Vulvo-vaginal swabs (VVSs) are widely used as a non-invasive sample type for nucleic acid amplification testing (NAAT) for Chlamydia trachomatis. Although currently only the APTIMA system (Gen-Probe Inc, San Diego, California, USA) is approved for use with this type of sample, there have been many studies showing that self-taken VVSs perform well when using other NAAT platforms.1^–3

In addition to facilitating screening programmes there is also the possibility of using this sample type in clinical settings. In a multicentre study4 it was demonstrated that a self-taken VVS was at least as good as a clinician-taken cervical swab. This type of sample removes the need for a speculum investigation for many patients and has been suggested as an alternative to combined cervical/urethral swabs.5 This would also expedite the processing of patients to reach the 48 h access target.

Laboratories in the UK need to provide validation data when introducing a new assay or a modification of an existing test. In the case of C trachomatis this often tends to rely on citing published studies using various platforms. As there appears to be no data on the use of VVSs with the Roche Taqman system we undertook a study to evaluate results obtained with this type of sample. We compared self-collected VVSs with endocervical swabs, which have generally been considered to be the optimal sample type for detection of infection in females.

METHOD

Female attendees at two genitourinary medicine (GUM) clinics, aged 18–24 years, were asked to take part in the study. They were provided with an information sheet and invited to discuss any issues. Participants were required to sign a consent form and the healthcare staff then took the endocervical swab. The patient was given a leaflet showing how to take a self-taken swab and then allowed to take the sample.

Swabs were collected using Remel RT collection kits. All samples were tested using COBAS Taqman 48 CT (Roche Diagnostics, Branchburg, USA) and the tests were carried out following the manufacturer’s instructions.

RESULTS

A total of 267 women provided both sample types for the study. Results are shown in table 1.

There was concordance for 255/267 (96%; 95% CI 92 to 98%) sets of samples. Samples from two women gave an invalid/inhibitory result for both types of swab. A further 10 VVSs were invalid/inhibitory, two of which were positive at the endocervix. This means that 0.7% of endocervical swabs gave an invalid/inhibitory result compared with 4.5% for VVSs. The overall positivity rate was 9.3%.

DISCUSSION

The results obtained in this study using VVSs were equivalent to those found with endocervical swabs. Keane et al, using the Cobas Amplicor system, recently showed a slightly higher percentage of positives (16.8%) with VVSs compared with endocervical samples (15.9%), although the difference is not significant.3

In this study, more VVSs samples (4.5%) gave invalid/inhibitory results compared with endocervical swabs (0.7%), although this is within the limit of 5% suggested in the manufacturer’s kit insert. However, invalid/inhibitory results mean that a definitive diagnosis is not immediately available and re-testing may be inconvenient. Therefore, levels of inhibitory samples should be monitored to ensure that they do not compromise the usefulness of the assay. The issue of inhibitory samples in NAATs is well recognised and has been reported elsewhere, although at varying levels for different platforms and sample types.2 6

Many studies report that women find it easy to provide a VVS and have no objections to this type of sample.7 8 Anecdotally, we did have some difficulty recruiting to this study and we attributed this to the genitourinary medicine setting and also to the need for written consent. However, the aim of this project was not to investigate acceptability of self-taken swabs and we feel that, for the reasons stated, a study done in this setting would not truly reflect acceptability.

CONCLUSION

VVSs provided equivalent results to those obtained with endocervical swabs although with a higher number of invalid/inhibitory results. This sample type is suitable for use with COBAS Taqman 48 CT (Roche Diagnostics).