Variability in diagnosis of clue cells, lactobacillary grading and white blood cells in vaginal wet smears with conventional bright light and phase contrast microscopy.

OBJECTIVE: Study the reproducibility of wet smear interpretation of clue cells, lactobacillary grades and leukocyte dominance with conventional bright light and phase contrast microscopy. STUDY DESIGN: Sets of vaginal specimens were taken from unselected consecutive women attending an outpatient gynaecology clinic. Air-dried vaginal fluid on a microscope slide was rehydrated with isotonic saline before examination by six independent international investigators. Some investigators initially used a conventional bright light microscope, followed by phase contrast technique. RESULTS: Using phase contrast microscopy, an excellent inter-observer agreement was obtained among all investigators for clue cells detection (Kappa values from 0.69 to 0.94) and lactobacillary grades (Kappa 0.73-0.93). When conventional light microscopes were used, poor agreement was obtained for these criteria (Kappa index 0.37-0.72 and 0.80, respectively), but switching to phase contrast microscopy by the same investigators, improved Kappa to 0.83-0.85 and 0.88, respectively. The inter-observer agreement for estimation of the leukocyte/epithelial cell ratio (Kappa index 0.17-0.67) was poor, irrespective of the type of microscopy applied. Intra-observer agreement of clue cell detection and lactobacillary grading was also found to be excellent if phase contrast microscopy was used (Kappa 0.87-0.93), and poor with conventional bright light microscopy (Kappa 0.45-0.66). CONCLUSION: Clue cells and the lactobacillary grades are reliably identified by phase contrast microscopy in wet smears, with excellent intra- and inter-observer reproducibility agreement, and better than when simple bright light microscopy was used. Evaluation of leukocyte grading, on the other hand, was inconsistent among the different microscopists, irrespective of the type of microscope used. We propose to grade the leukocytes in a different way than searching for leukocyte dominance over epithelial cells, namely by counting them per high power field and per epithelial cell.

Validation of the use of Pap-stained vaginal smears for diagnosis of bacterial vaginosis.

Papanicolaou-stained cervicovaginal smears (Pap smears) are used to screen for cervical cancer. Since there is a lack of consensus in published reports respecting the efficacy of Pap-stained smears in BV diagnostics, there is a need to validate their use for diagnosis of BV. Slides from the international BV00 workshop were Pap stained and independently analyzed by four investigators under a phase-contrast microscope. All workshop slides--whether Pap-stained, Gram-stained or rehydrated air-dried smears--were scored according to the same Nugent classification. The diagnostic accuracy of Pap smears for diagnosis of BV had a sensitivity of 0.85 and a specificity of 0.92, with a positive and negative predictive value of 0.84 and 0.93, respectively. The interobserver weighted kappa index was 0.86 for Pap-stained smears compared to 0.81 for Gram-stained smears, and 0.70 for rehydrated air-dried smears using the mean Nugent score as the criterion standard. Provided that the samples are taken from equivalent locations (the vaginal fornix) and analyzed according to the same scoring criteria, there is no discernable difference in the diagnostic accuracy of the three smear-staining methods. The Pap-stained vaginal smears can be used as a wholly adequate alternative to Gram-stained smears for BV diagnosis.

A randomised double blind trial comparing misoprostol or placebo in the management of early miscarriage.

OBJECTIVES: To study if misoprostol 400 microg, administered vaginally, increased the successful resolution of early miscarriage compared with placebo. DESIGN: Randomised, double blind placebo controlled study. SETTING: Sahlgrenska University Hospital, Göteborg, Sweden. SAMPLE: One hundred and twenty-six women seeking medical attention for early miscarriage. METHOD: Women with a non-viable, first trimester miscarriage were randomised to vaginal administration of misoprostol 400 microg or placebo. MAIN OUTCOME MEASURES: Main outcome measure was the proportion of successful complete resolution of miscarriage. Secondary outcomes were incidence of infection, bleeding, gastrointestinal side effects, pain, use of analgesics and length of sick leave between groups. RESULTS: Sixty-four patients were randomised to misoprostol and 62 to placebo. Eighty-one percent in the misoprostol and 52% in the placebo group had a complete miscarriage within one week of the primary visit (RR 1.57; 95% CI 1.20-2.06). Patients in the misoprostol group reported more pain as assessed on a visual analogue scale (60.4 [31.0] vs 43.8 [37.1] mm; P < 0.007) and required analgesics more often (83%vs 61%, RR 1.35; 95% CI 1.08-1.70). There were no significant differences in the occurrence of gastrointestinal side effects, infection, reduction in haemoglobin or sick leave between the groups. CONCLUSIONS: Treatment with 400 mug misoprostol administered vaginally increased the success rate of resolvement of uncomplicated early miscarriages compared with placebo. However, women who received misoprostol experienced more pain and required more analgesics than those who did not.

