Epidemiological impact of covid-19 fully vaccinated population
The epidemiological impact of increased vaccination on covid-19 pandemic is investigated by a study of the long-term cross-sectional correlation of 98 countries using the database of Our World in Data.
Higher percentages of fully vaccinated population are found in 94 countries to be associated with higher increases in infections in 2021 relative to 2020.
High levels of vaccination have been the main strategy in many countries around the world to reduce transmission of the covid-19 disease. A recent letter indicated that the epidemiological relevance of covid-19 vaccinated persons is increasing, suggesting that decision makers cannot ignore the vaccinated population as a relevant source of transmission when deciding public health policy.
Subramanian and Kumar recently found that increases in covid-19 cases are unrelated to levels of vaccination across 68 countries, as well as across 2947 counties in the United States. The absence of short-term (7 days) correlation in their study shows vaccination has no immediate epidemiological impact.
However, due to differences in definitions of vaccination status, and possible lags in biological responses, data reporting and publication, the epidemiological impact of vaccination may only be seen in a longer-term correlation, which is explored here. Instead of only a limited selection of countries, epidemiological impact of vaccination is examined globally for 98 countries.
A covid-19 data source has been downloaded from Our World in Data (OWID) for all countries on November 5, 2021, to construct a 98-country dataset using two filtering criteria: (1) the country is necessary to have a population exceeding two million and (2) the country is necessary to have submitted required data to OWID on November 1 or October 30 (the end date), where alternative dates are to allow for time zone differences in data collection.
On OWID a person is considered fully vaccinated if they have received a single-dose vaccine or both doses of a two-dose vaccine. The data are only available for countries which report the breakdown of doses administered by first and second doses.
In this paper, vaccination status is defined by the percentage of population fully vaccinated at the end date. Epidemiological impact of vaccination is defined by the 2021 change in reported cases measured by the total number of reported cases per thousand population in 2021 (to November 1) minus that of 2020. This 2021 change in cases quantifies impact by the long-term change in covid-19 transmissions, with and without, vaccination. The 2020 and 2021 data periods have roughly the same lengths of about ten months.
The data show that reported case increases in 2021 with vaccination, have higher covid transmissions than 2020 without vaccination, in 94 out of 98 countries or entities, with the exceptions being, Panama, Hong Kong, Saudi Arabia, and Switzerland, which have slightly fewer cases in 2021.
The reported 2021 case increases over 2020 could be understated, because vaccinated cases (breakthrough cases) had to satisfy a higher PCR test criterion (CT value < 28) than unvaccinated in many countries, thus biasing the data to fewer vaccinated cases in 2021 than would be otherwise. The observed increase in 2021 covid transmission is robust against this source of data bias.
The fact that most countries in the dataset have increased 2021 total cases from 2020, is consistent with a large increase in the global aggregate, which saw 2020 total cases of about 84 million nearly doubled to give 2021 total cases (at November 1) of about 164 million. On geographic distribution over the globe, there is a 45 percent positive correlation between vaccination status and 2021 increase in covid transmission. More fully vaccinated countries have tended to have higher increases in the number of covid cases.
Hence, so far, covid vaccines may have exacerbated the pandemic, enhancing the risk of infection, a phenomenon which could possibly be explained by escape from vaccine-mediated immunity due to viral mutations.
On standard analysis of the relationship, the linear regression between the variables is statistically significant with R-square 0.2, F-statistics 24.2, a p-value < 0•0001 and an equation: y = 3.89 + 0.37x, where y is the rate of increased infection in 2021 over 2020, x is the vaccination status or percent of population fully vaccinated. The t-values for the intercept and coefficient are respectively 1.1 and 4.9.
The most prominent outlier is Mongolia, which had nearly zero cases in 2020, but undertook a strong vaccination drive starting in April 2021. In contrast, central African countries which continued in 2021 with limited vaccination have remained substantially covid-free.
The data show that fully vaccinated populations have a strong epidemiological impact globally, not only in selected countries, but also over time and in most countries. While correlation is not causation, correlation is a necessary but not sufficient condition for causality.
The data with significant adverse correlation clearly falsify the assumption that universal vaccination can cause a reduction in covid transmission and end the pandemic.
Rather, the data suggest other epidemiological control measures such as prophylaxis and therapy traditionally used successfully for coronaviruses, should be encouraged globally by health policy makers.
May 15, 2022