ABSTRACT
Screening assays for Neisseria gonorrhoeae exhibit low positive predictive values, particularly in low-prevalence populations. A new real-time PCR assay that detects and identifies individual Neisseria spp. using melt curve analysis was compared to two previously published supplementary assays. NsppID, a 16S rRNA real-time PCR/melt curve assay developed to distinguish N. gonorrhoeae from other Neisseria spp., was compared to real-time PCR assays targeting genes reportedly specific for N. gonorrhoeae, the cppB gene and the porA pseudogene. A total of 408 clinical specimens (324 female endocervical swabs and 84 male urine or urogenital swab specimens) were screened using the COBAS Amplicor assay for Chlamydia trachomatis and N. gonorrhoeae (CT/NG) (Roche Diagnostics, Indianapolis, IN) followed by confirmatory testing via real-time PCR. The NsppID assay detected Neisseria spp. in 150/181 COBAS-positive specimens (82%), including six dual infections, and identified N. gonorrhoeae in 102 (56%) specimens. Sixty-nine of 181 (38%) specimens were positive for N. gonorrhoeae by porA pseudogene, and 115/181 (64%) were positive for cppB. However, cppB was also positive in 15% of COBAS-negative specimens, more than either NsppID (4%) or porA pseudogene (2%) assays. The porA pseudogene assay had the highest specificity for both genders but the lowest sensitivity, especially in female specimens. NsppID had a slightly lower specificity but greater sensitivity and overall accuracy. The least desirable confirmatory assay was cppB, due to poor specificity. The NsppID assay is an accurate confirmatory assay for N. gonorrhoeae detection. In addition, the NsppID assay can identify the non-N. gonorrhoeae species responsible for the majority of false-positive results from the COBAS Amplicor CT/NG assay.
Neisseria gonorrhoeae is the second most prevalent sexually transmitted bacterium after Chlamydia trachomatis (4). Many N. gonorrhoeae infections are asymptomatic, especially in women, and if left untreated can develop long-term consequences, including chronic pelvic pain, pelvic inflammatory disease, and ascending infection of the fallopian tubes. Accurate diagnosis of symptomatic and asymptomatic infection is required to prevent these complications and to control the transmission of infection.
Molecular assays such as the COBAS Amplicor CT/NG have high sensitivity and specificity in detecting both C. trachomatis and N. gonorrhoeae infections, and screening of high-prevalence populations results in positive predictive values (PPVs) of >80%. However, even with very specific assays, screening of low-prevalence populations lowers the positive predictive value of any test. For N. gonorrhoeae analysis of low-prevalence populations performed by the COBAS Amplicor CT/NG assay (6, 12, 32), unacceptable positive predictive values have been reported, particularly when the optical density (OD) reading is in the gray-zone of ≥0.2 to <3.5 (6). Repeat testing of the gray-zone and positive specimens has been shown to improve the specificity of the COBAS Amplicor CT/NG assay to some extent (24). Still, several cross-reacting nonpathogenic organisms, primarily other Neisseria spp., with homologous sequences are known to contribute to false-positive results (6, 7, 13, 24, 25). As recommended by the Centers for Disease Control and Prevention (5), confirmation of N. gonorrhoeae infection using independent testing for a different N. gonorrhoeae-specific target greatly improves the positive predictive value (14).
Other real-time PCR assays have been developed that target genes reportedly specific for N. gonorrhoeae, with confirmation frequencies of <60% of the presumptive N. gonorrhoeae-positive specimens (22, 29, 31). Subsequent studies indicated that several real-time confirmatory assays may lack the specificity and/or sensitivity required for accurate diagnosis (3, 23). Also, none of these molecular assays identify the cross-reacting species responsible for false-positive results with the COBAS Amplicor CT/NG assay. Thus, this study describes a newly developed assay, NsppID, that not only detects and identifies N. gonorrhoeae but also identifies cross-reacting Neisseria spp., in an effort to ultimately reduce the number of false positives, and compares NsppID to two previously published confirmatory assays (21, 29).
(This work was presented at the Annual Meeting of the Association for Molecular Pathology, Scottsdale, AZ, 10 to 13 November 2005.)
