Simplified gram stain interpretive method for diagnosis of bacterial vaginosis.

OBJECTIVE: Two methods which only consider bacterial morphotypes and require counting various types of bacteria, for interpreting Gram stains of vaginal secretions for the purpose of diagnosing bacterial vaginosis were previously described. A new interpretative method, which combines clue cells with bacterial morphotypes as diagnostic markers, is offered. STUDY DESIGN: One hundred twenty patients were tested with the new interpretative method and clinical criteria of bacterial vaginosis as the reference standard. RESULTS: The new method was found comparable to older methods when all were compared with clinical diagnostic methods. The sensitivity of the clue cell method was 86.1% and the specificity was 92.8%. The predictive value of a positive test was 83.7% for a population with a prevalence of bacterial vaginosis of 30%. The negative predictive value was 94.0%. CONCLUSION: If clue cells are present and the nonlactobacilli morphotypes exceed lactobacilli morphotypes, bacterial vaginosis can be diagnosed without the need for exact bacterial counts.

Is trichomoniasis often associated with bacterial vaginosis in pregnant adolescents?

The same criteria for identifying bacterial vaginosis are often present in women with trichomoniasis. These criteria include elevated vaginal pH, vaginal odor, homogeneous discharge, increased anaerobic bacteriologic vaginal flora, and elevated levels of bacterial enzymes. Clinically mixed vaginal infections occur, and because the treatment for these two conditions can be different, it is important to distinguish between them. Trichomoniasis can interfere with a Gram stain diagnosis or the proline aminopeptidase test for bacterial vaginosis. Clue cells are not generally found in women with Trichomonas vaginalis, but when present, they strongly indicate the concomitant presence of bacterial vaginosis.

Bacterial vaginosis: current review with indications for asymptomatic therapy.

Bacterial vaginosis is a definable clinical entity whose exact origin is unknown. A shift in normal vaginal flora from aerobic, predominantly but not exclusively lactobacilli, to a predominantly anaerobic flora characterizes the condition. More than one half of all women with bacterial vaginosis have no symptoms. The condition is not entirely benign. The potentially pathogenic bacteria present in the vagina in large numbers place these women at risk for postoperative morbidity and adverse obstetric outcome. Sexual transmission has not been proved, but therapeutic cures sometimes require that patient and partner be treated simultaneously. Recommended therapy is with metronidazole or clindamycin and must be given for 7 days for maximal effectiveness. Recurrence of disease can be a problem.

PIP: Bacterial vaginosis, the most common infectious cause of vaginitis, is characterized by a shift in normal vaginal flora from predominantly aerobic to mainly anaerobic flora. Incidence rates in various studies have ranged from 1-0-45%. Unclear are both the pathophysiology of the observed reduction in lactobacilli and whether bacterial vaginosis is sexually transmitted. Evidence for sexual transmission is provided by a significantly greater incidence of bacterial vaginosis in women with more than 5 sexual partners compared to those in a lifelong monogamous relationship, while bacterial detection in virgins and the failure to demonstrate benefits of partner treatment argue against sexual transmission. Use of an IUD does appear to be a risk factor. The only symptom consistently reported by women with bacterial vaginosis is a malodorous vaginal discharge; however, over half of women with this condition are asymptomatic. The standard criterion for diagnosis includes the presence of 3 of the following signs: the clue cell, homogeneous discharge that adheres to but is easily wiped off the vaginal wall, an elevated vaginal pH, and the potassium hydroxide test for volatile amines. Of concern is the potential for serious sequelae. The abnormal bacterial flora associated with vaginosis significantly impede the white blood cell response to infection. Gynecologic-obstetric sequelae include pelvic inflammatory disease, endometritis, postoperative vaginal cuff infections, preterm labor, premature rupture of membranes, and chorioamnionitis. Metronidazole and clindamycin are the most effective treatment agents. Given the long-term risks associated with bacterial vaginosis, a full course of antibiotics should be considered in both obstetric patients with this condition and women undergoing outpatient ambulatory invasive surgical procedures.

Terconazole for the treatment of vulvovaginal candidiasis.

A double-blind, randomized trial was conducted to evaluate the efficacy and safety of terconazole for vulvovaginal candidiasis. Treatment consisted of daily intravaginal application of one of the following regimens: 80-mg terconazole suppositories for 3 days, miconazole nitrate suppositories for 7 days or placebo suppositories for 7 days. The terconazole and miconazole nitrate groups had significantly higher therapeutic cure rates than did the placebo group. Evaluation of vaginal secretions with microscopic examination showed no evidence of leukocyte proliferation. Proline aminopeptidase activity, present in patients who have bacterial vaginosis, could not be detected in the vaginal secretions from patients with yeast vulvovaginitis.

Is pH test paper as accurate as the electronic measurement of the pH of vaginal secretions?

A comparison was made of two brands of pH test paper and electronic instrumentation for measuring the pH of vaginal secretions. When the pH of vaginal secretions was greater than 4.5 (abnormal), there was no significant difference between the methods, showing that pH test paper is reliable for pH determination of vaginal secretions.

