Major changes are looming in diagnostic laboratory testing and are being driven by health-care reform, the Clinical Laboratory Improvement Act (CLIA), automation, and advances in molecular diagnostic testing. As federal regulations impact on office and other low-volume laboratories, more specimens are being sent to larger commercial and state laboratories where the trend is to use automated antigen testing methods such as enzyme immunoassay (EIA) or deoxyribonucleic acid (DNA) probe technology. Unfortunately, few of these methods have been studied extensively in infants or prepubertal children. In general, they cannot provide a definitive microbiological diagnosis for medicolegal cases, and owing to less than 100% specificity, these tests do not perform well in predictably low prevalence, prepubertal pediatric populations. As an example of this problem, in a population with a high prevalence of infection ( 10%), a test with a sensitivity of 80% and a specificity of 99% will yield a positive predictive value (PVP) of 88.9% and a negative predictive value (PVN) of 98.9%. In a population with a low prevalence (1%), the PVP would be only 40% (ie, 60% of positive tests would be false positive). Thus, a test appropriate for screening high-risk, sexually active adolescents may be a poor choice for children.
This article discusses the performance characteristics of commonly available diagnostic tests for the common sexually transmitted infections, the situations in which they are recommended, and their proper interpretation. Because costs vary tremendously by test, laboratory, volume, and purchasing power of the care provider, this important practical issue in diagnostic decision making will not be discussed.
Gonorrhea is the most common sexually transmitted disease found in abused children, accounting for 7% to 20% of infections.1 In neonates born to infected mothers, 26% to 40% will have a positive orogastric fluid. Gonococcal conjunctivitis occurs in 0% to 5% of infants who have received silver nitrate ocular prophylaxis, and 2% to 30% of infants with no ocular prophylaxis. Gonococcal scalp abscesses associated with intrauterine fetal monitoring have been reported.
The recommended sites for the recovery of Neisseria gonorrhoeae ate listed in Table 1. Laboratory diagnosis of genital gonococcal infections may be either presumptive or definitive. A definitive diagnosis requires isolation of the organisms and confirmation of its identity by carbohydrate utilization, rapid enzyme substrate tests, serologic methods, or DNA probe technique. A presumptive diagnosis requires that typical gram-negative intracellular diplococci be seen on a smear from a genital specimen and also requires growth of typical oxidase-positive colonies composed of gram-negative diplococci on selective media such as modified Thayer-Martin agar or detection by a nonculture laboratory test such as EIA or DNA probe.
When the specimen is from a child, a definitive diagnosis is always required, and the culture should be confirmed by two tests that are based on different principles (eg, carbohydrate utilization and fluorescent antibody). A false-positive identification of N gonorrhoeae can cause damaging emotional, social, and legal problems because of the implication of sexual abuse.
At this time, when properly performed, the traditional culture for N gonorrhoeae on selective media remains the most sensitive and specific test available. Specimens must be plated onto selective media or placed into a transport system immediately. The highest sensitivity is obtained when specimens are inoculated onto media immediately in the examining room. Improper incubation conditions or defective media decrease culture sensitivity. Plates should be incubated immediately at 340C to 36°C in an atmosphere of 3% to 10% carbon dioxide and 70% humidity for at least 48 hours. Selective media contains vancomycin to inhibit gram-positive organisms. If there is >3 µg/ml vancomycin in the media, growth of susceptible gram-negative organisms will be inhibited. The reported prevalence of these organisms has varied between 2% and 30%.
Speciation (confirmation) of isolates (eg, carbohydrate utilization, fluorescent antibody, etc) requires an additional 30 minutes to 48 hours. Although confirmation tests in general are reliable, there are problems with each method. In one report from the Centers for Disease Control, 14 of 40 isolates submitted as suspected N gonorrhoeae from children failed to be confirmed.2 The most common problem with carbohydrate utilization occurs when cultures contain other organisms.
The Syva monoclonal fluorescent antibody test (Sy va Company, San Jose, California) is highly specific and takes only 30 minutes to perform. However, its sensitivity has decreased since it was first evaluated in 1987 and approximately 5% of strains will not react with it. As the antigenic composition of prevalent strains can vary over time, so can the performance of serological tests used to detect them. The Gen-Probe DNA probe confirmation test (GenProbe Ine, San Diego, California) takes only 30 minutes. Reported sensitivity ranges from 85.7% to 100% and specificity from 99% to 100%. However, experience with this test is still limited.
