Self-sampling has similar accuracy to that of conventional testing for diagnosing Chlamydia trachomatis and Neisseria gonorrhoeae infections and may increase linkage and engagement in care when combined with digital innovations. These study findings were published in BMJ Sexually Transmitted Infections.
Researchers conducted a systematic review and meta-analysis using data captured between January 2006 and January 2023. They compared the diagnostic accuracy of self-sampling strategies – alone or in combination with digital innovations – with that of conventional testing for identifying Chlamydia trachomatis and Neisseria gonorrhoeae infections. Digital innovations were defined as online-based, mobile-based, and computing technologies designed to support health interventions. The researchers also assessed the feasibility, accessibility, and uptake associated with the use of self-sampling strategies.
A total of 45 studies were included in the final analysis. The studies were conducted in 18 countries, of which 10 were high-income countries (HICs; 75.6%) and 8 were low- and middle-income countries (LMICs; 24.4%). Most studies had an observational design (n=16), with the exception of 2 randomized control trials. The median sample size across all studies was 480 participants. Of all the studies, 12 included digital interventions (26.7%).
For chlamydia, self-sampling had an overall sensitivity of 88.2 (95% CI, 81.9-92.6), with the highest sensitivity observed for rectal self-sampling (93.1; 95% CI, 87.8-96.2) and the lowest for vaginal self-sampling (79.6; 95% CI, 62.1-90.3). Overall, self-sampling had a specificity of 99.3 (95% CI, 98.9-99.5).
For gonorrhea, self-sampling had an overall sensitivity of 88.9 (95% CI, 84.2-92.3), with the highest sensitivity observed for pharyngeal self-sampling (94.3; 95% CI, 89.6-96.9) and the lowest for vaginal self-sampling (79.9; 95% CI, 68.6-87.9). The specificity of self-sampling for diagnosing gonorrhea was consistent across all sampling sites, with an overall specificity of 99.1 (95% CI, 98.6-99.4).
In studies that assessed self-sampling alone (n=21), participants reported a high level of acceptability (range, 80.0%-100.0%). However, the percentage of participants indicating a preference for self-sampling over clinician-based sampling significantly varied across studies (range, 23.1%-83.7%). Overall, the feasibility of self-sampling alone was high, with a completion rate between 83.3% and 99.0%).
In studies that assessed self-sampling combined with digital innovations, most participants rated this testing method as acceptable (93.0%), very satisfactory (95.5%), or easy to use (89.0%). A high percentage of participants indicated a preference for this testing method over clinician-based sampling in 2 studies (51.5% and 77.0%). In regard to feasibility, digital innovations, such as text-based reminders and website-based questionnaires, combined with self-sampling resulted in high rates of engagement and test completion. Overall, self-sampling combined with digital innovations successfully detected 36.8% (33.1 per 100 person-years [py]) and 17.9% (29.9 per 100 py) of new chlamydia and gonorrhea diagnoses, respectively.
Study limitations include potential publication bias, the lack of data on potential confounders, inconsistent study quality, and the low number of studies conducted in LMICs.
According to the researchers, “SS [self-sampling] constitutes an innovative opportunity to reduce global STI burden by addressing obstacles associated with conventional testing.”
Vialard F, Anand A, Leung Soo C, et al. Self-sampling strategies (with/without digital innovations) in populations at risk of Chlamydia trachomatis and Neisseria gonorrhoeae: a systematic review and meta-analyses. Sex Transm Infect. Published Online, March 29, 2023. doi:10.1136/sextrans-2022-055557