MATERIALS AND METHODS
Specimens.Genomic DNA samples from Neisseria spp. were obtained either directly from American Type Culture Collection (ATCC [Manassas, VA]) or extracted in house from strains obtained from ATCC. Genomic DNA from seven N. gonorrhoeae strains (ATCC 53420, 53421 53422, 53423, 53424, 53425, and 700825), six Neisseria meningitidis strains (ATCC 53414, 53415, 53416, 53417, 53418, and BAA-335), and one strain each of Neisseria cinerea (ATCC 14685), Neisseria lactamica (ATCC 23970), Neisseria perflava (ATCC 14799), and Neisseria subflava (ATCC 49275) were used in assay development and as controls during analytical assay validation of NsppID, a 16S rRNA PCR/melt curve assay we developed that distinguishes N. gonorrhoeae from other Neisseria spp. Samples remaining after COBAS Amplicor CT/NG testing of 408 deidentified clinical specimens (324 endocervical swabs and 84 male urine or urogenital swab samples) were used according to the protocol approved by the Evanston Northwestern Healthcare (ENH) Institutional Review Board.
COBAS Amplicor testing.Clinical specimens were screened using the COBAS Amplicor CT/NG assay (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's instructions. Specimens were considered negative if the initial OD readings were <0.2 and the OD of the inhibition control was ≥0.2. Specimens with initial OD readings in the gray zone of ≥0.2 to <3.5 or positive readings of ≥3.5 were repeated twice more, and at least two of the three OD readings needed to be ≥2.0 before the specimen was reported as COBAS N. gonorrhoeae positive (24).
DNA extraction.Samples of genomic DNA from Neisseria strains were extracted in house using Puregene DNA extraction (Gentra Systems, Minneapolis, MN). Genomic DNA was extracted from 200 μl of the original clinical specimen using GENERATION DNA purification capture columns (Gentra Systems) and/or High Pure viral nucleic acid (Roche Diagnostics) according to the manufacturers' instructions. Some M4RT specimens received had a significant pH color change due to increased acidity; therefore, 10 μl of 0.5 M NaOH was added to all M4RT aliquots before extraction to ensure an alkaline sample and improve the recovery of the sample DNA.
Real-time PCR.Our laboratory includes geographically separate areas for reagent preparation, sample preparation, PCR amplification, and amplicon analysis to avoid carryover contamination. The real-time assays included negative, positive, and no-template controls in each run. The NsppID assay was developed and validated using genomic DNA extracted from ATCC controls, either individually or as combinations of N. gonorrhoeae with either N. meningitidis or N. perflava. Clinical validation included testing of the 408 clinical specimens to determine any Neisseria spp. present using the NsppID assay. These results were compared to the results from two previously published real-time PCR confirmatory assays targeting genes reportedly specific for N. gonorrhoeae: the cppB gene (21) and the porA pseudogene (pap) (29). All three assays were performed on the LightCycler instrument (Roche). The two previously published procedures were performed as described in the individual publications, and the NsppID assay is described below. Real-time PCR inhibition/detection failure was assessed via amplification of exon 8 of the human p53 gene (in-house assay) or universal bacterial 16S rRNA gene amplification.
NsppID real-time PCR assay.Consensus primers NG767F16 (5′-AAAgCgTgggTAgCAA-3′) and NG964R16 (5′-TTCTTCgCgTTgCATC-3′) (sequence numbering from GenBank accession no. X07714) were designed after comparing several partial 16S rRNA gene sequences for each of 10 Neisseria spp. The fluorescent resonance energy transfer (FRET) probe pair consisted of a fluorescein isothiocyanate (FITC)-labeled probe, NG839GR3 (5′-CAACCTGATTGCTTAGTAGCGTAGCTAACG-FITC-3′), that was specific for N. subflava, N. flavescens, and one strain of N. meningitidis but reliably discriminated between the different Neisseria spp. by mean temperature of melting (Tm), and an LC640-labeled anchor probe, NG870RED (5′-LC640-gTgAAATTgACCgCCTggggAgTACgg-phos-3′), common to most Neisseria spp. (Fig. 1).
Sequence comparisons between 10 Neisseria spp. encompassed by the FRET probes used in the NsppID assay. The underlined sequences are exact matches to the probe sequences and therefore have the highest theoretical Tm. The sequence numbering listed for the two probes is from GenBank accession no. X07714.