Direct Testing for Neisseria gonorrhoeae
The DNA probe assay Race II (Gen-Probe Ine, San Diego, California) is a presumptive test that currently is approved only for genital specimens. It is not recommended for use in pediatric patients.
The Gram stain is one of the oldest, least expensive, and most useful methods to diagnose gonorrhea in symptomatic urethritis (Table 2). However, in children it should only be used to initiate therapy while awaiting definitive culture results.
There are no practical serologic tests that can reliably distinguish current from past gonococcal infection.
Nonspecific Tests for Urethritis
The leukocyte esterase dipstick test uses a dipstick to detect the presence of polymorphonuclear leukocytes in specimens. Granulocytic enzymes are detected by their reaction with reagents on the strip. Strips can be read by minimally trained people and have been used to screen for urinary tract, gonococcal, and chlamydial infections. The leukocyte esterase test should be limited to settings where Gram stains are not possible and used mainly to determine the need for more definitive tests. It has no current role in the diagnosis of sexually transmitted diseases in infants or children.
Recommended Sites for the Recovery of Neisseria gonorrhoeae and Chlamydia trachomatis in Neonates and Children*
CHLAMYDIA TRACHOMATIS INFECTIONS
Chlamydia trachomatis is the most common sexually transmitted bacteria in the United States. Approximately 40% to 70% of infants born to infected mothers will acquire the organism at birth, 30% as conjunctivitis or pneumonia.3 The eye, nasopharynx, vagina, and anal canal can all be colonized, and untreated infants can carry the organism for 3 years or longer. After infancy, the carriage of C trachomatis becomes uncommon and is associated with sexual abuse. The recommended sites for recovery are listed in Table 1 . The primary target for C trachomatis is a nonciliated columnar, cuboidal, or transitional epithelial cell. The spectrum of chlamydial disease is limited to sites in which these cells are found, such as mucous membranes of the urethra, the endocervix, the anorectum, the posterior nasopharynx, and the conjunctivae. Whichever test is used for chlamydial detection, it is important that a good specimen containing these cells be obtained. Because of the presence of maternal antibody, the sensitivity of diagnostic tests may be lowered after birth for 3 months or longer. Current tests include cell culture and nonculture methods that detect antigen (eg, direct fluorescent antibody [DFA], EIA, or specific sequences of nucleic acids such as DNA probes or polymerase chain reaction [PCR]). Laboratory diagnosis may be definitive or presumptive.
Definitive diagnosis requires isolation of C trocho' maus in tissue culture or a positive EIA, DFA, DNA probe, or PCR test confirmed by a second method.
Cell culture is still considered the gold standard method. Although under the best of conditions, its sensitivity ranges from 70% to 90%, its specificity should be 100%. It is the recommended method for use in suspected cases of child abuse or in any case where a false-positive result would have serious consequences. It is also the recommended method for the detection of chlamydia in the urethra of asymptomatic boys, the anorectum in patients of all ages, the nasopharynx in infants, and the vagina of prepubertal girls.4 Technical factors affect sensitivity. The specimen must be inoculated into chlamydial transport media and immediately stored at 4°C. If the sample cannot be inoculated onto cells within 24 hours, it must be frozen at -700C. Optimal processing of specimens requires access to a laboratory experienced in chlamydia tissue culture work. The time required to process a culture is 48 to 72 hours.
Direct Fluorescent Antibody
The specimen is applied to a slide, stained with fluorescent antichlamydial antisera, rinsed, and examined under a fluorescent microscope for the presence of apple-green elementary bodies. The adequacy of the specimen can be determined by the presence of columnar epithelial cells. After fixing the specimen, slides can be kept at room temperature and held for 7 days before staining. It takes about 30 minutes to stain and read slides. Direct fluorescent antibody is approved for detection of chlamydia in neonatal conjunctival specimens. The disadvantages of this method are that a fluorescent microscope is required along with an experienced microscopist, and it does not lend itself well to large numbers of specimens. False-positive results have been reported to occur occasionally with all antisera.
Studies evaluating the Syva MicroTrak test (Syva Company, San Jose, California) for the detection of C trachomatis in conjunctival specimens, its only recommended use in infants and prepubertal children, have shown sensitivity ranging from 93% to 100% and specificity from 70% to 98%. Most discrepancies between culture and DFA results occur in specimens with low numbers of organisms.
