Estimating SARS-CoV-2 infections and associated changes in COVID-19 severity and fatality

Abstract

Background: The difficulty in identifying SARS-CoV-2 infections has not only been the major obstacle to control the COVID-19 pandemic but also to quantify changes in the proportion of infections resulting in hospitalization, intensive care unit (ICU) admission, or death.

Methods: We developed a model of SARS-CoV-2 transmission and vaccination informed by official estimates of the time-varying reproduction number to estimate infections that occurred in Italy between February 2020 and 2022. Model outcomes were compared with the Italian National surveillance data to estimate changes in the SARS-CoV-2 infection ascertainment ratio (IAR), infection hospitalization ratio (IHR), infection ICU ratio (IIR), and infection fatality ratio (IFR) in five different sub-periods associated with the dominance of the ancestral lineages and Alpha, Delta, and Omicron BA.1 variants.

Results: We estimate that, over the first 2 years of pandemic, the IAR ranged between 15% and 40% (range of 95%CI: 11%-61%), with a peak value in the second half of 2020. The IHR, IIR, and IFR consistently decreased throughout the pandemic with 22-44-fold reductions between the initial phase and the Omicron period. At the end of the study period, we estimate an IHR of 0.24% (95%CI: 0.17-0.36), IIR of 0.015% (95%CI: 0.011-0.023), and IFR of 0.05% (95%CI: 0.04-0.08).

Conclusions: Since 2021, changes in the dominant SARS-CoV-2 variant, vaccination rollout, and the shift of infection to younger ages have reduced SARS-CoV-2 infection ascertainment. The same factors, combined with the improvement of patient management and care, contributed to a massive reduction in the severity and fatality of COVID-19.

Keywords: IFR; IHR; SARS‐CoV‐2; infection ascertainment ratio; infection fatality ratio; infection hospitalization ratio.

FIGURE 1

The COVID‐19 pandemic in Italy. (A) Mean estimates of the time‐varying reproduction number Rt as obtained from epidemic curves of symptomatic cases by date of symptom onset collected by the National Integrated Surveillance System (mean, grey solid line; shaded area, 95% CI; y‐axis on the left). Horizontal dotted line: epidemic threshold (Rt = 1). Grey bars represent the daily incidence per 1000 individuals of SARS‐CoV‐2 confirmed symptomatic infections by date of symptom onset as reported to the Italian Integrated Surveillance System, (y‐axis on the right). Background colors indicate the classification in different phases, and the dates indicated within the graph denote the day of transition between consecutive phases. The vertical dotted line denotes the start of the vaccination campaign on December 27, 2020. (B) Daily number of vaccine doses administered in Italy per 1000 individuals (stacked bar chart, y‐axis on the left). Line and bar colors, from lighter to darker shades, respectively indicate first, second, and booster doses. Solid lines show the cumulative vaccination coverage in the Italian population (y‐axis on the right). In Italy, by February 2022, administration of two doses was recommended to all individuals aged 5 years or more; administration of one booster dose was recommended to all individuals aged 12 years or more.

FIGURE 2

Schematic representation of the metrics of interest. Infection ascertainment ratio (IAR), ratio between SARS‐CoV‐2 ascertained infections and total infections estimated by the model; infection hospitalization ratio (IHR), ratio between hospitalized COVID‐19 cases and total infections estimated by the model; infection intensive care unit (ICU) ratio (IIR), ratio between COVID‐19 cases admitted to an ICU and total infections estimated by the model; infection fatality ratio (IFR), ratio between COVID‐19 deaths and total infections estimated by the model.

FIGURE 3

SARS‐CoV‐2 cumulative incidence. Estimated phase‐specific SARS‐CoV‐2 cumulative incidence (%) between February 21, 2020 and February 20, 2022 in the overall population and by age classes. Colors indicate the considered phases. Bars, mean estimates; vertical lines, 95% CI; n = 300 stochastic model realizations.

FIGURE 4

Validation of SARS‐CoV‐2 seroprevalence. (A) Grey dots represent weighted point estimates of the infection‐induced seroprevalence in high‐income European countries, as reported in a published meta‐analysis of population‐based serological studies. Red asterisks represent the mean proportion of the Italian population who had a previous natural infection, independently of vaccination status as estimated through the model. Vertical lines indicate 95%CI. (B) Grey dots represent weighted point estimates of the overall seroprevalence, that is, either induced by natural infection or vaccination, in high‐income European countries as reported in a published meta‐analysis of population‐based serological studies. Red asterisks represent the mean proportion of the Italian population who had a previous natural infection or has received one or more vaccine doses as estimated through the model. Vertical lines indicate 95%CI. Background colors indicate the classification in different phases, the vertical dotted line denotes the start of the vaccination campaign on December 27, 2020.

FIGURE 5

SARS‐CoV‐2 infection ascertainment ratio (IAR). Estimated phase‐specific SARS‐CoV‐2 infection ascertainment ratio between February 21, 2020 and February 20, 2022 (%). Bars: mean estimates; vertical lines: 95% CI; n = 300 stochastic model realizations.

FIGURE 6

Changes in SARS‐CoV‐2 infection hospitalization ratio, intensive care unit (ICU) ratio and fatality ratio. (A) Infection hospitalization ratio (IHR). (B) Infection ICU ratio (IIR). (C) Infection fatality ratio (IFR). (D) Estimated relative reductions in the IHR compared with the first ancestral phase (%). (E) As D but for the IIR. (F) As D but for the IFR. Bars: mean estimates; vertical lines: 95% CI; n = 300 stochastic model realizations.