Each 20-μl reaction mix contained 5 μl of sample DNA, 2 μl of 10× FastSTART DNA hybridization probe reagent (Roche), 3 mM MgCl2 (final concentration), 0.2 mM NG767F16 primer, 1.0 mM NG964R16 primer, 0.2 mM NG839GR3 probe, and 0.4 mM NG2RED primer. Increased sensitivity was observed when the master mix was predigested with HpaII restriction enzyme (1 U per reaction; 37°C incubation for 10 min, followed by enzyme inactivation by heating to 65°C for 20 min) before addition of the specimen DNA. This restriction digest eliminated any genomic DNA in the master mix, including residual DNA found in commercial Taq polymerases, and has been shown to improve sensitivity in other bacterial 16S rRNA PCR assays (18; data not shown). PCR conditions consisted of an initial denaturation at 95°C for 10 min followed by amplification for 50 cycles of 10 s at 95°C, 15 s at 55°C, and 7 s at 72°C with fluorescence acquisition at the end of each extension step. Amplification was immediately followed by a melt program consisting of 15 s at 95°C, 15 s at 40°C, and a stepwise temperature increase of 0.2°C/s until 75°C with fluorescence acquisition at each temperature transition. The fluorescence data were analyzed using the F2/F1 setting. This assay for Neisseria sp. differentiation required approximately 1 h to complete, in addition to the sample preparation time. Melt curve analysis was used to determine the species-specific Tm based on values determined from the respective ATCC controls. Although we did not test a strain of N. flavescens, the sequence information obtained from GenBank would indicate that it would also be included with the group of N. subflava, N. perflava, and N. meningitidis.
Interassay comparison.Since no accurate “gold standard” was available, a consensus N. gonorrhoeae result was defined as positive when at least two of the three confirmatory assays yielded positive N. gonorrhoeae results. This imperfect standard approach has been used in other studies (2, 21, 26) when the ideal gold standard does not exist, with underestimation of the test's accuracy in most situations (33). The chi-square homogeneity (χ2H) test was used to determine if the distributions of N. gonorrhoeae-positive results were the same between the three confirmatory assays (8). Sensitivity, specificity, and the predictive values of both positive and negative results were calculated and compared using a large-sample binomial z test (8, 20); the accuracy of each confirmatory assay was also calculated (28).
RESULTS
The NsppID assay was developed as a confirmatory test to determine the presence of N. gonorrhoeae or other cross-reacting Neisseria spp. in clinical specimens that tested N. gonorrhoeae positive using the COBAS Amplicor CT/NG assay. In the analytical validation of the NsppID assay, the consensus primers were able to amplify 16S rRNA gene sequences from all ATCC isolates of N. gonorrhoeae, N. lactamica, N. cinerea, N. subflava, N. perflava, and N. meningitidis tested and FRET probes to a unique sequence between the primers allowed species differentiation using melt curve analysis (Fig. 2 and Table 1). The analytical sensitivity of the assay was calculated to be 4 genome copies of N. gonorrhoeae (10 fg DNA) per reaction for this multicopy gene target (data not shown). Reactions containing DNA from both N. gonorrhoeae and either N. meningitidis or N. perflava indicated that N. gonorrhoeae could still be detected in the melt curve analysis even in the presence of 40-fold concentrations of the non-N. gonorrhoeae species (data not shown). The average Tm in the melt curve analysis for each Neisseria sp. is as follows: N. gonorrhoeae, ∼58.0°C; N. lactamica, ∼51.5°C; N. cinerea, ∼55.0°C; N. meningitidis, ∼62.5°C or ∼64.5°C; N. perflava, ∼62.5°C; and N. subflava, ∼62.5°C.
Melt curve analysis in the NsppID assay using the ATCC controls, showing species-specific Tm peaks.
NsppID assay results from 408 patient specimens using DNA extracted either by Gentra GENERATION capture column or by Roche High Pure viral methodsa
Clinical validation was performed on a total of 408 ENH patient specimens screened using the COBAS Amplicor CT/NG assay, yielding 181 N. gonorrhoeae-positive specimens and 227 randomly selected N. gonorrhoeae-negative specimens for this study. All 408 specimens were positive for human DNA and/or universal bacterial gene amplification, indicating no substantial PCR inhibition (data not shown).
The results of testing all 408 specimens using the NsppID assay are shown in Table 1. The NsppID assay was able to detect a Neisseria spp. in 150 of the 181 COBAS-positive specimens (83%) but identified N. gonorrhoeae in only 102 (56%). Thirty-one specimens did not amplify detectable amplicons for any Neisseria spp.; this could indicate a lack of sensitivity, problems with real-time PCR, and fluorescent detection from M4 media, or it may also reflect cross-reacting organisms present in the COBAS-positive specimens other than Neisseria spp., such as Lactobacillus (25). Dual infections were identified in six specimens, with N. gonorrhoeae being one of the infectious agents in four of the six. Of the 111 clinical specimens that were positive for N. gonorrhoeae in the NsppID assay, 90 (81%) had initial OD readings of ≥3.5, 20 (18%) had readings in the gray zone of ≥0.2 to <3.5, and only 1 specimen (1%) had an initial reading of <0.2 in the COBAS Amplicor assay.