An international study of the interobserver variation between interpretations of vaginal smear criteria of bacterial vaginosis.

An international workshop on vaginal smear-based diagnosis of bacterial vaginosis was organized where 13 investigators scoring 258 slides with smears from vaginal fluid. Interobserver reproducibility of interpretations of Nugent scores, Hay/Ison scores and wet smear scores for the diagnosis of bacterial vaginosis was shown to be high. Detailed analysis of individual scoring results however indicated that basic standards of quality control to ensure robust individual readings of slides must be adhered to.

Treatment with 2% clindamycin vaginal cream prior to first trimester surgical abortion to reduce signs of postoperative infection: a prospective, double-blinded, placebo-controlled, multicenter study.

BACKGROUND: Bacterial vaginosis (BV) and intermediate flora is known risk-factor for postoperative infection after surgical termination of pregnancy. Vaginal application of 2% clindamycin cream is an efficacious treatment for BV, but it is not known whether preoperative administration of clindamycin cream might reduce the signs of post-abortion infection after surgical termination of pregnancy. AIM: To evaluate whether preoperative treatment with clindamycin cream might reduce the signs of post-abortion infection after legal abortion. DESIGN: Prospective, double-blinded, placebo-controlled, multicenter study. MATERIAL AND METHODS: Consecutive women attending for surgical termination prior to 11+4 gestational weeks were approached. We randomized participants to preoperative vaginal treatment with 2% clindamycin cream or placebo cream in a double-blinded fashion. At all visits vaginal smears were air dried on microscopy slides to be stored. The rate of postoperative pelvic infection according to our definition was the main outcome variable, the cure rates of BV and of intermediate flora were secondary outcome variables. RESULTS: Of 1655 enrolled women, 1102 were evaluable for analyses. Fifty-eight women developed signs of post-abortion infection. Preoperative treatment with clindamycin cream significantly (RR: 4.2, 95% C.I. 1.2-15.9) reduced the risk of post-abortion infection among women with abnormal vaginal flora (BV and intermediate flora). Treatment with clindamycin cream in women with normal lactobacilli flora did not demonstrate any difference compared to the non-treatment group. CONCLUSION: Preoperative treatment for at least three days with clindamycin cream significantly reduced the risk for developing signs of post-abortion infection only among women with preoperative abnormal vaginal flora (BV and intermediate flora).

Randomised trial comparing expectant with medical management for first trimester miscarriages.

OBJECTIVE: To compare the efficacy of antiprogesterone (mifepristone) in combination with a synthetic prostaglandin E1 analogue (misoprostol) for outpatient treatment of miscarriages. PARTICIPANTS: One hundred and twenty-two women with first trimester miscarriages. METHODS: The women were randomised to treatment with mifepristone 400 mg orally followed by a single oral dose of 400 microg misoprostol 48 hours later (n = 60) or expectant management (n = 62). Women were re-evaluated five days later. If retained intrauterine products of conception were found with an antero-posterior diameter above 15 mm on transvaginal ultrasound, surgical evacuation was performed. RESULTS: Eighty-two percent of the women randomised to pharmacological treatment and 76% of those randomised to expectant management had an empty uterine cavity after five days. Convalescence time was 1.8 days longer for women randomised to pharmacological treatment. Pain, bleeding, complications, and satisfaction with the treatment did not differ between the groups. CONCLUSIONS: Most cases of spontaneous incomplete miscarriage will become a complete miscarriage without intervention. This study shows that outpatient treatment with a combination of antiprogesterone and a prostaglandin E1 analogue did not increase the rate of complete miscarriage, compared with expectancy alone, by a clinical important degree.

IL-1beta, IL-6, TNFalpha, fetal fibronectin, and endotoxin in the lower genital tract of pregnant women with bacterial vaginosis.

