Population-level effect of COVID-19 full vaccination coverage on transmission and mortality during Omicron variant dominance: a global longitudinal analysis

doi:10.21203/rs.3.rs-4934548/v1

Download PDF

Research Article

Population-level effect of COVID-19 full vaccination coverage on transmission and mortality during Omicron variant dominance: a global longitudinal analysis

https://doi.org/10.21203/rs.3.rs-4934548/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Aim

Given the higher transmissibility of the SARS-CoV2 Omicron variant and associated concerns about reduced vaccine effectiveness, we assessed the population-level impact of COVID-19 vaccination on viral transmission and mortality during the period of global Omicron dominance.

Subject and Methods

: We used a longitudinal dataset of 110 countries over 16 months (January 2022 to April 2023), representing the period of global Omicron dominance. Applying country-level random effects regression models, we assessed the effect of lagged monthly full vaccination coverage on the monthly rates of new confirmed COVID-19 cases and deaths across these countries, adjusting for a wide range of country characteristics and policy interventions. We obtained the data from open-access databases, including the WHO COVID-19 Dashboard and the Oxford COVID-19 Government Response Tracker.

Results

On average, each 1 percentage point increase in full vaccination coverage was associated with a 1.4% reduction (95% confidence interval [CI]: 0.1–2.8%, p = 0.035) in the rate of new cases and a 5% reduction (95% CI: 3.6–6.4%, p < 0.001) in the rate of deaths. This protective effect was graded across levels of vaccination coverage: compared to countries with vaccination coverages < 50%, countries with coverages of 50–59%, 60–69%, 70–79% and ≥ 80% had 20.5% (95% CI: -16.4–45.7%, p = 0.20), 53.8% (22.6–72.5%, p = 0.003), 54.3% (15.5–75.3%, p = 0.01), and 69.6% (38.7–84.9%, p < 0.001) lower rates of new cases respectively, on average.

Conclusion

Our findings underscore the crucial role of vaccination in mitigating the impact of pandemics, particularly during the emergence of highly transmissible variants like Omicron.

COVID-19

COVID vaccination

Omicron wave

Impact evaluation

Public health

In March 2020, the World Health Organization (WHO) declared the coronavirus disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a pandemic [1]. The pandemic produced a myriad of devastating health and socioeconomic impacts globally before it officially ended in May 2023 [2, 3]. In addition to a combination of non-pharmacological policy measures –including lockdowns and travel restrictions– a key aspect of the global response to the COVID-19 pandemic was mass vaccination campaigns [4]. Within ten months of the transcription of SARS-CoV-2 genome in January 2020, there were over 180 vaccine candidates at various stages of development across the world [5]. By December 2020, the large-scale administration of approved vaccines began in developed countries with the Pfizer/BioNTech mRNA vaccine, followed by Moderna’s mRNA vaccine and AstraZeneca’s vectored vaccine [1].

These vaccines initially recorded high efficacies in stage III trials – up to 90% protection against symptomatic COVID-19 in many cases [6–8]. However, their real-world efficacy decreased with passing time and the introduction of more virulent SARS-CoV-2 variants, necessitating booster doses [9, 10]. Five main SARS-CoV-2 variants of concern (VoC) were responsible for successive global waves of the pandemic. The first four of these (Alpha, Beta, Gamma and Delta variants) emerged across the world in late 2020 and dominated COVID-19 global epidemiology until early 2021, with generally increasing transmissibility, reduced vaccine efficacy, and worse clinical outcomes compared to their progenitors [10]. These were mostly replaced by the Omicron variant in November 2021, which recorded a significantly higher frequency of breakthrough infections and onward transmission between vaccinated individuals [11]. By February 2022, Omicron and its lineages accounted for 98% of all publicly available genetic sequences worldwide [12]. This variant continued to dominate the global COVID-19 epidemiology until the end of the pandemic, causing about two-thirds of the more than 768 million confirmed COVID-19 cases by 26 June 2023, and responsible for about 20% of the nearly 7 million recorded COVID-related deaths by that date [13, 14].

While COVID-19 vaccine efficacy has been demonstrated in randomized controlled trials (RCTs), there is a need for complementary population-level studies assessing the effectiveness of the global COVID-19 vaccination efforts. In contrast to the direct individual-level efficacy estimates from vaccine trials, population-level studies capture the direct and indirect effects of population-wide exposure to vaccination programmes [15]. In the context of the COVID-19 pandemic however, such studies have been lacking and limited in scope. A cross-sectional ecological analysis in the USA showed that counties with more persons fully vaccinated had substantially lower rates of COVID-19 cases and deaths [16]. Using time-trend data from four European countries, another study showed that higher levels of full vaccination coverage was associated with a reduced COVID-19 case-fatality rate [17]. A third study conducted in the early wave of the pandemic across 90 countries showed similar results [18]. The most recent study using data from the delta wave period in 45 high-income countries also reported similar results [19]. However, all of these studies used data from the earlier waves of the pandemic characterised by less infective strains, and are generally limited to specific countries or regions of the world.

