Bacterial culture is the reference standard for the diagnosis of Neisseria gonorrhoeae infection and provides useful data on antimicrobial susceptibility. Strict transport and growth conditions, as well as a 48-hour bacterial growth time, can limit the sensitivity and value of culture. Nucleic acid amplification techniques are therefore increasingly used in diagnosis.
The BDProbeTec-strand displacement amplification (SDA) system (Becton Dickinson Microbiology Systems, Sparks, Md.) uses fluorescent probes to detect products in solution. It has been found to have sensitivity, specificity, positive predictive value, and negative predictive value in urine samples of 99.2%, 99.3%, 84.9% and 99.9%, respectively, and sensitivity and specificity of 100% each from genital swabs (1, 4). Our own laboratory has previously found similarly high sensitivity and specificity of N. gonorrhoeae SDA on endocervical swabs from females and first-void urines (FVU) from males (98.1% and 100%, respectively) when compared to N. gonorrhoeae culture (S. M. McIntyre, unpublished data). Samples with a method other than acceleration (MOTA) value of ≥2,000 and an amplification control MOTA of ≥1,000 are considered positive. A previous study found that, among SDA-positive culture-negative samples, 50% had “low positive” MOTA values (2,000 to 9,999) (1). Becton Dickinson recommends that, although all samples with a MOTA of ≥2,000 should be reported as positive, there is an increased likelihood of false positivity with scores between 2,000 and 9,999. MOTA values, however, are not quantitative and are not indicative of the sample organism load but are a product of both the amount and rate at which amplification occurs (2). A previous study has reported the lower reproducibility of samples with MOTA values between 2,000 and 9,999 (33.3%) compared to those with a MOTA value of ≥10,000, although the authors did not correlate these findings with N. gonorrhoeae culture (3).
The prevalence of N. gonorrhoeae in the population attending our genitourinary services has previously been found to be 3.4% (McIntyre, unpublished). Our laboratory has established a protocol for the use of the BDProbeTec-SDA system for N. gonorrhoeae screening on FVU from male patients and endocervical swabs from female patients. All SDA-positive samples are retested. Patients with risk factors for N. gonorrhoeae also have a culture swab taken at their first visit and inoculated onto Philips agar. In those who are SDA positive and did not have a swab taken, one is taken at recall prior to treatment. This has significantly reduced both the time to reporting negative results and the number of cultures processed.
We conducted a retrospective evaluation to establish the frequency of low positive MOTA values on initial screening, as well as the reproducibility and culture concordance of low positive samples compared to high positive MOTA (≥10,000) values. During a 9-month period, 30/270 (11.11%) SDA-positive samples had a low positive MOTA. Only 13/30 (43.34%) low positive samples were reproducible compared to all 240 (100%) high positive MOTA samples.
Culture results were not available for eight samples with high positive MOTA scores and five samples with low positive scores. Of the remaining high positive samples, 218/232 (93.97%) were culture positive. In comparison, only 5/25 (20%) of the low positive samples were culture positive. Two patients with low positive scores and negative cultures were found to be culture positive at another site (pharyngeal), and two patients were culture positive a few days earlier. If these are reclassified, 9/25 (36%) of low positive samples may be considered true infections. Of these low positive samples, 8/13 (61.54%) of repeatable positive samples were culture positive or considered true infections, compared to only 1/12 (8.34%) of nonrepeatable samples.
A total of 30/257 samples were SDA-culture discordant. Of the 14 patients with high positive scores and negative cultures, 6 were contacts of N. gonorrhoeae-confirmed patients, 2 had microscopy consistent with N. gonorrhoeae, and 4 had cultures taken after antimicrobial therapy. Of the 16 patients with low positive scores and negative cultures, 1 had microscopy consistent with N. gonorrhoeae and 1 had previously confirmed infection.
Our results show that low positive MOTA samples are less likely to be positive on repeat testing compared with high positive samples. The rate of culture discordance was also greater if samples had low positive scores, although a repeatable result improved the likelihood of culture concordance considerably. The SDA-culture-discrepant results may be due to contamination from other samples or the environment. Alternatively, SDA may be more sensitive than conventional culture at detecting a fastidious organism. Our findings that several of our patients with SDA-culture-discrepant samples had N. gonorrhoeae isolated from another site, had culture-positive samples a few days previously, had microscopy consistent with N. gonorrhoeae, or were contacts of patients with proven N. gonorrhoeae infection may support this suggestion. Although the reproducibility of samples with a high positive MOTA value in the samples included in our present study was 100%, we would be guarded against recommending that only low positive MOTA samples be repeated. In our experience there are occasionally samples (presumably due to contamination) with high positive MOTA values on initial screen but <2,000 on subsequent testing. This was also reported in one previous study (3). Given the potential adverse medical, social, and psychological impact of a false-positive result, we would therefore recommend that all samples positive on initial screen be repeated. Based on our findings above, we would also suggest that those that are reproducible be reported as N. gonorrhoeae SDA positive regardless of whether they have high or low positive MOTA scores. One could also add a comment to those samples reported as N. gonorrhoeae SDA positive with repeatable low positive MOTA values (and negative N. gonorrhoeae cultures) to alert the clinician to the situation, so a well-informed treatment decision can be made based on both laboratory findings and clinical symptoms and risk factors. Samples with discrepant MOTA values on repeat testing could also be tested a third time and a consensus result decided upon as has been described previously (3). This approach has now been adapted in our own laboratory.
- Copyright © 2006 American Society for Microbiology