BACKGROUND: In our studies on women with bacterial vaginosis (BV) in early pregnancy a strong association has been found between BV and the levels of endotoxin or interleukin-1alpha (IL-1alpha) in the lower genital tract. In the present study we investigated if an association could be found between BV and other cytokines (IL-1beta, IL-6, tumor necrosis factor alpha, TNF) or fetal fibronectin (FFN). The cytokine-inducing capacity of endotoxins present in the cervical mucus was explored in a monocytic cell assay. METHODS: Cervical mucus or cervicovaginal fluid was collected from women with (BV) and without BV (nonBV) attending a family planning unit for first trimester abortion. The concentrations of IL-1beta, IL-6, TNF and FFN were determined by quantitative enzyme immunoassays. TNF was determined in 63 women (BV, n=25) out of whom 37 (BV, n=11) were analyzed for IL-1beta and the remaining 26 for IL-6 (BV, n=14). FFN was determined in another 36 women (BV, n= 19). The cytokine-inducing capacity of endotoxin-containing cervical mucus and purified endotoxin of Prevotella bivia were studied by an in vitro cell assay using a human monocytic cell line (THP-1). RESULTS: IL-lbeta and IL-6 were found in almost all women. The levels of IL-1beta, but not IL-6, TNF or FFN, were significantly increased in women with BV compared with the nonBV women (p<0.05). Purified endotoxin from P. bivia, and cervical mucus from BV women containing high levels of endotoxin were able to induce a cytokine response (IL-6) in monocytic cells in vitro. CONCLUSION: BV is associated with increased levels of IL-1beta in the lower genital tract of pregnant women in the first trimester. The ability of BV-associated endotoxins to induce cytokine production in monocytic cells may partly explain the increased IL-1beta levels.

Interleukin-1alpha, interleukin-6 and interleukin-8 in cervico/vaginal secretion for screening of preterm birth in twin gestation.

BACKGROUND: The purpose was to determine the prognostic value of interleukin (IL) 1-alpha, IL-6 and IL-8 in cervico/vaginal secretion for preterm birth (<37 weeks of gestation) in twin pregnancies. METHODS: The study included screening of 121 women with twin pregnancies with sampling at 24, 26, 28, 30, 32 and 34 weeks of gestation. IL-1alpha, IL-6 and IL-8 was analyzed with ELISA immunoassays. The detection limit was 30 pg/mL for IL-1 and IL-8 and 40 pg/mL for IL-6. Vaginal fluid was smeared and dried for later evaluation of bacterial vaginosis (presence of clue cells). RESULTS: Spontaneous preterm birth occurred in 36 women and 65 women were delivered at term. IL-8 was significantly higher (p=0.03) in samples from women delivered preterm (median 3.72 ng/g mucus, range <0.07-220.00) compared with samples from women delivered at term (median 3.03 ng/g mucus, range <0.08-378.60). At 28 weeks of gestation, IL-8 (cut off 1.75 ng/g mucus) was associated with preterm delivery (relative risk 2.2, CI 95% 1.1-4.5) with a sensitivity, specificity, positive and negative predictive value of 78.8, 45.8, 44.8 and 79.4%, respectively. The levels of IL-1alpha and IL-6 were not significantly associated with preterm birth. Bacterial vaginosis was found in 47/541 (8.7%) samples analyzed. The levels of IL-1alpha and IL-8 were significantly higher in samples positive for bacterial vaginosis than in negative samples (p<0.0001 and p<0.01, respectively). There was no significant association between the level of IL-6 and bacterial vaginosis. CONCLUSIONS: IL-8, but not IL-1alpha and IL-6, was associated with preterm delivery but the relationship was too weak to be of predictive value for preterm birth in twin pregnancies. IL-1alpha and IL-8, but not IL-6, were associated with bacterial vaginosis.

Unsuccessful treatment of missed abortion with a combination of an antiprogesterone and a prostaglandin E1 analogue.

In a prospective clinical trial the effectiveness of a combination of 400 mg of mifepristone (antiprogesterone) and 400 g misoprostol (synthetic prostaglandin E1 analogue), both taken orally, was evaluated for the treatment of missed abortion. Of the 31 patients included, 16 (52%) had an empty uterine cavity at follow up six days after inclusion, 11 (35%) required surgical evacuation for retained intrauterine products of conception found at follow up, and four (13%) required emergency surgical evacuation due to severe pain or bleeding. The results do not support the use of mifepristone and misoprostol for women wishing the miscarriage to be resolved quickly.

