Routine universal testing versus selective or incidental testing for oropharyngeal Chlamydia trachomatis in women in the Netherlands: a retrospective cohort study.

BACKGROUND

Pharyngeal Chlamydia trachomatis in women might contribute to autoinoculation and transmission to sexual partners. Data for effectiveness of different testing practices for pharyngeal C trachomatis are scarce. We therefore aimed to assess the prevalence of pharyngeal C trachomatis, determinants, and effectiveness of different testing practices in women.

METHODS

We did a retrospective cohort study, in which surveillance data for all women visiting sexually transmitted infection clinics in all regions in the Netherlands between Jan 1, 2008, and Dec 31, 2017, were used. We collected consultation-level data and individual-level data from 2016 onwards for sociodemographic characteristics, sexual behaviour in the past 6 months, self-reported symptoms, and STI diagnoses. The primary outcome was the positivity rate of pharyngeal C trachomatis infection compared between routine universal testing (>85% tested pharyngeally per clinic year), selective testing (5-85% tested pharyngeally per clinic year), and incidental testing (<5% pharyngeally tested per clinic year). We calculated the number of missed infections by extrapolating the positivity rate assessed by routine universal testing to all selectively tested women. We used multivariable generalised estimating equations logistic regression analyses to assess independent risk factors for pharyngeal C trachomatis and used the assessed risk factors as testing indicators for comparing alternative testing scenarios.

FINDINGS

Between Jan 1, 2008, and Dec 31, 2017, a total of 550 615 consultations with at least one C trachomatis test was recorded, of which 541 945 (98·4%) consultations (including repeat visits) were included in this analysis. Pharyngeal C trachomatis positivity was lower in the routine universal testing group than in the selective testing group (1081 [2·4%; 95% CI 2·2-2·5] of 45 774 vs 3473 [2·9%; 2·8-3·0] of 121 262; p<0·0001). The positivity rate was also higher among consultations done in the incidental testing group (44 [4·1%; 95% CI 3·1-5·5] of 1073; p<0·0001) than in the routine universal testing group. Based on extrapolation, selective testing would have hypothetically missed 64·4% (95% CI 63·5-65·3; 6363 of 9879) of the estimated total of C trachomatis infections. The proportion of pharyngeal-only C trachomatis was comparable between routinely universally tested women (22·9%) and selectively tested women (20·4%), resulting in a difference of 2·5% (95% CI -0·3 to 5·3; p=0·07). When using risk factors for pharyngeal C trachomatis as testing indicators, 15 484 (79·6%) of 19 459 women would be tested to detect 398 (80·6%) of 494 infections.

INTERPRETATION

No optimal testing scenario was available for pharyngeal C trachomatis, in which only a selection of high-risk women needs to be tested to find most pharyngeal C trachomatis infections. The relative low prevalence of pharyngeal-only C trachomatis (0·5%) and probably limited clinical and public health effect do not provide support for routine universal testing.

FUNDING

Public Health Service South Limburg.