No global studies have been conducted using data from the more recent waves of COVID dominated by the more infective Omicron variant [14], although calls have been made for such studies [16]. Furthermore, while two studies have assessed case fatality rate, there has been little focus on the effect of vaccination on COVID-19 transmission [17, 18]. Thus, using comprehensive time-series data for 110 countries between January 2022 (beginning of the global Omicron dominance) [12, 13] and May 2023 (official end of the pandemic) [3], we assessed the effect of the levels of full vaccination coverage on the rates of new COVID-19 cases and deaths across the world, while controlling for a broad set of country-level sociodemographic, economic, health status-related, and government policy characteristics.

Data Sources & Study period

The data for this study were obtained from Our World in Data (OWID) coronavirus repository, a frequently updated collection of COVID-19-related data on vaccinations, tests, confirmed cases, confirmed deaths, policy response, and other country-level characteristics, for over 200 countries/territories across the world [20, 21]. OWID data are collated from a range of official sources – including official governmental reports, the WHO COVID-19 Dashboard, Oxford COVID-19 Government Response Tracker, the Global Burden of Disease Collaborative Network, the United Nations Statistics Division, and the World Bank World Development Indicators – and have been extensively applied for COVID-19-related public health and policy research [18, 22–24]. The analysed dataset was downloaded from OWID (https://ourworldindata.org/coronavirus) on 29th of December, 2023. A description of the variables and their original sources is provided in the Online Resource (Table S1), and we subsequently reference the original data sources where applicable. The study period covered by this analysis ranges from 1 January 2022 (taken as the beginning of the global Omicron variant dominance, since Omicron and its lineages accounted for 98% of all publicly available genetic sequences worldwide by February 2022 [12, 13]) to 30 April 2023 (since the WHO declared the end of COVID-19 as a public health emergency on May 4, 2023 [3]). This is illustrated in Fig. 1, which shows the global average monthly new laboratory-confirmed COVID-19 cases (per 1 million population) and deaths (per 50 million population), with the study period highlighted in grey.

Outcomes

Two outcomes were assessed in this study: monthly new confirmed COVID-19 cases and monthly confirmed COVID-19-related deaths between January 1, 2022 and April 30, 2023. These country-level data were obtained from the WHO COVID-19 dashboard, which presents data submitted by WHO Member States [25]. These counts primarily reflect laboratory-confirmed cases and deaths, based upon WHO-published case definitions [25]. These cumulative outcomes were obtained as crude counts for the available countries and expressed as rates per million population for the analyses.

Exposure

For each country and in each month, we obtained estimates of national full vaccination coverage from OWID [21]. This is expressed as the total number of people in a country who received all doses prescribed by the country’s vaccination protocol per 100 people in the total population. In order to further investigate a potential “dose-response” relationship, we further categorized this variable into five mutually exclusive levels: countries with < 50%, 50–59%, 60–69%, 70–79%, and ≥ 80% vaccination coverage, as in a previous study conducted in the USA [16]. These continuous and categorical exposure variables were applied in regression analyses to observe the change in the rates of new cases and deaths in response to increases in full vaccination coverage. The categorized analysis also aimed to assess the point of attainment of herd immunity, which experts have hypothesized to be approximately 60 to 70% full vaccination coverage in the context of the COVID-19 pandemic [26, 27].

Covariates

We controlled for a broad range of country-level sociodemographic, economic, health-related and government policy characteristics. The sociodemographic characteristics include the most recent (by 2020) estimates of population size, proportion of population aged 65 years or older, median age of population, and population density, obtained from the United Nations (UN) Statistics Division. Economic factors include the most recent estimates of gross domestic product (GDP) per capita at purchasing power parity (constant 2011 international dollars), also obtained from the UN. We further adjusted for factors representing the health status of populations, viz.: life expectancy at birth in 2019; the national diabetes prevalence (% of population aged 20 to 79 with diagnosed or undiagnosed type 1 or type 2 diabetes in 2017), obtained from the International Diabetes Federation [28]; the national death rate from cardiovascular disease in 2017 (per 100,000 people), obtained from Global Burden of Disease Collaborative Network [29]; and the number of hospital beds per thousand population, obtained from the Organisation for Economic Co-operation and Development (OECD), Eurostat, World Bank, national government records and other sources. As a measure of non-pharmaceutical COVID-19 policy interventions, we also controlled for the Government Response Stringency Index, obtained from the Oxford COVID-19 Government Response Tracker (CGRT) [30]. The stringency index is a composite score measuring the intensity of nine non-pharmaceutical policy response indicators: school closures, workplace closures, cancellation of public events, restrictions on gathering size, closures of public transport, stay-at-home requirements, restrictions to internal movement, restrictions on international travel, and coordinated public information campaigns. Lastly, since OWID stopped updating its global COVID-19 testing dataset in June 2022 [20], we accounted for the testing volume in each country by calculating the total tests per 1,000 people in the year preceding the study period (January to December, 2021) and assigning this value to each month in the time series.