Sensitivity and Specificity of Diagnostic Tests for Sexually Transmitted Infections*
Enzyme immunoassays use either plastic beads or wells to capture chlamydial lipopolysaccharide antigen from clinical specimens. Antisera that has been labeled with an enzyme reacts with the antigen. An appropriate substrate is added and a color reaction takes place that is measured in a spectrophotometer. The amount of color produced is proportional to the amount of antigen in the specimen. The method is approved by the Fcxod and Drug Administration (FDA) for testing conjunctival and nasopharyngeal specimens from neonates. Data are limited, however. Conjunctival specimen sensitivity has ranged from 87% to 98% and specificity from 94% to 100% with Chlamydiazyme (Abbott Laboratories, North Chicago, Illinois). It is not recommended for nasopharyngeal specimens (sensitivity 66.7% and specificity 95.7% ).5
Advantages of the EIA are that it requires less technical expertise than culture or DFA, several hundred specimens can be assayed in 4 hours, and test results are measured objectively, thus eliminating the subjective element present in fluorescent microscopy. Enzyme immunoassay is not approved for rectal specimens. Swab specimens may be stored for 1 week before testing. Like other direct-antigen tests, performance is best when high antigen levels are present. False-positive reactions have been documented showing cross reactions between N gonorrhoeae, Escherichia coli, group A beta hemolytic streptococci, and other bacteria. Recently, rapid filtration immunoassays designed for lowvolume testing have been developed. Limited studies have shown sensitivity and specificity similar to standard EIA methods. Further testing is needed to determine their performance in various clinical settings.
This method, which uses a chemoluminescent probe complementary to the ribonucleic acid (RNA) of C trachomatis is becoming widely used by larger laboratories. It is approved only for specimens from adults and is not yet recommended for neonates or prepubertal children. Further testing is needed to evaluate its performance in low-prevalence populations and with specimens from nongenital sites.
Polymerase Chain Reaction
The polymerase chain reaction test Amplicor (Roche Diagnostic Systems, Branchburg, New Jersey) is a direct DNA probe test that uses nucleic acid amplification and nucleic acid hybridization for the detection of C trachomatis cryptic plasmid DNA in endocervical, male urethral, and urine specimens. The reported sensitivity ranges from 93% to 97%, and the specificity ranges from 99% to 100%. There are insufficient clinical data to recommend its use in neonates or children.
There are no practical serologic tests that can distinguish current from past infection. Immunoglobulin M (IgM) microimmunofluorescence is useful in the diagnosis of infant pneumonia caused by C trachomatis (titer of ^1:32 is strongly suggestive). However, the test is not widely available.
Laboratory diagnosis may be made either by direct identification of Treponema pallidum in clinical specimens or by serology. Acquired syphilis in children is diagnosed as in adults.
A positive microscopic identification of typical treponemes in a lesionai specimen is considered a definitive diagnosis of syphilis. Methods include darkfield examination, direct fluorescent antibody, hematoxylin and eosin stain, and silver stain. The darkfield examination is regarded as highly sensitive and specific in nonoral, moist lesions of primary and secondary syphilis (mainly condylomata lata). It is the method of choice when the wet mounts can be examined immediately (within 10 to 30 minutes) so that characteristic motility can be observed. Due to confusion with commensal organisms, specificity may be decreased when examining rectal lesions. Oral lesions near the gingival margins cannot be reliably interpreted if Treponema denticol, which is microscopically similar to T pallidum, is present. A negative dark field does not rule out syphilis. False-negative results may be due to the age or condition of the lesion (dry skin rashes are not suitable), prior antibiotic treatment, debris, erythrocytes obscuring the field, or most commonly, poor technique in collecting the sample or in identifying the organism.
Serologic tests are classified as either nontreponemal or treponemal. Neither is suited to occasional office use.
Nontreponemal tests are flocculation tests that measure anticardiolipid antibodies. The most commonly used are the Venereal Disease Research Laboratory (VDRL) and the rapid plasma reagin (RPR) tests. The VDRL is the only test standardized for use with cerebrospinal fluid. A false-positive rate of 1% to 2% is found in the general population regardless of the nontreponemal test used.6 A broad common denominator for false-positive cardiolipin antibodies may be the inflammation of vascular endothelium that occurs with certain systemic infections, autoimmune diseases, and chronic injection drug abuse. Falsenegative results occur when syphilis is acquired recently and seroconversion has not yet occurred, and uncommonly during secondary or congenital syphilis as a result of extremely high levels of antibody (the prozone phenomenon). Only nontreponemal tests are used to monitor the response to treatment.
