The last decade has seen a rapid increase in online postal self-sampling (OPSS) for sexually transmitted infections (STIs) in the UK. Evaluating OPSS is important to understand its uptake and use by different populations.
We analysed pseudonymised routine data from online sexual health providers in three case study areas (CSAs) in England, conducting descriptive analysis of the OPSS testing cascade for service users aged 16+ years who submitted an online kit order (1 January 2022 to 30 November 2022). We used logistic regression to determine crude and adjusted associations by CSA between service user characteristics and progression along the cascade.
Of 610 876 orders submitted and kits dispatched, 51.5–78.4% of STI kits across CSAs were returned. 98.5% (n=459 304/466 392) of returned STI kits contained 1+ processable swab/urine sample. Chlamydia test positivity as a proportion of kit orders ranged from 3.7% (CSA2) to 5.7% (CSA3), with gonorrhoea test positivity ranging from 1.29% (CSA1) to 1.63% (CSA2).
Return of blood samples varied from 36.3% to 65.3% and return of processable blood samples varied from 22.8% to 57.9% of orders submitted. Older people (45+ years) were more likely than younger people (16–19 years) to return a processable blood sample (CSA1-adjusted OR (aOR)=1.85, CSA2-aOR=1.48, CSA3-aOR=1.43, all p<0.001) as were those in less deprived versus those in most deprived areas (CSA1-aOR=1.24; CSA2-aOR=1.10; CSA3-aOR=1.37, all p<0.001). Women, and men who did not report sex with men (respectively), were less likely than men who have sex with men to return a processable blood sample (CSA1-aOR=0.42, 0.39; CSA2-aOR=0.75, 0.87; CSA3-aOR=0.40, 0.54, all p<0.001), as were those reporting symptoms compared with those with no symptoms (CSA1=0.75; CSA2=0.81; CSA3=0.92, all p≤0.004). Positive predictive value for HIV screening via OPSS (excluding those known to be living with HIV) ranged from 4.4% (CSA2) to 7.1% (CSA3).
We found consistent differences across CSAs in return of OPSS kits and processable blood samples by age, deprivation level, sex, sexual behaviour and symptom status. Interventions are needed to ensure that the testing pathways available meet the needs of all populations.