Influence of COVID-19 Vaccination Coverage on Case Fatality Risk

Background: It is well known that COVID-19 vaccines demonstrate higher efficacy against mortality than mild acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The estimation of the proportion of mortalities among morbidities is a measure of the case fatality risk (CFR). To date, few studies have estimated the real-world CFR in relation to COVID-19 vaccination coverage. This study aims to evaluate the change in CFR estimates among different countries following the introduction of COVID-19 vaccines, and to identify the threshold dose of vaccines that changed the CFR as early as April 3, 2021. Furthermore, this study sheds light on the influence of COVID-19 vaccinations on the attack rate (AR), death rate, and, ultimately, CFR. Material and Methods: We collected publicly available data concerning all countries and territories that implemented COVID-19 vaccination at least for 100 days, with an end date of April 3, 2021. In total, we found 16 countries and territories. The CFR was measured as the number of deaths per 100 COVID-19 confirmed cases, while vaccine coverage was defined as the number of doses of vaccine per 100 people in the total population. We performed descriptive data analyses, including the mean value, standard deviation, and graphical presentation, using bar charts. Performed inferential data analyses included the onesample Kolmogorov–Smirnov (K-S) test and general linear model procedure (GLM). Results: Our findings showed a significantly associated decrease in the mean CFR in countries with > 18 COVID-19 vaccine doses per 100 inhabitants. We found a decrease from 1.88 % to 1.45 % with a (p-value =0.03), indicating a decrease in the proportion of total deaths to total cases. There was a decrease in the 95% confidence interval from 0.742-3.006 to 0. 718-2.179. The Original Research Article Raham; AJRID, 9(2): 21-31, 2022; Article no.AJRID.83117 22 decrease in CFR was greater among the total deaths than total cases. Conclusion: COVID-19 vaccination was found to decrease the COVID-19CFR. Recommendations: Post-interventional CFR monitoring may constitute a parameter for measuring vaccination effectiveness and progress of the current pandemic or future pandemics. Furthermore, post-interventional CFR estimates can be used as a parameter for assessment effectiveness of interventions e.g. COVID-19 vaccination effectiveness.


INTRODUCTION
The ongoing global coronavirus 2019 (COVID-19) pandemic was initially reported in Wuhan, China, in December 2019. After a few weeks, COVID-19 had spread to several countries and became a significant public health problem [1,2,3]. The rapid spread of this disease has caused a substantial burden on morbidity, with a variable case fatality risk (CFR). CFR is an important parameter to understand the epidemiological features of an outbreak or epidemic [4,5], and is frequently used as a COVID-19 indicator to measure the mortality of the disease and its variants [6,7].
In late 2020, COVID-19 vaccines became crucial tools in the pandemic response. COVID-19 vaccines protect against the transmission of the disease, severe disease, and death [8]. Dozens of countries now have advanced vaccination campaigns as they rush to protect their people and revive their economies.
Measures used to monitor countries' vaccination progress include measuring the daily or 7-day average decrease in the number of cases, measuring the daily or 7-day average decrease in the number of deaths, and measuring the number of confirmed COVID-19 hospital admissions [9].
The COVID-19 vaccine has had a substantial impact in reducing the incidence, hospitalizations, and deaths related to COVID-19, especially among vulnerable individuals with comorbidities and risk factors associated with severe COVID-19 [10].
Although numerous primary studies conducted before the implementation of COVID-19 mass vaccination programs have reported variable CFR values for COVID-19 across different countries, information about the CFR after the implementation of COVID-19 mass vaccinations is scarce.
Several factors are suggested to be associated with temporal and spatial variances in the COVID-19 CFR, including comorbidity risk; demographic, socioeconomic, and political variables; and the age distribution of the community [11].
We conducted this study to determine the influence of COVID-19 vaccines on the CFR in different countries, and to shed light on the vaccine influence on disease transmission among different countries.
As a global real-world study conducted in the middle of the first week of April 2021, this study evaluates the influence of COVID-19 vaccines in the beginning of the launched vaccination program.