Specific treponemal tests (eg, fluorescent treponemal antibody absorption [FTA-ABS]), and microhemagglutination assay for T pallidum detect specific antibodies against T pallidum cellular components. They are recommended as confirmatory tests and to diagnose very early syphilis (FTA-ABS only) before nontreponemal tests have become reactive, as well as to diagnose late syphilis when nontreponemal tests may have reverted to negative. When the general population is tested, a 1% false-positive rate is found.6 False-positive FTA-ABS results have been associated with Lyme disease, genital herpes, autoimmune diseases, and injected drug use. Treponemal tests usually remain reactive many years after effective treatment.
The risk of congenital syphilis infection varies with the timing and stage of infection, increasing rapidly after 20 weeks gestation and approaching 100% with untreated secondary syphilis. A definitive diagnosis can be made when typical treponemes are detected in the placenta or the baby (eg, umbilical cord, skin lesions, and nasal discharge). A presumptive diagnosis can be made in infants of mothers with untreated syphilis at delivery or evidence of reinfection or relapse after therapy. Additional criteria include7: physical evidence of active disease, evidence of syphilis on radiographs of the long bones, elevated cerebrospinal fluid cell count (>5 leukocytes/mm3), or protein (>40 mg/dL) without other cause.
The presence of nontreponemal antibodies in an infant's serum is not necessarily diagnostic of congenital syphilis. A rising nontreponemal test titer in serial serum samples over a period of 8 months is required for diagnosis. Both nontreponemal and treponemal antibodies (immunoglobulin G [IgG]), may be acquired transplacentally from the mother. Indeed, for screening purposes, maternal serology is superior to cord blood analysis for identifying newborns at risk of congenital syphilis. False-positive and false-negative results can occur with cord blood specimens, and in one study, cord blood detected only 67% of babies at risk while maternal blood detected almost all.8
It is important to remember that because of late maternal infection and prozone reactions, even a negative maternal and infant serological result does not rule out the possibility of congenital syphilis. Since IgM detected in a baby is not of maternal origin, a more specific test is the FTA-ABS-IgM. Specificity has been reported at 100%. However, because of low sensitivity (73%), a negative test is inadequate to rule out congenital syphilis.
HUMAN IMMUNODEFICIENCY VIRUS
Human immunodeficiency vims (HIV) infection usually is diagnosed by detection of specific antibodies. Other techniques, which are limited to a few centers, include isolation of the vims and amplification of nucleic acid sequences of HIV from infected cells or fluid. A presumptive diagnosis is made by detection of antibodies to HIV on the same semm specimen by two EIAs and confirmation by Western blot or another specific test. Most HIV EIA kits on the market are automated, take approximately 4 hours to perform, and are 98.1% to 100% sensitive and 99.6% to 100% specific.9 False-positive results usually are found with moderate or weakly reactive specimens that have an optical density of 1.0 to 2.2. 10 Western blot kits are available commercially and detect antibodies to proteins of the three major gene regions. The definition of a positive Western blot is arbitrary; however, the Centers for Disease Control and Prevention consider a Western blot to be positive if bands to any two of p24, gp41, or gpl20/P60 are present.11 False-positive, false-negative, and indeterminate Western blot results are very uncommon. Patients with indeterminate results should be retested in 3 to 6 months. Over-the-counter kits that test saliva for HIV antibodies have been developed by several companies, but are not yet approved by the FDA.
Diagnosis of HIV infection in patients older than 18 months of age is the same as in adults. Diagnosis of HIV infection in babies of less than 18 months can be difficult because of the transfer of maternal IgG to the fetal circulation. False-negative results can occur, and seroconversion can take up to 3 years after delivery. Polymerase chain reaction, which can be as sensitive as culture (97%), is a promising method for the diagnosis of HIV in infants.
It is estimated that the annual incidence of neonatal herpes simplex virus (HSV) infection in the United States is approximately one per 7500 live births, with the most serious infections occurring during or shortly after vaginal delivery to a woman with recent genital infection. Genital HSV infections (type 1 or type 2) are rare in prepubertal children. While it can be transmitted nonsexually, its presence should raise the possibility of child abuse. A definitive diagnosis of HSV infection requires either a positive culture or detection of the virus by direct antigen methods. A Tzanck (cytologic) smear is at best a presumptive test for current infection, and serologic tests cannot yet distinguish current from past infection or type 1 from type 2 confidently.