PIP: A prospective clinical trial involving 31 Swedish women failed to demonstrate the effectiveness of a combination of 400 mg of mifepristone and 400 g of misoprostol for the treatment of missed spontaneous abortion. Transvaginal ultrasound examination showed intrauterine products of conception with an antero-posterior diameter of 15-50 mm in all study participants. 6 days after treatment, only 16 women (52%) had an empty uterine cavity. Surgical evacuation was required in the remaining 15 women, 11 because of retained products of conception (mean diameter, 26.7 mm). An additional 4 women underwent emergency surgical evacuation due to severe pain or bleeding. Women who successfully aborted after treatment had a significantly greater number of prior pregnancies than women who failed to abort. There were no differences between groups, however, in gestational length, size of the pregnancy products, or serum levels of human chorionic gonadotropin and progesterone. These results do not support the use of mifepristone and misoprostol for pharmacologic treatment in women who want their spontaneous abortions to be resolved quickly.

Prostaglandin E and F2 alpha concentration in the cervical mucus and mechanism of cervical ripening.

OBJECTIVE: To study the mechanism of cervical ripening by determination of prostaglandin E (PGE) and F2 alpha (PGF2 alpha) concentrations in cervical mucus during the course of pregnancy. STUDY DESIGN: Cervical mucus was collected from 99 pregnant women attending the mother care unit of the department. Women with sexual intercourse within the last 24 hours before sampling and subjects with bacterial vaginosis were analysed separately. RESULTS: Eleven women had sexual intercourse within 24 hours before sampling. The concentration of PGE in their cervical mucus was high corresponding to 2000-4000 pg/mg w w lasting for a period of 10-12 hours postcoitally, whereas the levels of PGF 2 alpha only increased slightly. Bacterial vaginosis was accomplished by a slight but significant elevation of PGF2 alpha levels but only of a minor increment of the PGE values. The prostaglandin concentrations in the mucus from the remaining 68 women were for PGE 102.75 +/- 111.51 and for PGF2 alpha 97.54 +/- 82.48 pg/mg w w (mean +/- SD). Although the values were scattered the concentrations remained at approximately the same level throughout pregnancy and there was no tendency towards an increment during the last weeks of pregnancy when cervical maturation is apparent. CONCLUSION: Cervical softening seems not to be accomplished by a massive local release of prostaglandins but rather the result of a number of different mechanisms more or less influenced by minor alterations of prostaglandin synthesis and release. Involved in these mechanisms are probably neutrophil-derived collagenases.

Fetal fibronectin, endotoxin, bacterial vaginosis and cervical length as predictors of preterm birth and neonatal morbidity in twin pregnancies.

OBJECTIVE: To evaluate the predictive values of fetal fibronectin, bacterial vaginosis, endotoxin and cervical length for preterm birth (< 35 and < 37 weeks) and neonatal morbidity in twin pregnancies. PARTICIPANTS: One-hundred and twenty-one women with twin pregnancies recruited into a prospective longitudinal study at three antenatal clinics in the southwest of Sweden. METHODS: Cervical or vaginal fluid was sampled and determined for fetal fibronectin (> or = 0.05 microgram/mL was used as cutoff), endotoxin (> or = 100 pg/mL) and bacterial vaginosis (presence of clue cells) at two week intervals from 24 to 34 weeks of gestation. The cervical length was measured with transvaginal sonography at the same time intervals. MAIN OUTCOME MEASURES: Occurrence of preterm birth (< 35 and < 37 weeks of gestation) and neonatal morbidity. RESULTS: All positive fetal fibronectin samples obtained at screening between 24 and 34 weeks predicted birth < 35 weeks (RR 18.0; 95% CI 2.2-145.9). A positive fetal fibronectin at 28 weeks of gestation predicted delivery < 35 weeks (RR 6.3; 95% CI 2.6-15.1) with a sensitivity, specificity, positive and negative predictive value of 50.0, 92.0, 62.5 and 87.3%, respectively. An independent association between fetal fibronectin at 28 weeks and preterm birth (< 35 weeks) was verified with logistic regression (P = 0.03). A positive fetal fibronectin at 28 weeks of gestation predicted neonatal morbidity (RR 5.1; 95% CI 2.4-11.0) and a longer period of care at the neonatal intensive care unit. The predictive power of cervical sonography was generally low but cervical length (cutoff < or = 33 mm) measured at 28 weeks of gestation was significantly associated with birth < 37 weeks (RR 2.2; 95% CI 1.1-4.2). The presence of endotoxin correlated to bacterial vaginosis, but these tests were not significantly related to preterm birth or neonatal morbidity. CONCLUSIONS: Fetal fibronectin predicted preterm birth and neonatal morbidity in twin pregnancies. The predictive value of cervical length determinations was low. Endotoxin and bacterial vaginosis had no predictive power for preterm delivery in this study.