Comparisons between the NsppID, cppB, and pap assays are shown in Table 2. The numbers of N. gonorrhoeae-positive specimens among the 181 COBAS-positive specimens according to the other two confirmatory tests were 69 (38%) positive for pap and 115 (64%) positive for cppB. However, cppB was also positive in a significant number (15%) of COBAS-negative specimens as compared to either the NsppID (4.0%) or pap (2%). This may be due to the detection of non-N. gonorrhoeae spp. carrying the cryptic plasmid pJD1 (2, 17, 29); 13% of cppB positives yielded non-N. gonorrhoeae spp. in the NsspID assay. Since the NsppID assay detects and identifies other Neisseria spp. in addition to N. gonorrhoeae, while the other two confirmatory assays only detect N. gonorrhoeae, the percentage of COBAS-positive specimens with no resolution is only 17% using the NsppID assay as compared to 37% and 61% for the cppB and pap assays, respectively.
Results from 408 patient specimens screened using the COBAS Amplicor CT/NG assay and the three N. gonorrhoeae confirmatory assaysa
Using positive results in at least two of the three confirmatory assays as a consensus N. gonorrhoeae-positive specimen, we can determine the sensitivity and specificity for the detection of N. gonorrhoeae in all four PCR assays. Comparisons between the COBAS Amplicor, NsppID, cppB, and pap assays are shown in Table 3. Of the 181 COBAS-positive specimens, 95 were also consensus N. gonorrhoeae positive, and the percentage concordance differed along gender lines: 45 of 61 male specimens (74%) were concordant, whereas only 51 of 120 female specimens (43%) agreed. These results are in agreement with those of other published studies analyzing COBAS Amplicor detection of N. gonorrhoeae in low-prevalence populations (6, 14, 21, 32) and highlight the need for a confirmatory test to ensure accurate reporting of infections. Six of the 227 COBAS-negative specimens (3%) were consensus N. gonorrhoeae positive—all from female specimens. Five of the six were in the gray zone of the COBAS Amplicor assay on the initial run but were categorized as negative after repeat testing. Although 3% is approximately the false-negative rate reported in the COBAS Amplicor CT/NG product information, it is higher than the prevalence rate of 0.6% in our population.
Results from 408 patient specimens screened using the consensus result of at least two of the three confirmatory assays as the gold standarda
Statistical analysis indicates a significant difference between the NsppID and pap assays, with the NsppID assay being more sensitive and less specific than pap (P < 0.001, χ2Η = 33.8, df = 1); the cppB assay had significantly lower sensitivity and specificity than the other two assays (P < 0.01, χ2Η = 10.69, df = 1). Limiting the analysis to only those specimens that were N. gonorrhoeae positive by COBAS Amplicor testing, both NsppID and pap assays identified N. gonorrhoeae more accurately than the cppB assay, with higher sensitivity and specificity (P < 0.001, χ2Η = 12.558, df = 1), but the NsppID assay did not significantly exceed the pap assay in accuracy (P > 0.1, χ2Η = 2.032, df = 1).
Statistical assessment of clinical assays includes common parameters of sensitivity, specificity, PPVs and negative predictive values (NPVs), and accuracy, as shown in Table 4. The key statistic for confirmatory assays after an initial screen is PPV. The pap assay has the highest PPV for both genders (97% in female specimens and 100% in male specimens), followed closely by NsppID (90% in female specimens and 97.8% in male specimens). However, the sensitivity of the pap assay is significantly lower than that of NsppID (P < 0.001), particularly in female specimens (59% versus 92% in female specimens and 87% versus 100% in male specimens, respectively). Therefore the accuracy of the pap assay is lower (overall, 85% versus 92%). An added feature not represented in these statistics is that the NsppID assay is capable of identifying non-N. gonorrhoeae species responsible for false-positive results of the COBAS Amplicor testing, reducing the number of specimens with unresolved COBAS-positive results. Although the sensitivity of the cppB assay is higher than that of the pap assay, the lack of specificity of the cppB assay makes it significantly lower in PPV, NPV, and accuracy than the NsppID assay and lower in PPV than the pap assay (P < 0.001).