Growth of HSV in tissue culture is still the standard and most reliable diagnostic method. Specimens may be stored at 4°C for up to 3 days before processing. For longer periods, they may be stored at - 70°C. The median time required for development of characteristic cytopathic effect is 2 to 3 days. Ninety-five percent of all positive specimens will be detected within 7 days. Sensitivity depends on the amount of virus in the specimen, which varies with the stage of disease: vesicle (95%), pustule (70% to 85%), early first episode ulcers (70% to 80%), ulcers from recurrent infection (35% to 40%), and crusted lesions (25%).12 Confirmation and HSV typing can be accomplished with commercial immunofluorescent antisera.
Direct Antigen Methods
Rapid diagnosis of HSV infection is particularly important in neonatal disease and in immunocompromised patients. Several direct antigen method tests are available commercially. The completion time ranges from 15 minutes to approximately 3 to 4 hours. The tests perform best in symptomatic patients with early lesions, and according to the commercial test used, sensitivity ranges from 64% to 99% and specificity from 89% to 99%. Performance is dependent on the quality of the specimen (ie, whether enough cells were obtained).
A definitive diagnosis of human papillomavirus (HPV) infection requires either identification of characteristic changes (koilocytosis) on histologic examination of a biopsied lesion or detection of HPV antigen or nucleic acid. Papanicolaou smears of cervical cells that show typical cytologic changes are only suggestive of HPV infection. Biopsy tissue samples of approximately 3 mm3 are required for adequate analysis. Exfoliated cells obtained from the cervix and vagina by use of swabs, scrapers, or aspirators also are suitable. Cervical specimens should include cells from the transformation zone. All specimens need to be immediately frozen on dry ice, and stored at - 700C to prevent degradation. Either a hematoxylin and eosin or a Papanicolaou stain can detect characteristic cytologic changes associated with HPV. However, they cannot identify the HPV type present.
Approximately 10% of women with normal cervical cytology will demonstrate HPV DNA.13 Sixty-six types of HPV have been characterized. A subset that includes types 16, 18, 31, 33, and 51 is frequently associated with anogenital cancers. Specific typing only rarely will be indicated in children, and while a number of sophisticated nucleic acid hybridization tests are used in research laboratories, the commercially available ViraPap and ViraType HPV detection and typing kits (Digene, Gaithersburg, Maryland) are among the simplest and fastest. Currently under development is the PCR test for HPV DNA. This test can theoretically detect a single copy of HPV DNA. However, its extreme sensitivity also may be a problem because false positives due to miniscule amounts of contamination may occur.
Molluscum contagiosum is caused by a large pox virus that appears to spread by direct contact (possibly through abrasions), although pools and public baths have been implicated. Culture techniques for isolation of the molluscum contagiosum virus are unavailable. Lesions have a distinctive appearance and confirmation can be made by microscopic examination of a Wright's, Giemsa, or Gram stain of the expressed white core of the papule or by a hematoxylineosin stained punch biopsy. Hypertrophied epidermal cells contain multiple cytoplasmic inclusions made up of vacuoles containing virus particles. While the virus may be sexually transmitted, its diagnosis should not be considered evidence of sexual abuse.
The anatomy, physiology, and indigenous microbial flora of the vagina are age dependent. During the first 4 to 6 weeks after birth, the neonatal vagina, under the influence of maternal estrogen, is lined by stratified squamous epithelial cells. After estrogen disappears, the vagina becomes lined with cuboidal epithelial cells. At puberty, with the production of estrogen, the vagina again becomes lined with squamous epithelial cells. The neonatal vagina is resistant to perinatal gonococcal and chlamydial infections but susceptible to yeast and Trichomonas vaginalis.
From infancy to puberty, the vagina is susceptible to chlamydial and gonococcal infections but resistant to Candida and Trichomonas. Other causes of vaginitis in this age group are Streptococcus pyogenes (group A ß-hemolytic streptococcus) and Shigella. After puberty, the most common causes of vaginitis are Candida albicans, T vaginalis, and bacterial vaginosis, an infection in which Gardnerella vagnialis acts synergistically with other anaerobes that normally inhabit the vagina.
The epidemiology of trichomoniasis is not completely understood. Although prevalence rates in adolescents and young adults correspond with the usual risk factors for sexually transmitted diseases, prevalence rates in older women remain atypically high into the fourth and fifth decades of life.