Infections after hysterectomy. A prospective nation-wide Swedish study. The Study Group on Infectious Diseases in Obstetrics and Gynecology within the Swedish Society of Obstetrics and Gynecology.

BACKGROUND: An increased use of prophylactic antibiotics to avoid postoperative infections in women undergoing hysterectomy has been observed in Sweden. This investigation was performed a) to study the infection rate to enable future evaluation of the effect of prophylactic antibiotic regimens and b) to identify subgroups suitable for intervention. METHOD: A prospective study comprising all women undergoing pelvic surgery with hysterectomy during a two month period at forty-two Departments of Obstetrics and Gynecology in Sweden. Relevant information regarding the surgical procedure and the postoperative course was included in a standardized form at discharge from hospital and at a follow-up visit 4 to 6 weeks after surgery. RESULTS: Of the 1060 women included in the study, 23% developed postoperative infections: 9.4% had wound-, cuff-, and/or deep infections, 13% urinary tract infections, and 4% other infections unrelated to the surgical site. Only half of the wound/cuff-/deep infections were diagnosed before discharge from the hospital. Wertheim-Meigs procedures, bleedings per-operatively exceeding 1000 ml, and presence of bacterial vaginosis (BV) were associated with an increased risk of postoperative infections. Among women undergoing abdominal hysterectomy for benign reasons, (n = 159), wound-cuff/deep infection was significantly associated with preoperative BV, i.e. 7 of 28 (25%) vs. 11 of 131 (8%), respectively (relative risk = 3.0, p = 0.01). Infection was associated with prolonged postoperative hospital stay. Antibiotics were given pre- or postoperatively to 236 (22%) of the 1060 women. Reduction in the postoperative infection rate was seen among women undergoing vaginal hysterectomy who were given pre- or postoperative antibiotics. CONCLUSION: The postoperative infection rate after hysterectomy was clinically significant in this population. Wertheim-Meigs procedures, peroperative bleeding > 1000 ml and BV were identified as risk factors for postoperative infections after hysterectomy.

Is misoprostol the drug of choice for induced cervical ripening in early pregnancy termination?

BACKGROUND: To study the effectiveness of three different cervical ripening agents in terms of dilatation ability and patient discomfort in connection with legal first trimester abortion. METHODS: Three randomized trials among unselected nulliparous women were performed 1: hygroscopic tent versus gemeprost, 2: misoprostol versus gemeprost and 3: misoprostol administered orally 17 versus 10 hours before vacuum curettage was performed. MAIN OUTCOMES: Dilatation ability, frequent gastrointestinal side effects, severe pain (patients' perception). RESULTS: In Trial 1, there was a tendency of a greater dilatation ability using the hygroscopic tent, while the experience of pain was a greater problem with gemeprost. In Trials 2 and 3, there were no significant differences in the dilatation abilities or gastrointestinal patient discomfort. There was a tendency towards a higher demand for narcotic analgesics in patients treated with gemeprost compared with all other groups. CONCLUSIONS: Gemeprost and misoprostol showed almost identical ability to dilate and caused patient-experienced discomfort to the same degree. The use of misoprostol may be preferred as it has the advantage of being both less expensive and easier to administer.

