Evaluating the impact on health outcomes of an event that resulted in a delay in contact tracing of COVID-19 cases

Objective

In September 2020, records of 15,861 SARS-CoV-2 cases failed to upload from the Second Generation Laboratory Surveillance System (SGSS) to the Contact Tracing Advisory Service (CTAS) tool, resulting in a delay in the contact tracing of these cases. This study used CTAS data to determine the impact of this delay on health outcomes: transmission events, hospitalisations, and mortality. Previously, a modelling study had suggested a substantial impact.

Design

Observational study

Setting

England.

Population

Individuals testing positive for SARS-CoV-2 and their reported contacts.

Main outcome measures

Secondary attack rates (SARs), hospitalisations, and deaths amongst primary and secondary contacts were calculated, compared to all other concurrent, unaffected cases. SGSS records affected by the event were matched to CTAS records and successive contacts and cases were identified.

Results

The initiation of contact tracing was delayed by 3 days on average in the primary cases in the delay group (6 days) compared to the control group (3 days). This was associated with lower completion of contact tracing of primary cases in the delay group: 80% (95%CI: 79-81%) in the delay group and 83% (95%CI: 83-84%) in the control group. There was some evidence to suggest an increase in transmission to non-household contacts amongst those affected by the delay. The SAR for non-household contacts was higher amongst secondary contacts in the delay group than the control group (delay group: 7.9%, 95%CI:6.4% to 9.2%; control group: 5.9%, 95%CI: 5.3% to 6.6%). There was no evidence of a difference between the delay and control groups in the odds of hospitalisation (crude odds ratio: 1.1 (95%CI: 0.9 to 1.2) or death (crude odds ratio: 0.7 (0.1 to 4.0)) amongst secondary contacts.

Conclusions

The delay in contact tracing had a limited impact on population health outcomes.

Strengths and limitations of the study

• Shows empirical data on the health impact of an event leading to a delay in contact tracing so can test hypotheses generated by models of the potential impact of a delay in contact tracing

• Estimates the extent of further transmission and odds of increased mortality or hospitalisation in up to the third generation of cases affected by the event

• The event acts as a natural experiment to describe the possible impact of contact tracing, comparing a group affected by chance by delayed contact tracing to a control group who experienced no delay

• Contact tracing was not completed for all individuals, so the study might not capture all affected contacts or transmissions