Background In New Zealand it has been usual practise to detect Neisseria gonorrhoeae (NG) by culturing endocervical and urethral specimens obtained by pelvic examination. However there is a significant false negative rate. The use of newer nucleic acid amplification tests (NAATS) increases the detection of NG and allows testing of non-invasively collected samples. A large retrospective audit was performed on 18,913 microbial culture and cobas 4800 NG PCR results with the aim to determine if urogenital and non-genital specimens could be screened without the need for supplementary testing of positive results.

Methods Results from culture and PCR were compared; discrepancies were resolved by clinical correlation and/or an in-house assay targeting the opa gene and the porA psuedogene.

Results NG PCR diagnosed 33% more urogenital and 25% more rectal infections than culture; and testing of non-invasive specimens by PCR resulted in 37% more patients being screened for infection. Female urine is not suitable as a sole screening specimen by this assay as sensitivity was only 86.7%. There were insufficient pharyngeal or eye swabs available for the study to rule out the need for supplementary testing by additional DNA targets.

This study also showed an association between ‘failed’ cobas 4800 results and NG positive culture results, likely caused by mucopurulent discharge. Treating specimens with 1.4% Dithiothreitol enabled resolution of 89% of these specimens, of which 18% were positive for CT and/or NG.

In our population, 8% of NG positives were porAnegative, and 22% were opa negative. Confirmatory testing of a pharyngeal specimen identified a cross-reacting commensal Neisseria which gave a false positive cobas 4800 NG result.

Conclusion The cobas 4800 NG test is acceptable for urogenital and rectal specimens without supplementary testing in our low prevalence (< 1%) population, however other non-genital sites require confirmation.