Sensitivity, specificity, PPV, NPV, and accuracy of the three confirmatory assays in testing COBAS Amplicor N. gonorrhoeae-positive specimens
DISCUSSION
The COBAS Amplicor assay targets the N. gonorrhoeae cytosine methyltransferase gene. This assay has been shown to be a reliable target in a widely used screening assay in asymptomatic patients, as NPVs greater than 99% occur with assays with sensitivities greater than 98%, regardless of the prevalence (13, 14, 25). However, using this and similar assays in low-prevalence populations significantly lowers the PPV of N. gonorrhoeae analysis below satisfactory levels, requiring the application of a second assay to confirm the existence of N. gonorrhoeae in patient specimens (12). The COBAS Amplicor CT/NG assay used in the initial screen and the three confirmatory assays used in this study are all nucleic acid amplification tests, but each of the four has a different genetic target reportedly specific for N. gonorrhoeae and can be considered an independent assay, fulfilling the CDC requirement for a second independent test for confirmation (5). Since this study was designed to find the best N. gonorrhoeae confirmatory assay after screening with COBAS Amplicor, the gold standard for designating N. gonorrhoeae-positive specimens was positivity for N. gonorrhoeae in at least two of the three confirmatory tests. Generally, bacterial isolation using selective culture media would be used as the gold standard. However, successful culture of N. gonorrhoeae requires stringent collection, transport, and storage of clinical specimens and often produces false-negative results (1, 14, 16, 27). Under optimal laboratory conditions, the sensitivity of culture is 85 to 95% of acute N. gonorrhoeae infections (19) but drops to 50% of chronically infected females (1). In a high-prevalence population (6.6% of women and 20.1% of men), the sensitivities of culture were 84.8% for endocervical specimens, 92.7% from symptomatic male urethral specimens, and 46.2% from asymptomatic male urethral specimens (16). Additionally, the sensitivity of culture for N. gonorrhoeae has been reported as 88.9% for on-site collection/testing compared with 77.8% for off-site testing (11). Due to these and other limitations, culture results were not included in this study.
We encountered unexpected difficulty with amplification of some samples, though whether this was due to PCR inhibition, interference with fluorescence detection, or other issues remains unclear. Significant effort went into determining the best DNA extraction method to use with M4RT transport medium, the most common specimen received for clinical C. trachomatis/N. gonorrhoeae testing in our laboratory. In our hands, the most successful extraction used was High Pure viral nucleic acid extraction (Roche), followed closely by GENERATION capture column extraction (Gentra). Recently, the EZ1 virus kit (QIAGEN, Valencia, CA) has also been used successfully (data not shown). All 408 specimens analyzed in this study showed successful amplification of human and/or universal bacterial genes. However, endocervical samples, even by the most successful extraction method used, have a lower positivity rate for identification of N. gonorrhoeae or other species than urine or urethral samples.
We developed a real-time PCR assay (NsppID) that could identify N. gonorrhoeae as well as at least five other Neisseria spp. using genomic DNA from ATCC as controls, in order to confirm positive N. gonorrhoeae specimens and identify COBAS Amplicor false positives due to cross-reacting Neisseria spp. Repeated testing of seven strains of N. gonorrhoeae and four other Neisseria spp. resulted in distinct melting curve Tm peaks for N. gonorrhoeae and two other non-N. gonorrhoeae Neisseria spp. (N. cinerea and N. lactamica) and a single overlapping peak for three other non-N. gonorrhoeae Neisseria spp. (N. meningitidis, N. perflava, and N. subflava). Dual infections were also detected using the NsppID assay, with the capability of detecting N. gonorrhoeae even when the non-N. gonorrhoeae Neisseria spp. was 40× the concentration of N. gonorrhoeae. Therefore, the detection of non-N. gonorrhoeae Neisseria spp. in COBAS Amplicor N. gonorrhoeae-positive specimens without detection of an N. gonorrhoeae peak would indicate the positive result in COBAS was most likely due to a cross-reacting Neisseria spp. If no amplification occurred in the NsppID assay, either N. gonorrhoeae was present but below the sensitivity of the NsppID assay, or an organism besides Neisseria is cross-reacting in the COBAS Amplicor assay to produce a false N. gonorrhoeae-positive result. Some species of Lactobacillus have been reported as causing false N. gonorrhoeae-positive readings in the COBAS Amplicor CT/NG assay (25). Based on sequence comparisons, it is highly unlikely that any Lactobacillus spp. would be detected in the NsppID assay presented in this report and is a possible cause for lack of amplification from specimens positive for N. gonorrhoeae after COBAS Amplicor CT/NG screening that are negative using the NsppID assay.