Growth of Trichomonas vaginalis in commercial culture media.

There are only two commercially available, ready-to-use culture media which are approved by the Food and Drug Administration for clinical diagnosis of vaginal trichomoniasis: Kupferberg's STS and Diamond's medium (modified). Diamond's medium (Klaas modification), recommended by the Centers for Disease Control for the isolation of Trichomonas vaginalis, was compared in vitro to Kupferberg's (STS) medium. Growth studies using six fresh clinical isolates, all from different patients, showed that while generation time was about 6 h in both STS and Diamond's, the period of exponential growth was longer in Diamond's. More important, in STS there was a 4-h lag period during which the population significantly decreased prior to exponential growth. This did not occur in Diamond's medium. Three hundred organisms inoculated into Diamond's reached a population of over 10(5) organisms in 72 h. In STS, the same inoculum could multiply to only 6 x 10(3) organisms. The fact that there is a lag phase in STS which is not seen in Diamond's could explain why low numbers of T. vaginalis do not multiply in STS but do multiply and can be detected in Diamond's. We conclude that because Diamond's medium (modified) allows more prolific growth over a shorter period of time, it is more suitable than Kupferberg's (STS) for detecting T. vaginalis in patients with vaginitis.

Statistical evaluation of diagnostic criteria for bacterial vaginosis.

Bacterial vaginosis is the most common cause of vaginitis in women of reproductive age. In an attempt to clarify diagnosis of this condition, various parameters of signs and symptoms and groups of parameters were compared with classical diagnostic criteria in 310 patients. There was no significant difference in positive diagnosis rates between the Amsel et al. criteria and those of Thomason et al. (p = 0.25). The single most reliable indicator of bacterial vaginosis was the presence of clue cells on wet mount examination of vaginal secretions (sensitivity 98.2%, specificity 94.3%, positive predictive value 89.9%, negative predictive value 99.0%). The best two combinations of parameters for rapid accurate clinical diagnosis were clue cells and odor on alkalinization (sensitivity 99.5%) positive predictive value 98.8%, negative predictive value 92.1%). Gram stain criteria (bacterial morphologic types) were less accurate predictors of the disease (sensitivity 97.0%, specificity 66.2%, positive value 57.2%, negative predictive value 97.9%). Even when the bacterial morphologic type criteria were combined with presence of clue cells, predictive accuracy did not exceed that of clue cells on wet mount examination alone (sensitivity 93.9%, specificity 84.7%, positive predictive value 74.2%, negative predictive value 96.8%). Homogeneous discharge was found to be of little diagnostic value.

Trichomonas vaginalis.

Although not life-threatening, trichomoniasis is a major health problem, especially in women. It is transmitted sexually and has a very high incidence worldwide. We argue that Trichomonas vaginalis is not a commensal. Attempts should be made to identify this pathogen in all suspicious cases. Although culturing remains the most sensitive method to detect T vaginalis, newer methodologies such as monoclonal antibody fluorescent and direct staining make it possible to diagnose an infection with high sensitivity and specificity and to treat the patient at the same visit. Treatment is effective, approaching a 100% success rate when sexual consorts are also treated. Only by aggressive attempts to diagnose and treat this disease can we stop its spread.

Advances in the understanding of bacterial vaginosis.

Bacterial vaginosis is a disease that results from massive overgrowth of vaginal bacterial flora. The exact etiology is unknown. The major components of the normal bacterial flora, peroxide-producing lactobacilli, are replaced by non-peroxide-producing lactobacilli, allowing overgrowth of anaerobic and gram-negative aerobic flora. Anaerobic bacteria produce enzymes, aminopeptidases, that degrade protein and decarboxylases that convert amino acids and other compounds to amines. Those amines contribute to the signs and symptoms associated with the syndrome, raising the vaginal pH and producing a discharge odor. The excessive amounts of bacteria characteristic of the syndrome attach to epithelial cell surfaces, resulting in "clue" cells. Nearly half the patients report no noticeable symptoms, but many develop a characteristic copious, malodorous discharge within six months if untreated. Serious infectious sequelae are associated with bacterial vaginosis, including salpingitis, abscesses, endometritis and pelvic inflammatory disease. There is also a danger to pregnant women because of premature rupture of the membranes and premature labor. Clindamycin and metronidazole are considered effective therapy for the disease. Treatment of sexual partners remains controversial since sexual transmission has not been proven unequivocally.

Comparison of Diamond's medium modified and Kupferberg medium for detection of Trichomonas vaginalis.

Diamond's medium modified, the commercially available version of the Klaas modification of Diamond's medium, was compared to Kupferberg medium and to direct wet-mount examination for detection of trichomonads in symptomatic patients. Diamond's medium was found to be superior to both Kupferberg medium (P less than 10(-4) and wet mount (P less than 10(-6].

Quality control standards for the proline aminopeptidase assay used to detect bacterial vaginosis.

We describe the preparation and usage of positive and negative quality control standards for the proline aminopeptidase assay of vaginal secretions as an indicator of bacterial vaginosis.