Missing Data

The study includes 110 countries for which data was available for all the variables. Except for vaccination coverage and stringency index, the data for the two outcomes all other covariates were 100% complete. While available for all 110 countries, vaccination coverage and stringency index were 77% and 75% complete, respectively, with data missing for some months within countries. Instead of listwise deletion, we imputed missing vaccination coverage data using linear interpolation between non-missing values within countries. We further deterministically imputed trailing missing values with the preceding non-missing value, with the conservative assumption that vaccination coverage does not decrease over time. Since the Oxford CRGT provided stringency index data for 2020 to 2022 (and not 2023), we had missing stringency data for all countries for the included four months of 2023 (January to April). As before, we addressed this using imputation method. First, we deterministically assigned a stringency score of zero for all countries for the last month (April 2023), since most countries had lifted restrictions at this point [31]. Then, we applied linear interpolation to fill in stringency values for January, February and March of 2023, since countries gradually “eased” out of the restrictive policies [30, 31]. In sensitivity analyses, we tested the robustness of this assumption by imputing a stringency score of zero for all months in 2023.

Statistical Analysis

We first assess the characteristics of the country-months of observations using descriptive statistics and graphs, stratified by the categories of full vaccination coverage. Next, we fit random effects negative binomial models on the panel data to assess the association of full vaccination coverage with the rates of COVID-19 cases and deaths across the world, with random intercepts for continents and countries, allowing the baseline rates of the outcomes to vary across countries and accounting for correlation of observations within continents and countries. We assumed an autoregressive correlation structure between observations within each country over time. In the regression models, we lagged vaccination coverage and stringency index by one period, since previous research identified a 40-day and 14-day lag in their respective effects on COVID-19 outcomes [32]. We used an offset to account for (the natural log of) country population size. All models adjusted for the country-level covariates, in addition to dummy variables for each month of the year to account for temporal or seasonal variations in the outcomes. Given our Poisson-distributed COVID-19 outcome $\:{y}_{ij,t}$ for country i located in continent j at time t, our multi-level negative binomial regression models took the following functional form:

$$\:In\left({y}_{ij,t}\right)=\alpha\:+{\beta\:vac}_{ij,t-1}+{\sum\:}_{k=1}^{K}{\gamma\:}_{k}{x}_{k,ij}+{{\beta\:}^{1}stringency}_{ij,t-1}+\:{\vartheta\:}_{month}\:+\text{\:log}\left(pop\right)+\:{\theta\:}_{i}+{e}_{it}\:$$

Where $\:{vac}_{i,t-1}\:$denotes the one month-lagged vaccine coverage for country i in month t-1; $\:{x}_{k,ij}$ represents time-invariant country characteristic k (1, ..., K) for country i in continent j (e.g., population density); $\:{stringency}_{ij,t-1}$ refers to time-varying non-pharmaceutical policy intervention represented by the stringency index for country i at time t-1; and $\:{\vartheta\:}_{month}$ represents dummy variables for each month of the year. Furthermore, $\:{\theta\:}_{i}$ is an error term accounting for between-country variation in the outcome and $\:{e}_{it}$ is the error term for each country-month observation, which is uncorrelated with $\:{\theta\:}_{i}$ or any covariates. We treated $\:{\theta\:}_{i}$ as random using random-effects models, primarily because the use of country fixed effects precluded us from including time-invariant country characteristics in the models [18]. Nonetheless, we also used fixed effects models in sensitivity analyses. All analyses were performed in R statistical program (version 4.3.3 - Windows). The code used for the analyses are provided in a GitHub repository (https://github.com/pharmsteve/COVID).

Sensitivity Analyses

We conducted four sensitivity analyses to test the robustness of our study results to important assumptions. First, instead of the random effect models used in the primary analysis, we fit negative binomial models with country and time fixed effects, accounting for across-country variations in time-invariant factors influencing the study outcomes, as well as time-varying factors that are constant across countries, respectively. Second, we addressed missing stringency data for the year 2023 by imputing a score of zero for all four months in 2023. Third, we varied the lags for vaccination and stringency to 0 and 2 months, instead of the one-month lag implemented in the main analysis. Lastly, we began the study period from November 2021, since the Omicron variant was first identified as a VOC by the WHO in late November 2021, although it only became the globally dominant variant at the start of 2022 [12, 13].

