolates belonging to the Bacteroides fragilis group to 9 antibiotics, i.e. ampicillin, piperacillin, ceftriaxone, cefotaxime, cefoxitin, imipenem, erythromycin, clindamycin, and metronidazole, were tested using a standard agar dilution method and the E‐test. Overall, 81% of the E‐test MICs were within one log2 dilution step of the agar dilution MICs and 95% were within two log2 dilution steps. The E‐test showed significantly lower MIC values than the agar dilution method for ampicillin, cefotaxime, and imipenem, and significantly higher MIC values for piperacillin, ceftriaxone, erythromycin, clindamycin, and metronidazole. However, this only resulted in minor discrepancies that did not change the susceptibility status. The E‐test is easy to perform and read, and the MIC values correlated well with the MICs obtained by the agar dilution method when testing susceptibility of the Bacteroides fragilis group.

Comparison of detection speed and yield in BACTEC NR‐860® and Roche BCB® blood culture systems

Detection speed and yield were compared between BACTEC NR‐860® and Roche BCB® blood culture systems. From a total of 1,550 paired blood cultures inoculated with the same volume of blood, 161 (10.4%) grew 180 organisms from 140 adult patients. Sixty‐six percent of the isolates were clinically significant. The BACTEC® system detected more clinically significant isolates than the BCB® system, 91% vs 76% (p < 0.01), especially Enterobacteriaceae. In 39 paired blood cultures, 42 significant organisms were detected at different times in the two systems. The vast majority (93%) were detected first (on average 1 day faster) by the BACTEC® system (p < 0.05), especially Escherichia coli and streptococci. Generally, detection times were shorter in the BACTEC® system. Three fourths (76%) of all significant organisms were detected within the first incubation day as compared to less than half (42%) in the BCB® system (p < 0.001). Generally, the agitated aerobic resin medium BACTEC 26 Plus® had the fastest detection and highest yield in our study. The contamination rate, below 2% in both systems, was acceptable. The BACTEC NR‐860® system with resin‐containing media proved to be a reliable, sensitive and fast blood culture system, which deserves further investigation.

Susceptibility testing of anaerobic bacteria

The relationship between susceptibility testing by an agar dilution test and a tablet diffusion test was studied for 60 anaerobic bacteria (20 B. fragilis, 20 anaerobic cocci, 20 Clostridium species). For cefoxitin, no prediffusion and prediffusion times of one h, three h, 12 h, 24 h and 48 h were examined. For metronidazole, erythromycin, clindamycin, penicillin and imipenem, only 24 h prediffusion and no prediffusion were studied. Measurements were made after incubation for 24 h and 48 h. Prediffusion improved the correlation for all antibiotics tested, and 24 h prediffusion gave the best results. The slope of the regression line increased and the influence of the individual growth parameters on zone size was reduced. Prediction of susceptibility based on three zone size breakpoints to estimate MIC was also better with 24 h prediffusion. However, the variation about the regression line for many of the antibiotics was still extremely high. Measurements after 24 h and 48 h incubation times showed almost identical regression equations, except for erythromycin, where the regression lines differed.