Quantifying SARS-CoV-2 Omicron variant spread and the impact of non-pharmaceutical interventions in Newfoundland and Labrador, Canada

The highly transmissible Omicron variant of SARS-CoV-2 caused many infections in Newfoundland and Labrador (NL), and the fraction of infections that were unreported varied as PCR testing capacity was exceeded and eligibility rules changed. Due to these inconsistencies in the testing rate, we developed a mechanistic model that was calibrated to seroprevalence data (Dec 2021-May 2022) to estimate underreporting and understand the impact of non-pharmaceutical interventions on transmission. Our model considers the epidemiology of SARS-CoV-2 spread, natural and vaccine-derived immunity, and the booster dose vaccination campaign that was ongoing in NL during the study. We found that during the early spread of the Omicron variant, when the eligibility for tests that were reported in the provincial counts was less restrictive, fewer than 3 infections were unreported per reported case. After March 17, 2022, when test eligibility was more restrictive, the underreporting rate increased steadily, with an average of 21 infections unreported infections per reported case. We found that Omicron transmission was lower when schools were closed (mean reproduction number, 1.44, 95% CI: 1.41 - 1.47), higher when open (mean reproduction number, 2.03, 95% CI: 1.60 - 2.46), and the strictest alert level (ALS-4) reduced transmission the most (mean reproduction number, 1.52, 95% CI: 1.35 - 1.69). When underreporting rates vary, the impact of non-pharmaceutical interventions (NPIs), such as alert level systems and school closures, cannot be determined from reported cases. Our findings highlight the value of combining seroprevalence data with modelling to determine the impact of NPIs during pandemics when surveillance systems are constrained.