Country-level characteristics

This study sample consisted of 110 countries for which data was available for all variables. This includes countries from across all continents of the world, with a total represented population of 5.3 billion people. These countries generated 1,756 country-months of data between January 2022 and April 2023. Table 1 shows the characteristics of these country-months of data overall and stratified by the categories of full COVID-19 vaccination coverage in country-month observations. It can be observed from Table 1 that 80% of the observations from the African continent were in the had < 50% full vaccination coverage, while the observations from Europe, Oceania and the Americas mostly had ≥ 70% vaccination coverage. This is further illustrated in the map in Fig. 2, which shows the spatial distribution of full vaccination coverage across the globe at the start of the study period (January 1, 2022). Furthermore, the Wilcoxon rank sum tests in Table 1 suggest significant positive associations of vaccination coverage with developmental indices such as Human development index, GDP per capita and life expectancy, as well as with the number of COVID-19 tests and hospital beds.

Table 1

Characteristics of the 1,756 country-month observations overall and grouped by percentage of the population fully vaccinated against COVID-19
	OVERALL	By percent of eligible population fully vaccinated against COVID-19 within country-month (Jan 2022 – April 2023)
	OVERALL	< 50%	50–59%	60–69%	70–79%	≥ 80%	P-val
	Median (IQR)
Number of country-months	1,756	528 (30%)	209 (12%)	345 (20%)	342 (19%)	332 (19%)
Median Full Vaccination coverage (%)	65 (46, 78)	34 (22, 43)	54 (52, 56)	66 (64, 68)	75 (73, 78)	85 (82, 87)	< 0.001
Continent (row %)
Africa	254	203 (80%)	32 (13%)	19 (7.5%)	0 (0%)	0 (0%)
Asia	526	102 (19%)	61 (12%)	156 (30%)	57 (11%)	150 (29%)
Europe	560	116 (21%)	63 (11%)	94 (17%)	197 (35%)	90 (16%)
North America	192	64 (33%)	32 (17%)	52 (27%)	18 (9.4%)	26 (14%)
Oceania	48	0 (0%)	0 (0%)	16 (33%)	17 (35%)	15 (31%)
South America	176	43 (24%)	21 (12%)	8 (4.5%)	53 (30%)	51 (29%)
Outcome Variables
Rate of confirmed COVID-19 cases per million	1,110 (118, 6,209)	273 (26, 1,531)	444 (65, 2,930)	693 (74, 3,439)	5,072 (870, 17,665)	3,525 (893, 11,128)	< 0.001
Rate of confirmed COVID-19 deaths per million	7 (0, 28)	3 (0, 15)	4 (0, 20)	3 (0, 19)	25 (5, 53)	14 (2, 31)	< 0.001
Sociodemographic Variables
Population density (people per Km²)	85 (31, 158)	87 (44, 143)	74 (16, 124)	80 (44, 308)	83 (22, 137)	96 (25, 227)	< 0.001
Median age (years)	32 (28, 41)	26 (20, 37)	28 (25, 40)	32 (29, 40)	40 (34, 42)	36 (32, 42)	< 0.001
Percentage of population aged 65 or older (%)	8 (5, 17)	5 (3, 11)	7 (4, 14)	8 (5, 15)	15 (9, 20)	14 (5, 19)	< 0.001
Economic and Developmental Indices
Gross domestic product per capita (X $1,000)*	17 (8, 33)	8 (3, 16)	11 (7, 24)	17 (9, 30)	36 (19, 45)	36 (21, 45)	< 0.001
Human development index	0.8 (0.7, 0.9)	0.7 (0.6, 0.8)	0.7 (0.7, 0.8)	0.8 (0.7, 0.9)	0.9 (0.8, 0.9)	0.9 (0.8, 0.9)	< 0.001
Life expectancy	76 (72, 80)	72 (64, 74)	74 (70, 77)	77 (72, 78)	81 (77, 82)	80 (77, 83)	< 0.001
Health Status Variables
Adult diabetes prevalence in 2017	7 (5, 10)	7 (5, 10)	7 (6, 8)	7 (5, 11)	6 (5, 8)	7 (6, 10)	0.2
Death rate from cardiovascular disease in 2017 (per 100,000)	219 (145, 304)	280 (219, 371)	254 (202, 335)	260 (167, 343)	134 (118, 156)	138 (108, 201)	< 0.001
Total COVID-19 tests in 2021 (per million)	747 (209, 1,579)	335 (65, 596)	306 (197, 910)	898 (267, 1,909)	1,448 (646, 2,909)	1,391 (766, 2,366)	< 0.001
Hospital beds per thousand	3 (1, 4)	2 (1, 4)	2 (1, 6)	3 (1, 5)	3 (2, 5)	3 (2,3)	< 0.001
Government COVID-19 policy strength
Stringency index (%)†	14 (9, 32)	21 (11, 37)	11 (6, 32)	18 (11, 32)	11 (8, 27)	12 (8, 26)	0.046
P-values computed from Wilcoxon rank sum tests comparing medians and Chi-square test for “Continent”. *GDP per capita is expressed as constant 2011 international USD (at purchasing power parity). †Stringency Index is a composite score measuring the intensity of nine non-pharmaceutical policy response indicators on a scale of 1 to 100.

