key: cord-0841781-htzw42h2
authors: Hsu, Sylvia H.; Chang, Su-Hsin; Gross, Cary; Wang, Shi-Yi
title: Relative risks of Covid-19 fatality between the first and second waves of the pandemic in Ontario, Canada
date: 2021-07-01
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2021.06.059
sha: fe58294360670bef5bc0d771408ad559199817c2
doc_id: 841781
cord_uid: htzw42h2

OBJECTIVES: To examine whether the case fatality rate (CFR) of coronavirus disease 2019 (Covid-19) decreased overtime, and whether the Covid-19 testing rate is a driving factor for the changes if the CFR decreased. METHODS: Analyzing Covid-19 cases, deaths, and tests in Ontario, Canada, we compared the CFR between the first wave and the second wave across 26 Public Health Units in Ontario. We also explored whether a high testing rate was associated with a large CFR decrease. RESULTS: The first wave CFR ranged from 0.004 to 0.146; whereas the second wave CFR ranged from 0.003 to 0.034. The pooled RR estimate of the second wave Covid-19 case fatality, compared to first wave, was 0.24 (95% CI: 0.19-0.32). Additionally, Covid-19 testing percentages were not associated with the estimated RR (p-value = 0.246). CONCLUSIONS: The Covid-19 CFR decreased profoundly in Ontario during the second wave, and Covid-19 testing was not a driving factor for this decrease.

The coronavirus disease 2019 pandemic continues to surge globally. Recently, many countries experienced a substantial second wave of Covid-19 (James et al., 2021) . While much higher infection numbers, and in some countries more deaths were seen, a substantial decrease in the case fatality rate (CFR) was observed (https://www.ft.com/content/b3801b63-fbdb-433b-9a46-217405b1109f ). It is interesting to note that research reported the CFR decreased during the first wave, indicating that Covid-19 severity might lessen with time (Schmidt et al., 2021; Anesi et al., 2021) . Prior studies, however, only examined the declining CFR in hospitalized patients during the first wave. Since there is a higher likelihood of detecting asymptomatic and mildly symptomatic infections during the second wave, health experts suspect that increased testing numbers may partially explain such declines in CFR (https://www.ft.com/content/b3801b63-fbdb-433b-9a46-217405b1109f ). Little is known regarding the CFR changes between the first and second waves at the population level or whether high testing rates were associated with CFR decrease. Determining the magnitude in CFR decrease and the association between Covid-19 testing rates and CFR changes could provide implications for future public health interventions.

We conducted a cross-sectional study using publicly reported Covid-19 cases, deaths, and tests available on the Ontario Public Health website as of February 10, 2021 (https://covid-19.ontario.ca/data). In Ontario, Canada, the peak 7-day moving average during the first and second waves was 640 in April 2020 and 4,249 in January 2021. As the number of new cases in Ontario dropped below 200 per day between late June and early September, 2020, we used July 31 to separate the first and second waves. We also conducted sensitivity analyses using June 30, 2020 and August 31, 2020 as the cut-off date. We determined the CFR per Public Health Unit (PHU), for both first and second waves, defined as the number of deaths divided by the number of new cases. Those PHUs which had no Covid-19 deaths in the first or second wave were excluded. We calculated the relative risk (RR) of Covid-19 CFR between the two waves (that is, the second wave vs. the first wave) and the corresponding 95% confidence interval (95% CI) for each PHU. Using a random-effects meta-analysis with inverse variance weighting, we estimated the pooled RR on population-level CFR across PHUs. Using meta-regression techniques, we regressed the logarithm form of effect size by the second-wave testing percentage (the number of testing divided by the total population). Acknowledging that second-wave patients were younger, we also analyzed individual-level data. Applying hierarchical generalized linear models, we estimated adjusted odds ratio (AORs), clustering by PHUs and controlling for patient age, wave, and testing percentage.

Of the 34 PHUs in Ontario, Canada, 26 PHUs had more than one death in the first and second waves, and were included in the analyses. The first wave CFR ranged from 0.004 to 0.146; whereas the second wave CFR ranged from 0.003 to 0.034. There was substantial variation in the RRs between the two waves across PHUs (Figure 1 ). The pooled RR estimate of the second wave Covid-19 case fatality, compared to first wave, was 0.24 (95% CI: 0.19-0.32). Metaregression models showed that Covid-19 testing percentages were not statistically significantly associated with the estimated RR (p-value = 0.246), despite a trend that the higher the percentage of Covid-19 testing, the lower the effect size; i.e., RR close to 1 (Appendix Figure) . Individuallevel data analyses, adjusting for age, revealed that the AOR of the second wave, compared to the first wave, was 0.41 (95% CI: 0.38-0.43; Table 1 ). Sensitivity analyses using two different dates to define the second wave reached similar conclusions (data not shown).

We found that the Covid-19 CFR decreased profoundly in Ontario during the second wave.

Young age contributed to this decrease, yet high Covid-19 testing was not a driving factor for this decrease. Potential reasons for the decrease in COVID-19 CFRs may be attributed to improvement in clinical care: For example, there was increased evidence to support corticosteroid use (Chris Baraniuk, 2021) , and Canada approved effective treatments, such as Remdesivir and Bamlanivimab in July and November, 2020, respectively (https://www.canada.ca/en/health-canada/services/drugs-health-products/covid19-industry/drugsvaccines-treatments/authorization/list-drugs.html). Additionally, our findings suggest that coronavirus, despite of becoming more transmissible, might decline in virulence over time.

Indeed, researchers have demonstrated a dynamic shift in COVID variants to those with increased survival (Esper et al., 2021) . As coronavirus genome sequencing in Ontario during the study period is lacking, research on the variants' CFR is needed. Increased public awareness during the second-wave period, such as compliance with wearing masks, could also decrease the viral inoculum, leading to more mild and asymptomatic infection manifestations (Gandhi et al., 2020) . While our random-effects models account for PHU-level characteristics, we acknowledge that we are unable to control for confounding factors. To evaluate Covid-19 intervention programs, researchers and policy decision-makers should account for the changes in Covid-19 severity. 

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Characteristics, Outcomes, and Trends of Patients With COVID-19-Related Critical Illness at a Learning Health System in the United States

Where are we with drug treatments for covid-19?

Drug and vaccine authorizations for COVID-19: List of authorized drugs, vaccines and expanded indications

Genomic Epidemiology of SARS-CoV-2 Infection During the Initial Pandemic Wave and Association With Disease Severity

Masks do more than protect others during COVID-19: reducing the inoculum of SARS-CoV-2 to protect the wearer

COVID-19 second wave mortality in Europe and the United States

Summary Report. SARS-CoV-2 Variants of Concern: Results of Point Prevalence Study

Clinical characteristics and day-90 outcomes of 4244 critically ill adults with COVID-19: a prospective cohort study