The real-time analysis of infectious disease surveillance data is essential in obtaining situational awareness about the current dynamics of a major public health event such as the COVID-19 pandemic. This analysis of e.g., time-series of reported cases or fatalities is complicated by reporting delays that lead to under-reporting of the complete number of events for the most recent time points. This can lead to misconceptions by the interpreter, for instance the media or the public, as was the case with the time-series of reported fatalities during the COVID-19 pandemic in Sweden. Nowcasting methods provide real-time estimates of the complete number of events using the incomplete time-series of currently reported events and information about the reporting delays from the past. In this paper we propose a novel Bayesian nowcasting approach applied to COVID-19-related fatalities in Sweden. We incorporate additional information in the form of time-series of number of reported cases and ICU admissions as leading signals. We demonstrate with a retrospective evaluation that the inclusion of ICU admissions as a leading signal improved the nowcasting performance of case fatalities for COVID-19 in Sweden compared to existing methods.

Background: Vaccination was the single most effective measure in mitigating the impact of the COVID-19 pandemic. Our study aims to quantify the impact of vaccination programmes during the initial year of vaccination (2021) by estimating the number of case fatalities avoided, using Sweden as a case study.

Methods: Using Swedish data on age-specific reported incidence and vaccination uptake, along with vaccine efficacies, age-specific contact patterns and under-reporting from the literature, we fit a Bayesian SEIR epidemic model with time-varying community contact rate for COVID-19 incidence. Age-specific fatality rates from the literature are adjusted proportionally to fit the observed number of case fatalities in the factual analysis, resulting in 5,510 (95% PI: 5,370-5,650) matching the observed number 5,430. The estimated time-varying community contact rate is then used in a counterfactual analysis where the population is unvaccinated, leading to more infections and fatalities. A sensitivity analysis is performed to identify which parameters influence our conclusions.

Findings: The counterfactual analysis result in a severe epidemic outbreak during the early autumn of 2021, resulting in about 37,100 (36,700–37,500) number of case fatalities. Consequently, the number of lives saved by the vaccination programme is estimated to be about 31,600 (31,300–32,000), out of which 5,170 are directly saved and 26,400 are indirectly saved, mainly by drastically reducing the severe outbreak in the early autumn of 2021, which would have occurred without vaccination and unchanged community contact rate.

Interpretation: Our mathematical model is used to analyse the impact of COVID-19 vaccination on lives saved in Sweden during 2021, but the same methodology can be applied to other countries. The counterfactual analysis offers insights into an alternative trajectory of the pandemic without vaccination. The results show the direct impact of vaccination on reducing deaths for infected individuals and shed light on the indirect effects of reduced transmission dynamics.

Identifiability is the property in mathematical modelling that determines if model parameters can be uniquely estimated from data. For infectious disease models, failure to ensure identifiability can lead to misleading parameter estimates and unreliable policy recommendations. We examine the identifiability of a modified SIR model that accounts for under-reporting and pre-existing immunity in the population. We provide a mathematical proof of the unidentifiability of jointly estimating three parameters: the fraction under-reporting, the proportion of the population with prior immunity, and the community transmission rate, when only reported case data are available. We then show, analytically and with a simulation study, that the identifiability of all three parameters is achieved if the reported incidence is complemented with sample survey data of prior immunity or prevalence during the outbreak. Our results show the limitations of parameter inference in partially observed epidemics and the importance of identifiability analysis when developing and applying models for public health decision making.