MATERIAL AND METHODS
We selected all countries and territories that implemented COVID-19 vaccination for at least the last 100 days, with an end date of April 3, 2021. In total, we found 16 countries and territories that were fit with the inclusion criteria. The publicly available data included the total doses, vaccine doses per 100 people, total deaths, and accumulative COVID-19 cases. The supplementary file contains these data.
The CFR was computed as the total accumulative deaths divided by the accumulative total cases x 100.

Statistical Analysis
The statistical data analysis approaches were used with the statistical package for social sciences (SPSS), version (21), through:

1-
Descriptive data analysis, which included mean value, standard deviation, and graphical presentation by using bar charts.

2-
Inferential data analyses: A: The One-Sample Kolmogorov-Smirnov (K-S) test to accept or reject the statistical hypothesis. B: Through (SPSS), version (21). We also used the general linear model procedure (GLM), through which we incorporated normally distributed dependent variables and interpreted the results using profile plots of the estimated means. In addition, we customized the linear model so that it directly addressed the research question. Table 1 shows a higher initial CFR mean value than the final CFR value. The results also show that the values of the AR, number of cases, and number of deaths were the lowest on April 3, 2021, after subtracting the values on the first day of initiating vaccination. The deaths during study period were 45.82% of the total deaths, while the number of cases was 47.66% of the total cases for the same period, showing a decrease in the proportion of deaths to cases.   Table 2 shows the normal distribution function (goodness-of-fit test). It represents a onesample Kolmogorov-Smirnov test procedure comparing the observed cumulative distribution function for the studied readings with a specified theoretical distribution, which proposes a normal shape (i.e., bell shape).

RESULTS
The results show the distribution of the studied readings regarding CFR marker distribution function in relation to different locations. Since the (Pvalue ) was set to (P>0.05), this enabled us to apply the convention statistical methods (the parametrical methods).
In Table 3 and Figure 1, the results show that the mean CFR was lower in countries with > 18 vaccine doses per 100 people compared to countries with ≤ 18 vaccine doses per 100 people.
We found that countries and territories that had a level of coverage of > 18 doses per 100 people showed a decreased mean CFR compared to the countries' corresponding CFR at the time of initiating the vaccine. The mean CFR was decreased from 1.88 to 1.45. On the other hand, the CFR for countries with a coverage rate of ≤ 18 doses per 100 inhabitants showed decreases in the mean CFR to a lesser extent, with decreases ranging from 3.31 to 3.28. The findings showed that a significant decrease in the mean CFR in countries with > 18 doses per 100 inhabitants. There was a decrease in the 95% confidence interval of the mean from 0.742-3.006 to 0. 718-2.179 ( Fig. 2).  Table 4 shows our tests and analyses of the studied marker for CFR with different sources of variation (SOV), such as the two different dose categories, countries' starting vaccine times, interaction factors represented by applying the GLM of the fixed effects model, and testing the effectiveness of other source of variations not included in the studied model (i.e., the intercept). The R-squared value was 0.157, which determines the proportion of variance in the dependent variable that can be explained by the independent variable.

Fig. 2. 95% confidence interval of CFR mean values
The results show significant differences at P<0.05 related to the studied vaccine dose categories per 100 people, while no significant differences at P>0.05 found for both the time that countries started the vaccine and the interaction factor. In addition, the intercept (the other sources of variations not included in the studied model) recorded highly significant effectiveness at P<0.01.

DISCUSSION
This study demonstrated a decrease in CFR in countries with >18 doses per 100 populations COVID-19 vaccination coverage in a significant association (p-value =0.033).
Since the CFR values were significantly decreased within countries (as a function of  Vaccines provide at least some protection from infection and transmission, but vaccines generally offer higher protection against serious deaths [12]. This study provides evidence on how vaccinations reduce infection and transmission to some extent and deaths to a greater extent. In other words, the effect of the COVID-19 vaccine on deaths outweighs its effect on cases, leading to a decrease in the CFR.

