Self-completed online contact tracing for COVID-19 is associated with reporting fewer contacts: an observational study

BACKGROUND: Contact tracing was central to England's approach to controlling coronavirus disease 2019 (COVID-19). England's COVID-19 contact tracing system included a step where cases were invited to provide information through an online 'self-service' portal. It is important to understand whether this had any effect on the numbers of contacts reported. METHODS: This study analyses contact tracing data for one local authority in England for 2021. Univariable negative binomial and logistic regression models were used to test for associations between contact tracing tier and average numbers of contacts reported, and key demographic factors and date of reporting. The association between contact tracing tier and number of contacts reported adjusted for age, gender, ethnicity, deprivation and the date on which the case was reported was estimated using a multivariable negative binomial generalized additive regression model with smooth terms for date of reporting and age. RESULTS: Controlling for age, gender, ethnicity, deprivation and date of reporting, self-completed contact tracing was associated with reporting 23% fewer contacts (95% confidence interval 20-25% fewer contacts). CONCLUSION: The use of online self-completed contact tracing may contribute to under-reporting of close contacts. Unmeasured confounding is likely so caution is needed in assuming that the results reflect a causal relationship.