Literature review > Issue_6 > Review on Palusci et al. 

The commentary for this article is a letter to the editor published in The Pediatric Infectious Disease Journal, (C) 2003 Lippincott Williams & Wilkins, Inc., Volume 22(11), November 2003, pp 1028-1029, and used by permission of the publishers. The authors' reply (pp 1029-1030) is also included.

To The Editors:

Laboratory testing for sexually transmitted disease in prepubertal children being evaluated for suspected sexual abuse is a complicated issue, because in addition to medical implications, the results of testing also have legal implications. Thus only tests with the highest specificity should be used [1]. One cannot extrapolate the performance of a test in urethral and endocervical sites in adults to genital sites in children, especially the vagina.

The sequential testing procedure for Neisseria gonorrhoeae recently proposed by Palusci and Reeves [2] using a nonculture test followed by culture, is unfortunately based on several erroneous assumptions. First the authors state that the sensitivity of culture for detection of N. gonorrhoeae from a genital site (not specified) ranges from 40 to 80%. A reference in their article, The 2002 Sexually Transmitted Disease Treatment Guidelines from the Centers for DiseaseControl and Prevention (CDC) [1] does not contain any statements about the relative sensitivity of culture of N. gonorrhoeae. A review article cited by the authors actually states, "The sensitivity of culture, is nevertheless, regarded as approaching 95-100% in clinics and laboratories where there is experience in taking suitable specimens and expertise in isolation and identification procedures" [3]. Certainly there are no data on the relative sensitivity of culture for detection of N. gonorrhoeae in vaginal specimens. PACE 2 is a DNA hybridization test, not an amplification test, and numerous studies in adults have found sensitivities of 90 to 94% compared with culture when used for genital sites in adults [4, 5]. The 2002 CDC Guidelines on laboratory testing for Chlamydia trachomatis and N. gonorrhoeae recommend that DNA probe tests for N. gonorrhoeae only be used as screening tests for endocervical/urethral swabs from adults [6]. It was not recommended for any other anatomic site, including the vagina. It is rather surprising that the authors chose a nonculture test thought to be insufficiently sensitive for vaginal specimens in adults for use in vaginal specimens from children.

Finally the authors reveal what appears to be a lack of familiarity with the recommendations for N. gonorrhoeae testing of prepubertal children in the 2002 CDC Guidelines. In the discussion they state that the AAP and CDC stated that only cultures be used and "the use of nonculture methods such as DNA probes or enzyme-linked immunosorbent assays (EIAs) for N. gonorrhoeae is considered investigational." Actually the guidelines state that only standard culture methods be used; there is no reference to DNA probes or EIAs. EIAs for detection of N. gonorrhoeae, such as Gonozyme, which was manufactured by Abbott Diagnostics, were briefly on the market in the 1980s. They were withdrawn because of very poor performance with sensitivities averaging 50% compared with culture [7]. Even so, DNA probes are also old technology. As of this writing there are three Food and Drug Administration-approved nucleic acid amplification tests (NAATs) for N. gonorrhoeae and C. trachomatis on the market: PCR (Amplicor; Roche); strand displacement amplification (ProbeTec; BD Biosciences); and transcription-mediated amplification (Gen-Probe). A fourth assay, ligase chain reaction (LCX; Abbott), which the authors refer to in the discussion, was withdrawn from the market last year because of significant performance problems [8]. However, whereas NAATs may be anywhere from 10 to 30% more sensitive than culture for detection of C. trachomatis, they do not have a significant advantage over culture for detection of N. gonorrhoeae [6]. The major advantage they have relates to specimen handling given that one does not have to be as concerned about preserving viability, but one will not have an isolate available for further testing, which could be of forensic importance in the setting of child sexual abuse. The CDC did suggest that in rare circumstances a NAAT could be used for detection of N. gonorrhoeae in children, provided the result be confirmed with a second NAAT [1].

Although most genital gonococcal infections in children are symptomatic, it is not 100%; asymptomatic infection does occur. Pharyngeal and rectal infections are more likely to be asymptomatic, and as mentioned above, none of the currently available NAATs or PACE 2 is approved for use at those sites. Furthermore the results of several large studies of sexually transmitted diseases in prepubertal children have found rates of gonococcal infection to be ~1% [9]. Thus even with a sensitivity of 90%, the positive predictive value of PACE 2 may be only 70%, which is not acceptable in a forensic situation. In conclusion it is possible that NAATs may have a role in diagnosis of gonorrhea in children under very specific circumstances when confirmation is available. Use of a two step testing procedure with a relatively insensitive assay as proposed by Palusci and Reeves [2] without confirmation of positive results is not appropriate at this time.

References:

1. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines 2002. MMWR 2002; 51 (RR-6):1-80.

2. Palusci VJ, Reeves MJ. Testing for genital gonorrhea infections in prepubertal girls with suspected sexual abuse. Pediatr Infect Dis J 2003; 22: 618-23.

3. Ison CA. Laboratory methods in genitourinary medicine: methods of diagnosing gonorrhoea. Genitourin Med 1990; 66: 453-9.

4. Schwebke JR, Zajackowski ME. Comparison of DNA probe (Gen-Probe) with culture for the detection of Neisseria gonorrhoeae in an urban STD programme. Genitourin Med 1996; 72: 108-10.

5. Ciemans EL, Borenstein LA, Dyer IE, et al. Comparisons of costs and accuracy of DNA probe test and culture for the detection of Neisseria gonorrhoeae in patients attending public sexually transmitted disease clinics in Los Angeles County. Sex Transm Dis 1997; 24: 422-8.

6. Centers for Disease Control and Prevention. Screening tests to detect Chlamydia trachomatis and Neisseria gonorrhoeae infections: 2002. MMWR 2002; 51(RR-15):1-38.