Multivariable modeling

Table 2 presents the estimates from multivariable random effects negative binomial regression models assessing the association between full vaccination coverage and the rates of confirmed COVID-19 cases and deaths across the world, during the study period. In general, after adjusting for a wide range of country-level factors and seasonality, each one percentage point increase in full vaccination coverage is associated with a 1.4% reduction (95% confidence interval [CI]: 0.1–2.8%, p = 0.035) in the rate of new cases and a 5% reduction (95% CI: 3.6–6.4%, p < 0.001) in the rate of new deaths from COVID-19.

We also assessed these associations for graded categories of vaccination coverage. Compared to a full vaccination coverage of less than 50%, vaccination coverages of 50–59%, 60–69%, 70–79% and ≥ 80% were associated with 20.5% (95% CI: -16.4–45.7%, p = 0.20), 53.8% (22.6–72.5%, p = 0.003), 54.3% (15.5–75.3%, p = 0.01), and 69.6% (38.7–84.9%, p < 0.001) lower rates of new cases, respectively. Similarly, the death rates were 35.1% (95% CI: 7.9–54.3%, p = 0.02), 69.3% (51.1–80.8%, p < 0.001), 71.0% (49.1–83.5%, p < 0.001), and 81.1% (64.1–90.1%, p < 0.001) lower in the respective groups compared to the reference group.

When we used fixed effect models instead of the primary random effects models, the associations were considerably stronger (Online Resource, Table S2). The dot-and-whisker plots in Fig. 3 graphically depicts these associations for the random effects models for COVID-19 cases and deaths, suggesting a dose-response relationship between full vaccination coverage and the COVID-19 outcomes. The “human development index” variable was dropped from the models due to a high multicollinearity with GDP per capita and life expectancy. Similarly, the “proportion aged 65 years or older” variable was dropped due to collinearity with “median age”. The previously described sensitivity analyses produced results that are generally consistent with those of the main analysis (Online Resource, Tables and Figures S2 to S6).

Table 2

Regression estimates from four random effects negative binomial models assessing the association between one-month-lagged full vaccination coverage (%) and the rates of new COVID-19 cases and deaths during the period of global Omicron variant dominance.
Variable	CASES			DEATHS
Variable	Rate Ratio	95% CI^†	p-value	Rate Ratio	95% CI	p-value
Full vaccination coverage (%)	0.986	0.97, 0.99	0.035	0.95	0.94, 0.96	< 0.001
Categorized coverage^‡
<50% (reference)	1.00	-	-	1.00	-	-
50–59%	0.80	0.54, 1.16	0.200	0.65	0.46, 0.92	0.016
60–69%	0.46	0.28, 0.77	0.003	0.31	0.19, 0.49	< 0.001
70–79%	0.46	0.25, 0.85	0.013	0.29	0.17, 0.51	< 0.001
≥80%	0.30	0.15, 0.61	< 0.001	0.19	0.09, 0.36	< 0.001
Stringency index	1.06	1.06, 1.07	< 0.001	1.05	1.05, 1.06	< 0.001
GDP per capita	1.00	1.00, 1.00	0.046	1.00	1.00, 1.00	0.200
Life expectancy	1.13	1.00, 1.35	< 0.001	1.22	1.10, 1.35	< 0.001
Population density	1.00	1.00, 1.01	0.600	1.00	0.99, 1.00	0.700
Median age	1.07	1.00, 1.14	0.06	1.16	1.08, 1.25	< 0.001
Hospital beds per thousand	1.14	1.01, 1.29	0.04	0.99	0.87, 1.13	0.900
Adult diabetes prevalence in 2017	1.03	0.94, 1.12	0.60	0.99	0.91, 1.08	0.800
Death rate from cardiovascular disease in 2017 (per 100,000)	0.99	0.99, 1.00	0.60	0.99	0.99, 1.00	0.600
Total COVID-19 tests in 2021 (per million)	1.00	1.00, 1.00	0.06	1.00	1.00, 1.00	0.400
The sample includes 1,756 country-months of observations from 110 countries between Jan 2022 and April 2023. The models include an offset (log of population) and dummy variables for months of the year, with standard errors clustered at the country level. ^†95% CI = 95% confidence interval. ‡Obtained from an identical model using the categorized vaccination coverage variable instead of the continuous one.