Several other real-time N. gonorrhoeae confirmatory assays have been published (22, 29, 31). The cppB gene has frequently been used as an N. gonorrhoeae-specific target in PCR assays (10), but the specificity for N. gonorrhoeae of the cppB gene has recently been challenged (2, 3, 15, 17, 23, 29). Although found on the cryptic plasmid pJD1, which is often at higher copy numbers than chromosomal genes, there are some N. gonorrhoeae strains that are cppB negative by PCR due to lack of the plasmid or chromosomal integration (2, 3, 15, 17, 23). Adding to its lack of specificity, the cppB gene has also been identified in Neisseria spp. other than N. gonorrhoeae as well (2, 17, 29). The pap assay that detects the porA pseudogene has reportedly a greater specificity for N. gonorrhoeae but slightly lower analytical sensitivity than the cppB assay (29, 30). In this study, we compared the NsppID assay with these two previously published real-time PCR confirmatory assays; studies of real-time confirmatory assays based on a third target that is also used commercially for N. gonorrhoeae detection (the opa gene) were recently published but not included in this comparison (9, 22, 29, 30).
The application of any of the three confirmatory assays studied to the COBAS Amplicor N. gonorrhoeae-positive results significantly improves the PPV beyond that of COBAS Amplicor alone (from 53% to 73 to 99% overall). The differences in improvement were most evident in the female specimens. The PPV for female specimens tested by COBAS Amplicor alone was only 43%. The COBAS plus pap assay results from female specimens had the highest PPV at 97%, higher than 90% for NsppID (P = 0.0410) and dramatically higher than 62% found using the cppB assay (P < 0.001). Although the PPV for the pap assay is higher than those for either of the other confirmatory assays due to its specificity, the lower sensitivity of this assay resulted in negative results for over 40% of the consensus N. gonorrhoeae-positive female specimens, compared to less than 8% that were not detected using the NsppID assay. With the added feature of being capable of identifying non-N. gonorrhoeae species responsible for false-positive results of the COBAS Amplicor testing, the NsppID assay also reduced the number of specimens with unresolved COBAS-positive results.
Since the development and clinical validation of this test, our laboratory has utilized NsppID for confirmation of all repeatedly N. gonorrhoeae-positive COBAS Amplicor specimens. The specimen is reported as positive for N. gonorrhoeae if an N. gonorrhoeae-specific melt peak is observed and is reported as negative if a melt curve peak associated with a Neisseria sp. other than N. gonorrhoeae is detected without an accompanying N. gonorrhoeae peak. If no amplification is detected with the NsppID assay, the specimen is reported as equivocal, and a repeat specimen is recommended for N. gonorrhoeae testing by either molecular diagnostic analysis or microbial culture. Since implementing this strategy, we have tested over 15,000 specimens with over 200 specimens that were positive by the COBAS Amplicor assay. We are able to confirm 50% of the COBAS-positive N. gonorrhoeae results. Otherwise, findings are reported as negative in 17% of the specimens or reported as equivocal (33%) if the specimen does not amplify or clearly indicate in the NsppID assay the species present. The positivity rate has dropped from 0.9% to 0.3% of specimens submitted for N. gonorrhoeae testing from asymptomatic females and from 3.9% to 3.1% of specimens from symptomatic males, as predicted. Therefore, in our low-prevalence (0.6% overall) population, the COBAS Amplicor CT/NG screen plus the NsppID confirmatory assay have been used together to achieve high NPV and high PPV while greatly reducing the number of equivocal specimens that require retesting of the patient.
ACKNOWLEDGMENTS
This study was funded in part by a grant from Roche Diagnostics Corp., Indianapolis, IN, and the Department of Pathology & Laboratory Medicine of ENH.
We thank the technologists in the ENH Microbiology laboratory for their invaluable assistance in providing control strains of Neisseria spp. and the technologists in the ENH Molecular Diagnostics laboratory for clinical specimens and data.
FOOTNOTES
- Received 19 April 2006.
- Returned for modification 24 July 2006.
- Accepted 7 March 2007.
- Copyright © 2007 American Society for Microbiology