Trichomonas vaginalis infection may be diagnosed by mixing vaginal fluid with several drops of saline and examining immediately under a 10 X or 40 X power light microscope. If examination is delayed, the trichomonads become nonmotile and may be confused with white blood cells. Douching within the past 24 hours also may decrease sensitivity. When the wet mount is compared with culture in modified Diamonds media, sensitivity is usually around 50% to 60%.
Culture is sensitive and specific, but very labor intensive as specimens must be examined daily for 5 days before a negative result can be reported. A pouch method for T vaginalis has been developed (InPouch TV Culture System, Biomed Diagnostics, Santa Clara, California) that allows delay in examination. A microscope slide is built into the side of the pouch, permitting examination without opening and removing a sample. It requires 3 days of incubation and microscopic examination and has a sensitivity of 88.25%, similar to that of culture.14 A probe method (Affirm VP Microbial Identification Test, MicroProbe Corporation, Bothell, Washington) has been marketed for the detection of T vaginalis and G vaginalis. Processing time is 30 minutes. The sensitivity and specificity for T vaginalis have been reported as 83% and 100%, respectively, higher than a wet mount but not as good as culture.15
1. Jenny C. Child sexual abuse. In: Holmes KK, Mardh PA1 Sparling PT, Wiesner PJ, eds. Sexually Transmitted Diseases. 2nd ed. New York, NY: McGraw Hill Ine; 1990:895-900.
2. Whinington BA, Rice RJ, Biddle JW, Knapp JS. Incorrect identification of Neisseria gonorrhoeae from inânes and children. Pediatr inject Dis J. 1988;7:3-10.
3. Alexander ER, Harrison HR. Role of Chlamydia trachomatis infection in perinatal infection. Review of Infectious Diseases m Gynecology. 1983;5:713-719.
4. Centers for Disease Control and Prevention. Recommendations for the prevention and management of Chlamydia trachomatis infections. MMWR. 1993;42(RR-12).
5. Hammerschlag MR, Roblin PM, Gelling M, Worku M. Comparison of two enzyme immunoassays to culture for the diagnosis of chlamydial conjunctivitis and respiratory infections in infants. Journal of Clinical Microbiology. 1990;28:1725-1727.
6. Cohen P, Stout G, Ende N. Serologic reactivity in consecutive patients admitted to a general hospital. Arch Intern Med. 1969;124:364-367.
7. Centers for Disease Control and Prevention. Syphilis. In: Sexually Transmitted Diseases Clinical Practice Guidelines 1991. Washington, DC: US Government Printing Office; 199136-41. US Dept of Health and Human Services.
8. Rawstron SA, Bromberg K. Comparison of maternal and newborn serologic tests for syphilis. AmJ Dis Child. 1991;145:1383-1388.
9. Jackson JB, Balfour HH Jr. Practical diagnostic testing for human immunodeficiency virus. CIm Microbiol Rev. 1988;1:122-138.
10. Hou X, Bréese PL, Douglas JM. Utility of quantitative enzyme immunoassay reactivity for predicting human immunodeficiency virus seropositiviry in low and high prevalence populations. J Clin Microbiol. 199432:220-223.
11. Centers for Disease Control and Prevention. Sexually Transmitted Diseases Clinical Procace Guidelines. Washington, DC: US Government Printing Office; 1991:21-25. US Dept of Health and Human Services.
12. Levin MJ. Genital herpes simplex. In: Wentworth BB, Judson FN, Gilchrist MJR, eds. Laboratory Methods for the Diagnosis of Sexually Transmitted Diseases. 2nd ed. Washington, DC: American Public Health Association; 1991:128-164.
13. Lorincz AT. Human papilloma virus detection tests. In: Holmes KK, Mardh PA, Sparling PF, et al, eds. Sexually Transmitted Diseases. 2nd ed. New York, NY: McGraw Hill Ine; 1990:953-959.
14. Draper D, Parker R, Patterson E, et al. Detection of Trichomonas vaginalis in pregnant women with the InPouch TV Culture system. J CIm Microbiol. 1993;31:1016-1018.
15. BisekJen AM, Hillier SL. Evaluation of Affirm VP Microbial Identification Test for Gordnerelia vaginalis and Trichomonas vaginalis. J CUn Microbiol. 1994;32:148-152.
Recommended Sites for the Recovery of Neisseria gonorrhoeae and Chlamydia trachomatis in Neonates and Children*
Sensitivity and Specificity of Diagnostic Tests for Sexually Transmitted Infections*