The emergence of the Omicron variant posed significant challenges to global public health systems due to its increased transmissibility, raising concerns about reduced vaccine effectiveness and necessitating booster vaccine doses [13, 16]. Reassuringly, our study indicated an inverse association between full vaccination coverage and the rates of both COVID-19 cases and deaths across the world during this period of global Omicron variant dominance (Jan 2022 to April 2023). In addition, this effect appears to generally increase with higher full vaccination coverage. The sensitivity analyses suggest that our findings are robust to a range of assumptions regarding the modelling choices and the handling of missing data.

These results are consistent with those from other studies assessing the population-level effect of vaccination coverage in other geographical regions and earlier waves of the COVID-19 pandemic. Most recently, a study exploring COVID-19 epidemiology during the delta variant dominance period reported that higher vaccination coverage was protective against peak case rates (odds ratio 0.95, 95% CI 0.91–0.99) and against peak death rates (odds ratio 0.96, 95% CI 0.91–0.99) in 45 high-income countries [19]. Two other international studies conducted in earlier waves of the pandemic – one including four European countries [17] and the second including 90 countries globally [18] – reported that higher levels of full vaccination coverage was associated with a reduced COVID-19 case-fatality rate [17]. Lastly, a cross-sectional ecological analysis in the USA during Delta variant predominance showed that US counties with greater proportions of persons fully vaccinated had substantially lower rates of COVID-19 cases and deaths [16]. As in our analysis, this study reported a “dose-response” effect across the same categories of full vaccination coverage used in our study.

We further aimed to assess the level of vaccine coverage at which herd immunity against COVID-19 was achieved globally. Herd immunity is attained when a virus can no longer spread effectively because a sufficient proportion of the population has acquired immunity either through vaccination or naturally through infection [27]. Vaccination experts and modellers have proposed that, in the context of COVID-19, herd immunity is likely to be attained when approximately 60 to 70% of eligible individuals are fully vaccinated [26, 27]. This projection is supported by our study, where significant decreases in the rates of transmission and deaths were observed at full vaccination coverages of 60% or higher. Similar observations were also reported in a study conducted across counties of the USA [16]. While this 60% threshold aligns with previous estimates and models proposed by public health experts [26, 27], it is important to note that the concept of herd immunity is dynamic and may vary depending on factors such as the transmissibility of a viral strain, efficacy of specific vaccines, and population demographics. Additionally, achieving herd immunity in the global context requires sustained efforts to ensure equitable access to vaccines, address vaccine hesitancy, and overcome logistical challenges in vaccine distribution and administration.

Our study has some important limitations. First, OWID’s global COVID-19 testing data was only available until June 2022, which means that we did not have testing data for some months in the study period. We addressed this limitation by computing each country’s number of tests per thousand people in the year preceding the study period (January to December 2021) and assigning this number to every month for each country during the study period. However, given the potential correlation between testing volume and the rates of confirmed cases and deaths, we anticipate some residual confounding from this factor. Secondly, not all covariate data came from the same year, and some may be deemed rather dated. However, despite limited data availability, we endeavoured to control for a wide range of potential confounders, and all the data used were measured within 5 years of the start of the study period. Lastly, the generalizability of our results may be limited by our inclusion of only 110 countries for which data was available for all the variables of interest. However, Table 1 suggests that our sample is representative of the world in terms of geography and developmental indices. Some important strengths of our study include the longitudinal design, wealth of covariate data, large sample, and multiple sensitivity analyses to test the robustness of our results to our modelling assumptions and choices.

Our study suggests that on a global scale, higher levels of full vaccination coverage were associated with lower rates of new COVID-19 cases and deaths during the period of global Omicron variant dominance. These results were robust to modelling and other data processing assumptions. Our findings underscore the crucial role of vaccination in mitigating the impact of the pandemic, particularly during the emergence of highly transmissible variants like Omicron.

Funding: This research was supported by the Lady Davis Institute / Eileen and Louis Dubrovsky Graduate Fellowship Award, as well as the McGill Faculty of Medicine / George G Harris doctoral fellowship.

Competing interests: The authors have no relevant financial or non-financial interests to disclose.

Ethics approval: This study was conducted at McGill University, an institution under the purview of the Canadian Tri-Council Policy Statement on the Ethical Conduct of Research Involving Humans (TCPS-2). According to the TCPS-2, the current study is exempt from Research Ethics Board (REB) review since it relies exclusively on the secondary use of publicly accessible anonymous information, and because such data “is in the public domain and the individuals to whom the information refers have no reasonable expectation of privacy” (TCPS-2, Article 2.2).

Data Availability: All data used for this study are downloadable from the Our World in Data (OWID) COVID-19 database (https://ourworldindata.org/coronavirus) at no cost.

Code availability: All codes used for this analysis are available at a repository: https://github.com/pharmsteve/COVID.git.

Author’s contributions: All four authors were involved in the conception and design of the study. S.C.O and D.C.O carried out the statistical analysis. All authors jointly drafted and reviewed the manuscript, and approved the final version for submission.

