In women the vulvovaginal swab has been established as the sample of choice for Chlamydia trachomatis.1 ,2 However, there is no consensus on the optimal sample for the detection of Mycoplasma genitalium.3–6

Since 2008, all attendees to Olafia Clinic (Department of Venereology, Oslo University Hospital) have routinely been screened for C. trachomatis and M. genitalium. Following the publication of international guidelines recommending the vulvovaginal swab for C. trachomatis screening,1 we discontinued dual collection of first void urine (FVU) and genital swab in women in 2010 and continued with only the genital swab. However, the subsequent publication of Gesink et al’s6 study where 5 of 18 women (28%) tested positive for M. genitalium only in the urine sample and not the genital swab, prompted us to review our practice.

During a 3 month period (1 February 2013–30 April 2013) all women attending the clinic for a sexually transmitted infection screen were asked to provide FVU in addition to the routine genital swab (self-taken vulvovaginal if asymptomatic or physician-taken cervicovaginal during the gynaecological exam if symptomatic). As this was a quality assurance study, ethics committee approval was not required.

If either the FVU or genital swab was positive, this was interpreted as a true positive for calculating the relative sensitivity of the sample. Where the FVU was missing, the corresponding genital swab was excluded from the statistical analysis.

During the study time, there were 1110 attendances from women requesting a sexually transmitted infection screen. Of the attendances 1097 (98.8%) resulted in a paired genital swab and FVU. There were 83 (7.5%) C. trachomatis positive (multiplex assay http://www.geneproof.com) and 71 (6.5%) M. genitalium positive tests (inhouse PCR).

Application of χ² test with Yates continuity correction found no difference between the self-taken and physician-taken swabs.

Using McNemar’s χ² test on paired samples, the relative sensitivities for FVU versus genital swab for C. trachomatis were 75% vs 96% (p<0.001) and for M. genitalium 61% vs 100% (p<0.001), respectively.

In this study the genital swab had a significantly higher sensitivity compared with FVU when testing for M. genitalium. Specifically there were zero false negatives found from the genital swab.

The study also validates preferential use of the genital swab when screening for C. trachomatis in accordance with the general consensus in the published literature.

In conclusion, the genital swab is the single sample of choice when performing dual screening for C. trachomatis and M. genitalium in women.