95% CONFIDENCE INTERVAL OF CFR MEAN VALUES
Vaccination of a certain share of the population is essential for the reduction of epidemic transmission in society, as well as for protecting the unvaccinated individuals [13,14].
In recent literature, findings of a positive association between the COVID-19 AR and CFR have been raised [15,16,17]. These studies suggest that the increase in the attack rate (AR) may be correlated to the disease severity. The suggested hypothesis is that the clustering of cases and viral overload lead to an increased mortality rate and CFR. We believe that vaccinations can decrease the number of cases and viral overload, which can lead to a decreased mortality rate and CFR. It is clear that the relative reduction in mortality overcomes the relative reduction in morbidity. This might indicate that the AR plays a role in mortality per se, as stated in the literature.
In one study, COVID-19 vaccination reduced the overall AR from 9.0% to 4.6% over 300 days, which constituted a reduction of approximately 50%. Vaccination markedly reduced adverse outcomes by decreasing non-intensive care unit (ICU) hospitalizations, ICU hospitalizations, and deaths [10].
It was suggested that an increase in the fatality rate as the number of infected people increases is related to the overwhelming of the health care system [11,18]. This should be studied further, as clusters of COVID-19 infections have been associated with an increase in fatalities [19,20].
Furthermore, although the number of hospital beds per 1000 people had a negative association with COVID-19 mortality in certain countries, including European countries, the USA, Mexico, Brazil, and Bolivia, these findings were not global. The number of hospital beds per 1000 people did not have such a negative association in many Asian (excluding Japan) and African countries [21]. Asian and African countries displayed comparatively low mortality rates regardless of their limited bed capacity. The controversy in these findings might be biased by the high ARs in some countries, which makes these beds an insufficient measurement of the CFR. On the other hand, low ARs in other countries likely led to a low CFR regardless of the bed capacity.
In agreement with our study, a 10% increase in vaccine coverage was observed in a crosscounty quasi-experiment in some countries with higher vaccination rates, with a 7.6% reduction in the CFR (95% confidence interval (CI = -12.6 to -2.7%, P = 0.002). This study evaluated the effectiveness of the COVID-19 vaccine during the third week of April 2021 [22] (rather than the middle of first week, as in our study). Another comparable epidemiological study which evaluated the effectiveness of the COVID-19 vaccine on the AR, conducted on May 2, 2021, showed that when the accumulated vaccination rate reached 1.46-50.91 doses per 100 people, the weekly infection rate of the of disease was reduced [23]. The locally adopted strict measures of non-pharmaceutical interventions (NPIs) greatly affect the AR in addition to the vaccination coverage [23]. This study sheds light on the value of the CFR as an important epidemiological parameter to assess the effectiveness of the level COVID-19 vaccination coverage in the real word to protect a whole community from severe disease. This study addresses the effectiveness of COVID-19 vaccination in reducing the real-world CFR as an early post-licensure evaluation, making its order next to vaccine clinical trials.
The finding of low CFR in countries with relatively low vaccination coverage might indicate other mechanisms that decrease the CFR, i.e., through a relatively low reduction in the AR compared to a greater reduction in the death rate. The low Rsquared value and the presence of a highly significant intercept calls for further studies to identify the effect of other possible factors that may reduce the CFR.
The possible limitations in this study include: (1) the COVID-19 vaccine doses administered per 100 people may not equal the number of people that are vaccinated if the vaccine requires two doses, (2) the change in testing coverage within a country or across countries, (3) the difficulty in estimating asymptomatic cases, (4) the difficulty in estimating actual COVID-19 deaths for a variety of reasons, (5) differing COVID-19 preventive approaches across countries and within the same country from time to time, (6) the COVID-19 pandemic stage difference across countries, and (7) the contact-reducing interventions in place.

CONCLUSIONS
Countries with a higher rate of COVID-19 vaccination, indexed as >18 doses per 100 people, reported a significantly associated lower CFR on April 3, 2021, than on day 1 of implementing vaccination. Furthermore, data suggest that the CFR reduction is associated with a concomitant reduction in the ARs and number of cases. The decrease in CFR is related to a proportional decrease in the number of deaths and the decreased number of cases.

RECOMMENDATIONS
Post-interventional CFR monitoring may constitute a parameter for measuring vaccination effectiveness and progress of the current pandemic or future pandemics. Furthermore, post-interventional CFR estimates can be used as a parameter for assessment effectiveness of interventions e.g.