Background The role of non-invasive testing for gonorrhoea (GC) in women has not yet been fully established in the UK. Validation of NAATs in low prevalence populations has been recommended. Our study is the first to compare gonorrhoea detection on self-taken VVSs by AC2 assay with gold-standard culture of clinician-taken urethral and endocervical (endocx) samples. As it is unclear whether a VVS or endocx swab is better for the detection of chlamydia (CT) by AC2 assay in women needing an examination we also compared the sensitivities of these samples for CT detection.

Methods Women aged 16+ requesting STI testing consenting to perform a self-taken VVS prior to routine examination were included. Clinicians took urethral and endocx samples for gonorrhoea culture and an endocx swab for AC2 assay. AC2 positives were confirmed with Aptima GC and Aptima CT assays.

Results 3973 women included, 100 (2.5%) were infected with GC. Overall sensitivities were culture 82%; clinician taken endocx AC2 96%; self-taken VVS AC2 99% (p=0.0002). The specificity of all the AC2 tests was 100%. In women with symptoms the sensitivities were culture 84%; clinician taken endocx AC2 100% and VVS AC2 100% (p=0.003). In women with no symptoms 1.9% had GC. The sensitivities were culture 79%; clinician taken endocx AC2 91% and self-taken VVS AC2 97.5% (p=0.015). The endocx AC2 performed less well in women without symptoms, 91% vs 100% (p=0.031); the VVS AC2 assay performed equally well, 97.5% vs 100% (p=0.41). Overall sensitivities for CT were clinician taken endocx AC2 89%; self-taken VVS AC2 97% (p=0.0001). In women with symptoms the sensitivities were clinician taken endocx AC2 88%; self-taken VVS AC2 97% (p=0.001). In women with no symptoms the sensitivities were clinician taken endocx AC2 89%; self-taken VVS AC2 98% (p=0.002).

Conclusion AC2 assay of self-taken VVSs was significantly more sensitive for the detection of GC than culture of urethral and endocx samples and equivalent to detection by AC2 assay from clinician-taken endocx swabs. The specificity and PPV of the AC2 assay was very high in this low prevalence population. AC2 assay of self-taken VVSs was significantly more sensitive for the detection of CT than AC2 assay of clinician-taken endocx samples. On the basis of these findings a self-taken VVS is the sample of choice in women who do not need an examination. In those who are being examined either a self-taken or clinician-taken VVS is the sample of choice giving better detection rates of GC and CT than an endocx swab.