PIP: The findings of three randomized trials of different cervical ripening agents indicate that misoprostol may be the preferred drug for use in vacuum aspiration abortion. Evaluated were the performances of the hygroscopic tent versus gemeprost, misoprostol versus gemeprost, and misoprostol administered orally 17 versus 10 hours before vacuum curettage. Included in the trials were 406 randomly selected nulliparous women presenting for first-trimester induced abortion at a Swedish hospital. The 3-mm hygroscopic tent was inserted by a physician 17-19 hours before vacuum aspiration. The gemeprost vaginal suppository was inserted by the patients 3-5 hours before the procedure, while subjects in the misoprostol trials took three 200 mcg tablets 8-10 or 17-19 hours before abortion. The degree of cervical dilatation was greater with the hygroscopic tent (mean, 8.8 mm) than gemeprost (6.3 and 6.6 mm in trials 1 and 2, respectively) and misoprostol (7.5 and 6.9 mm in trials 2 and 3, respectively). The severity of pain measure was similar in all 3 groups, although gemeprost acceptors required more parenteral narcotic analgesics. The duration of pain was positively linked to the time between treatment and abortion. There were 13 cases of postoperative pelvic inflammatory disease, but no correlation with cervical ripening technique. Despite the slightly greater dilatation associated with use of the hygroscopic tent, misoprostol is recommended given its lower cost, lack of need for refrigeration, and potential for patient rather than physician administration. There was no difference in dilatation when misoprostol was administered 10 hours as opposed to 17 hours before vacuum aspiration; thus, the former regimen is urged to reduce the duration of pain.

Inflammatory signs in wet smear and Pap-smear compared with the histopathology from the female lower genital tract.

In order to define the relationship between various criteria of female lower genital tract inflammation, we examined wet smears, cervical smears and biopsies from 131 patients. The presence of clue cells in rehydrated dry smears showed a positive correlation to the presence of clue cells in Papanicolaou stained smears, and to some extent with cytological evidence of inflammation (Kappa 0.48). Cytological inflammatory findings correlated well with the presence of clue cells in rehydrated dry smears, but signs of inflammation diagnosed by histopathology did not correspond to findings in cytological smears, probably because these methods reveal inflammation at different sites.

Dilapan tent-gemeprost regimen vs. combinations of extra-amniotic Rivanol-Laminaria/Lamicel and oxytocin for second trimester abortion.

OBJECTIVE: To compare a new regimen for second trimester abortion using Dilapan and vaginal gemeprost suppositories with extra-amniotic Rivanol instillation and oxytocin i.v. immediately or 16 h after instillation. METHODS: A prospective study was performed in 153 women to analyze the induction-abortion interval, the use of analgesics and the complication rate. Wilcoxon's rank sum test was used for statistical evaluation. RESULTS: The mean induction-abortion interval was significantly shorter in the Dilapan-gemeprost-treated women than in the immediate or 16-h Rivanol-oxytocin-treated women, 12.5 vs. 23.3 and 26.8 h, respectively. The 24-h cumulative abortion rate was 91% in the former group vs. 49% and 61%, respectively. The use of analgesics was less frequent among the Dilapan-gemeprost-treated women, whereas the complication rate did not differ. CONCLUSIONS: The Dilapan-gemeprost treatment was advantageous with respect to a shorter induction-abortion interval and ease of handling. However a minority of women do not respond to this treatment and it is therefore necessary to employ alternative methods to complete the abortion in these cases.

Long, uniform Lactobacilli (Döderlein's Bacteria): a new risk factor for postoperative infection after first-trimester abortion.

OBJECTIVE: The production of hydrogen peroxide (H2O2) from different strains of lactobacilli in the vagina has been proposed to play one of the most important protective roles in the vaginal defense system. New data have, however, suggested that Döderlein's bacteria, with the morphological appearance of long lactobacilli, have a low production of H2O2 . The purpose of the present study was to correlate the morphology of lactobacilli with the incidence of infection following legal abortion. METHODS: Seven hundred sixty-nine women with lactobacilli but without Chlamydia trachomatis or bacterial vaginosis in their vaginal wet smears who were to undergo legal abortions were divided into 6 different groups according to the morphological appearance of the lactobacilli. The postoperative infection rates in these different groups were compared. A phenotypic classification of some of the lactobacilli was performed. RESULTS: The presence of Döderlein's bacteria compared with a mixed flora of lactobacilli increased the risk of postoperative infection 3 times [relative risk (RR) = 3.0; 95% confidence interval (CI), 1.5-6.3]. After a logistic regression analysis, the only independent risk factors were the presence of Döderlein's bacteria and earlier gestational age. CONCLUSIONS: We found that the lactobacilli regarded as commensal organisms and "normal, healthy lactobacilli" in the vagina were present in only 18% of these women and that their presence might be hazardous. Therefore, we must reconsider our concept of the "normal" lactobacilli in the vaginal wet smears of healthy women.