Acknowledgements: The authors acknowledge OWID for compiling and freely providing the comprehensive data used in this study. We also acknowledge technical support from the Partnership Grant Program of the Social Sciences and Humanities Research Council of Canada (SSHRC) for the Consortium on Analytics for Data-Driven Decision-Making (CAnD3) (#895-2020-1013; PI: Amélie Quesnel-Vallée).

Krammer F. The role of vaccines in the COVID-19 pandemic: what have we learned? Semin Immunopathol. 2023;1–18. https://doi.org/10.1007/s00281-023-00996-2.
Sarkodie SA, Owusu PA. Global assessment of environment, health and economic impact of the novel coronavirus (COVID-19). Environ Dev Sustain. 2021;23:5005–15. https://doi.org/10.1007/s10668-020-00801-2.
Sarker R, Roknuzzaman A, Nazmunnahar, Shahriar M, Hossain MJ, Islam MR. The WHO has declared the end of pandemic phase of COVID-19: Way to come back in the normal life. Health Sci Rep. 2023;6:e1544. https://doi.org/10.1002/hsr2.1544.
Onyeaka H, Anumudu CK, Al-Sharify ZT, Egele-Godswill E, Mbaegbu P. COVID-19 pandemic: A review of the global lockdown and its far-reaching effects. Sci Prog. 2021;104:00368504211019854. https://doi.org/10.1177/00368504211019854.
Krammer F. SARS-CoV-2 vaccines in development. Nature. 2020;586:516–27. https://doi.org/10.1038/s41586-020-2798-3.
Baden LR, El Sahly HM, Essink B, Kotloff K, Frey S, Novak R, et al. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med. 2021;384:403–16. https://doi.org/10.1056/NEJMoa2035389.
Polack FP, Thomas SJ, Kitchin N, Absalon J, Gurtman A, Lockhart S, et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med. 2020;383:2603–15. https://doi.org/10.1056/NEJMoa2034577.
Logunov DY, Dolzhikova IV, Shcheblyakov DV, Tukhvatulin AI, Zubkova OV, Dzharullaeva AS, et al. Safety and efficacy of an rAd26 and rAd5 vector-based heterologous prime-boost COVID-19 vaccine: an interim analysis of a randomised controlled phase 3 trial in Russia. Lancet. 2021;397:671–81. https://doi.org/10.1016/S0140-6736(21)00234-8.
Barda N, Dagan N, Cohen C, Hernán MA, Lipsitch M, Kohane IS, et al. Effectiveness of a third dose of the BNT162b2 mRNA COVID-19 vaccine for preventing severe outcomes in Israel: an observational study. Lancet. 2021;398:2093–100. https://doi.org/10.1016/S0140-6736(21)02249-2.
Thye AY-K, Law JW-F, Pusparajah P, Letchumanan V, Chan K-G, Lee L-H. Emerging SARS-CoV-2 variants of concern (VOCs): an impending global crisis. Biomedicines. 2021;9:1303. https://doi.org/10.3390/biomedicines9101303.
Carreño JM, Alshammary H, Tcheou J, Singh G, Raskin AJ, Kawabata H, et al. Activity of convalescent and vaccine serum against SARS-CoV-2 Omicron. Nature. 2022;602:682–8. https://doi.org/10.1038/s41586-022-04399-5.
World Health Organization. (2023a) Statement on the update of WHO’s working definitions and tracking system for SARS-CoV-2 variants of concern and variants of interest. https://www.who.int/news/item/16-03-2023-statement-on-the-update-of-who-s-working-definitions-and-tracking-system-for-sars-cov-2-variants-of-concern-and-variants-of-interest. Accessed Jan 25th 2024.
Dhama K, Nainu F, Frediansyah A, Yatoo MI, Mohapatra RK, Chakraborty S, et al. Global emerging Omicron variant of SARS-CoV-2: Impacts, challenges and strategies. J Infect Public Health. 2023;16:4–14. https://doi.org/10.1016/j.jiph.2022.11.024.
Flisiak R, Zarębska-Michaluk D, Dobrowolska K, Rorat M, Rogalska M, Kryńska JA, et al. Change in the clinical picture of hospitalized patients with COVID-19 between the early and late period of dominance of the omicron SARS-CoV-2 variant. J Clin Med. 2023;12:5572. https://doi.org/10.3390/jcm12175572.
Hanquet G, Valenciano M, Simondon F, Moren A. Vaccine effects and impact of vaccination programmes in post-licensure studies. Vaccine. 2013;31:5634–42. https://doi.org/10.1016/j.vaccine.2013.07.006.
McLaughlin JM, Khan F, Pugh S, Swerdlow DL, Jodar L. County-level vaccination coverage and rates of COVID-19 cases and deaths in the United States: An ecological analysis. Lancet 2022; Reg Health–Americas 9. https://doi.org/10.1016/j.lana.2022.100191