Effectiveness of Gonozyme for detection of gonorrhea in low-risk pregnant and gynecologic populations.

The Gonozyme Diagnostic Kit (Abbott Laboratories, North Chicago, IL) was evaluated in a population at low-risk for gonorrhea. Two hundred seventy-two female patients, consisting of 152 obstetrical patients and 120 gynecologic patients attending a family-planning clinic, were tested for Neisseria gonorrhoeae by use of culture and Gonozyme testing of specimens from the endocervical canal. The average ages were 16.1 years for the obstetric patients and 17.8 years for the gynecologic patients. The overall sensitivity of Gonozyme as compared with culture was 60.8% (57.1% for obstetric patients and 62.5% for gynecologic patients); the overall specificity was 98.4%. An association was observed between the number of colony-forming-units (cfu) per plate and a positive Gonozyme test; specimens with less than 60 cfu/plate usually yielded a negative Gonozyme result. We conclude that Gonozyme is a highly specific test but its low sensitivity as compared with that of culture does not allow it to replace standard culture media for screening of low-risk women for infection with N. gonorrhoeae.

Is analysis of vaginal secretions for volatile organic acids to detect bacterial vaginosis of any diagnostic value?

A study was undertaken to determine which fermentation products in vaginal secretions serve as the best markers for bacterial vaginosis. Three categories of markers had been previously identified, but due to cost considerations it was necessary to accurately determine if analysis for all three were necessary. With the use of vaginal secretions from well-defined patient populations, we applied gas-liquid chromatography to test for both volatile and nonvolatile acid fermentation products. It was found that only the detection of nonvolatile acids was necessary, because no additional patients with bacterial vaginosis were identified by subsequent analysis of volatile acids. Routine analysis of volatile acids to identify patients with bacterial vaginosis is labor intensive and costly and had no clinical diagnostic value.

Comparison of four methods to detect Trichomonas vaginalis.

Four methods for the detection of Trichomonas vaginalis in vaginal secretions from 88 symptomatic patients were compared: wet-mount examination, Kupferberg liquid medium, Hirsch charcoal agar, and the Papanicolaou smear. Hirsch diphasic slants and Kupferberg medium did not significantly differ in sensitivity from direct examination of wet mounts. The Papanicolaou smear identification of trichomonads was found to be the least sensitive method evaluated.

Evaluation of Chemstrip LN in a male geriatric population.

A dipstick method for bacteriologic screening of urine using a combination of leukocyte esterase and nitrite indicators, Chemstrip LN, was evaluated in a male inpatient geriatric population. Three hundred seventy-five urine samples from 198 symptomatic patients (age range, 45 to 98 years; mean, 69.6) were tested by LN and cultured using standard methods. LN showed 169 negatives, but in 7 of these cases, the culture was positive (greater than or equal to 5 X 10(4) col/mL). Organisms recovered from these cultures were Candida (not albicans), Group B Streptococcus, Staphylococcus aureus, and in four cases, Pseudomonas aeruginosa. In the four cases of missed Pseudomonas the patients were receiving aminoglycoside therapy for 2 to 6 days and/or drugs that dye the urine. Despite the fact that many patients were on medications known to interfere with the test strips, the negative predictive value was still 96.0% showing that the strips may be suitable for use with this population given the understanding that certain antibiotics and/or substances that dye the urine may give false negative results.

Proline aminopeptidase activity as a rapid diagnostic test to confirm bacterial vaginosis.

Two biochemical indicators of bacterial vaginosis, proline aminopeptidase activity and gas-liquid chromatographic analysis, were compared. Five hundred women had their vaginal secretions tested for pH, presence of a positive amine test, levels of volatile and nonvolatile short-chain organic acids, and proline aminopeptidase activity. In addition, direct microscopic and Gram stain examinations were performed. Of the 500 women, 349 (70%) had some form of vaginitis. One hundred sixteen were diagnosed as having bacterial vaginosis, and 69 of these (59%) had Mobiluncus sp on either direct microscopic or Gram stain examination. Two hundred thirty-three had either mixed or other forms of vaginitis. One hundred fifty-one patients were normal. The sensitivity of the proline aminopeptidase assay was 83 and 79%, respectively, in patients having bacterial vaginosis with and without Mobiluncus morphotypes. In contrast, gas-liquid chromatography of short-chain organic acids had sensitivities of 71 and 30%, respectively. Specificity of both assays was about 95%. The greater sensitivity of the proline aminopeptidase assay, especially in patients without Mobiluncus morphotypes, proves its superiority.

Vibrio cholerae non-01 cellulitis.

A 46-year-old man presented with pain and fever and a postphlebitic ulcer on his left leg. The wound was suppurative and open at the margins, but there was little underlying fasciitis and no apparent muscle or blood vessel involvement. Three separate wound cultures were obtained at two-day intervals, and all showed only Vibrio cholerae non-01. The patient was successfully treated with cefazolin sodium. This marks the second documented case of V cholerae non-01 type alone as a causative agent of cellulitis, and the first case where no saltwater origin could be demonstrated.