Detection of bacterial vaginosis in wet mount, Papanicolaou stained vaginal smears and in gram stained smears.

In a prospective study of 107 women, bacterial vaginosis was clinically diagnosed in 34 women. Compared with clinical diagnosis of bacterial vaginosis, detection of clue cells in Papanicolaou stained vaginal smears showed a sensitivity of 88.2%, a specificity of 98.6%, a positive predictive value of 96.8% and a negative predictive value of 94.7%. The corresponding values for detection of bacterial vaginosis in Gram stained smears compared with the clinical diagnosis were 100%, 97.3%, 94.4% and 100%, respectively. Compared with clue cells in wet smears, identification of clue cells in Papanicolaou stained vaginal smears showed a Kappa index of 0.87 and compared with Gram stain criteria a Kappa index of 0.94. The correlation between Gram stain and Papanicolaou stained vaginal smears showed a Kappa index of 0.89. In contrast to the results of earlier investigators our studies indicate that the demonstration of clue cells in Papanicolaou stained vaginal smears correlate reasonably well with the conventional clinical criteria. However, the Gram stain method may be more reliable than the Papanicolaou method.

Expanded gold standard in the diagnosis of Chlamydia trachomatis in a low prevalence population: diagnostic efficacy of tissue culture, direct immunofluorescence, enzyme immunoassay, PCR and serology.

OBJECTIVE: To evaluate the diagnostic efficacy of chlamydia culture, direct immunofluorescence (DFA), direct enzyme immunoassay (EIA), polymerase chain reaction (PCR) and serology by defining positive culture or at least two positive non-culture tests as true positive. SETTING: Three gynaecological departments located in separate areas of Sweden. PATIENTS AND DESIGN: All pregnant women requesting abortion during a six month period were included. In cases with unconfirmed non-culture tests, reculture with multiple passage and PCR on the culture transport medium was performed for confirmation. Serum was analysed for chlamydial antibodies type IgG, IgM and IgA using microimmunofluorescence. RESULTS: 18 of 419 (4.3%) patients were positive for chlamydia according to the defined criteria. Twelve of 419 (2.9%) were positive in standard culture (primary inoculation). The sensitivity of standard culture, DFA, EIA and PCR were 66.7%, 77.8%, 64.7% and 71.4% respectively. The specificity 100% (by definition), 99.5%, 100%, 100% respectively. The positive predictive value 100% (by definition), 87.5%, 100%, 100% respectively. Negative predictive value 98.5%, 99.0%, 98.5%, 98.9% respectively. Serum IgG titre of > or = 64 and > or = 1024 gave positive predictive values of 10% and 21% respectively. CONCLUSIONS: When an expanded gold standard is used, the specificity and positive predictive value of the non-culture tests used are comparable with that of standard culture even in this low prevalence population. Standard culture underestimated the chlamydia prevalence by 33%. The prevalence found represents a decrease from 10 to 2.9% of culture verified chlamydia during four years in comparable populations. Chlamydial antibodies of certain immunological classes are not necessarily present in cases with chlamydia.

Bacterial vaginosis and cervical intraepithelial neoplasia.

METHODS: In an attempt to investigate an association between the finding of clue cells in Papanicolaou-stained (PAP) smears and cervical intraepithelial neoplasia (CIN), a total of 6150 smears from 1976 were re-investigated. RESULTS: Clue cells representing bacterial vaginosis were present in 10% of the PAP-smears. CIN II and III alone, as well as all CIN cases, were more common in women with bacterial vaginosis (p < 0.001). Histologically CIN III/carcinoma in situ was found in nine patients with and in 16 patients without bacterial vaginosis. The relative risk of having CIN III/carcinoma in situ if the women had bacterial vaginosis was 5.0 with 95% confidence interval of 2.2-11.6. CONCLUSIONS: The possibility exists that bacterial vaginosis is in some way associated with the development of cervical intraepithelial neoplasia, i.e. as a cofactor to human papilloma virus. Therefore, bacterial vaginosis must be taken in consideration in future studies on CIN.