Tan J, Wu Z, Gan L, Zhong Q, Zhu Y, Li Y, et al. Impact of vaccination and control measures on the fatality of COVID-19: An ecological study. J Epidemiol Global Health. 2022;12:456–71. https://doi.org/10.1007/s44197-022-00064-2.
Liang L-L, Kuo H-S, Ho HJ, Wu C-Y. COVID-19 vaccinations are associated with reduced fatality rates: Evidence from cross-county quasi-experiments. J Global Health. 2021;11. https://doi.org/10.7189/jogh.11.05019.
Atherstone CJ, Guagliardo SAJ, Hawksworth A, O’Laughlin K, Wong K, Sloan ML, et al. COVID-19 Epidemiology during Delta Variant Dominance Period in 45 High-Income Countries, 2020–2021. Emerg Infect Dis. 2023;29:1757. https://doi.org/10.3201/eid2909.230142.
Edouard Mathieu HR, Lucas Rodés-Guirao C, Appel C, Giattino J, Hasell B, Macdonald S, Dattani D, Beltekian, Esteban Ortiz-Ospina and Max Roser. (2020) Coronavirus Pandemic (COVID-19) [dataset]. https://ourworldindata.org/coronavirus (accessed 29 December 2023).
Mathieu E, Ritchie H, Ortiz-Ospina E, Roser M, Hasell J, Appel C, et al. A global database of COVID-19 vaccinations. Nat Hum Behav. 2021;5:947–53. https://doi.org/10.1038/s41562-021-01122-8.
Aknin LB, Andretti B, Goldszmidt R, Helliwell JF, Petherick A, De Neve J-E, et al. Policy stringency and mental health during the COVID-19 pandemic: a longitudinal analysis of data from 15 countries. Lancet Public Health. 2022;7:e417–26. https://doi.org/10.1016/S2468-2667(22)00060-3.
Thakur V, Bhola S, Thakur P, Patel SKS, Kulshrestha S, Ratho RK, et al. Waves and variants of SARS-CoV-2: understanding the causes and effect of the COVID-19 catastrophe. Infection. 2021;1–16. https://doi.org/10.1007/s15010-021-01734-2.
Asundi A, O'Leary C, Bhadelia N. Global COVID-19 vaccine inequity: The scope, the impact, and the challenges. Cell Host Microbe. 2021;29:1036–9. https://doi.org/10.1016/j.chom.2021.06.007.
World Health Organization. (2023b) WHO Coronavirus (COVID-19) dashboard > Cases. https://data.who.int/dashboards/covid19/cases. Accessed Jan 25 2024.
Rubin R. Difficult to determine herd immunity threshold for COVID-19. JAMA. 2020;324:732–732. https://doi.org/10.1001/jama.2020.14778.
Randolph HE, Barreiro LB. Herd immunity: understanding COVID-19. Immunity. 2020;52:737–41. https://doi.org/10.1016/j.immuni.2020.04.012.
Cho NH, Shaw J, Karuranga S, Huang Y, da Rocha Fernandes J, Ohlrogge A et al. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes research and clinical practice. 2018; 138:271–81. https://doi.org/10.1016/j.diabres.2018.02.023
James SL, Abate D, Abate KH, Abay SM, Abbafati C, Abbasi N, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–858. https://doi.org/10.1016/S0140-6736(18)32279-7.
Hale T, Angrist N, Goldszmidt R, Kira B, Petherick A, Phillips T et al. A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker) Nature human behaviour. 2021; 5:529–38. https://doi.org/10.1038/s41562-021-01079-8
Tian C, Balmer L, Tan X. COVID-19 lessons to protect populations against future pandemics by implementing PPPM principles in healthcare. EPMA J. 2023;14:329–40. https://doi.org/10.1007/s13167-023-00331-7.
Li H, Wang L, Zhang M, Lu Y, Wang W. Effects of vaccination and non-pharmaceutical interventions and their lag times on the COVID-19 pandemic: Comparison of eight countries. PLoS Negl Trop Dis. 2022;16:e0010101. https://doi.org/10.1371/journal.pntd.0010101.

No competing interests reported.

SupplementaryMaterial.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Population-level effect of COVID-19 full vaccination coverage on transmission and mortality during Omicron variant dominance: a global longitudinal analysis

Status:

Version 1

Abstract

Aim

Subject and Methods

Results

Conclusion

Figures

INTRODUCTION

METHODS

Data Sources & Study period

Outcomes

Exposure

Covariates

Missing Data

Statistical Analysis

Sensitivity Analyses

RESULTS

Country-level characteristics

Multivariable modeling

DISCUSSION

CONCLUSIONS

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1