7. Thomason JL, Gelbart SM, Sobieski VJ, et al. Effectiveness of Gonozyme for detection of gonorrhea in low-risk pregnant and gynecologic populations. Sex Transm Dis 1989; 16: 28-31.

8. Centers for Disease Control and Prevention. Recall of LCx Neisseria gonorrhoeae assay and implications for N. gonorrhoeae and Chlamydia trachomatis testing. MMWR 2002; 51: 709.

9. Ingram DL, Miller WC, Schoenbach VJ, et al. Risk assessment for gonococcal and Chlamydia infections in young children undergoing evaluation for sexual abuse. Pediatrics 2001; 107: e73.

Authors' Reply:

We appreciate the opportunity to respond to Dr. Hammerschlag's comments about our article concerning testing for genital gonorrhea infections (GC) in prepubertal girls with suspected sexual abuse [1]. We agree that laboratory testing for sexually transmitted diseases in children with suspected abuse is complicated and that the medical and legal issues indicate the need for accuracy, both to protect the child from further abuse and to prevent the negative consequences of falsely identifying GC. Dr. Hammerschlag raises important questions about what we do and do not know about GC diagnosis in young children.

We also agree that it is important to be aware of the potential pitfalls when extrapolating test information from adults to children. Unfortunately, as we stated, the lack of available data specific to children forces one to make such extrapolations. We made certain assumptions about the sensitivity and specificity of culture and nonculture techniques with the goal of identifying which testing strategy would result in the best health outcomes. We noted these and other potential limitations in our paper. Clinical decision analysis explicitly uses test characteristics and other factors to arrive at a strategy that results in optimal health utilities or outcomes. Although Dr. Hammerschlag suggests that GC culture can attain sensitivity as high as 95 to 100%, this may not be achieved in clinical practice, and our sensitivity analyses showed that sequential testing with a nonculture test followed by culture confirmation was optimal over a wide range of culture sensitivity. Culture alone was preferable only when the test characteristics of nonculture tests were set below those of values for culture in our model. Culture, which may have high sensitivity when optimally performed, will be less useful in clinical practice under real world conditions (improper plating, delays in transport, inappropriate media or incubation), and we need to assess the potential impact of several strategies given the cost and actual implementation of current guidelines.

Dr. Hammerschlag notes that the CDC recommends that only standard culture methods be used, then later adds that the CDC did suggest that in rare circumstances a nucleic acid amplification test, a nonculture test, could be used with confirmation. The CDC 2002 Guidelines recommend that "all presumptive isolates of N. gonorrhoeae should be confirmed by at least two tests that involve different principles (i.e. biochemical, enzyme substrate, serologic, or DNA probe methods)" [2]. Our clinical decision analysis agrees with that strategy. We used a DNA test as an example of a nonculture test in our model because DNA probes were recommended as a definitive test for GC by the American Academy of Pediatrics [3]. Although recent AAP recommendations note that only culture using standard confirmation methods should be used and DNA probes should not be used as a diagnostic method [4], our results are generalizable to any other nonculture tests that have similar test characteristics, and the conclusions would not change.

We are unclear why Dr. Hammerschlag discusses rates of asymptomatic genital GC infection and testing of nongenital sites. We limited our analyses to symptomatic (discharge) genital infections in prepubertal girls based on published reports. A lower GC prevalence (1%) in asymptomatic children with suspected sexual abuse would decrease the positive predictive value of nonculture tests as Dr. Hammerschlag describes, but it would also affect culture positive predictive value in the same manner. Our sensitivity analyses address this very point directly, showing that sequential testing was the preferred strategy for all GC prevalence rates of 54% or less. Testing strategies for nongenital infections could be assessed in a new decision analysis with likely different assumptions.

Dr. Hammerschlag concludes that a testing procedure without confirmation of positive results is not appropriate at this time. Our decision analysis is absolutely consistent with this recommendation, because the results confirm that a two step testing strategy with a nonculture test followed by culture confirmation is optimal when compared with culture or nonculture alone. Fewer children with GC are missed. Undoubtedly progress in laboratory methods will result in improvements in test sensitivity and specificity over time, and clinical decision analysis offers a useful methodology for assessing the outcomes of those changes.

We recognize Dr. Hammerschlag's important contributions to the field and hope that she agrees with us that it is preferable to strive to not miss children with genital GC given the detrimental effects of ongoing child sexual abuse. We also recognize the need for financial support for research to answer the important questions raised by sexually transmitted diseases in children. Given current knowledge in the field, our analysis shows that a sequential testing strategy maximizes identification of genital GC with no false positive results, and this conclusion was not sensitive to plausible changes in a wide range of underlying conditions.

Vincent J. Palusci, M.D., M.S.
Mathew J. Reeves, B.V.Sc., Ph.D.
DeVos Children's Hospital
Grand Rapids, MI (VJP)
Department of Epidemiology
Michigan State University
East Lansing, MI (MJR)

References:

1. Palusci VJ, Reeves MJ. Testing for genital gonorrhea infections in prepubertal girls with suspected sexual abuse. Pediatr Infect Dis J 2003; 22: 618-23.

2. Centers for Disease Control and Prevention. Sexually transmitted disease treatment guidelines. MMWR 2002; 51 (RR-6):1-80.

3. American Academy of Pediatrics, Committee on Child Abuse and Neglect.Guidelines for the evaluation of sexual abuse of children: subject review. Pediatrics 1999; 103: 186-91.

4. American Academy of Pediatrics, Committee on Infectious Diseases Red Book.2003 Report of the Committee on Infectious Diseases. 26th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2003: 162.

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001