key: cord-0555910-3bzmwkzw authors: Rughinis, Cosima; Vulpe, Simona-Nicoleta; Flaherty, Michael G.; Vasile, Sorina title: Vaccination, life expectancy, and trust: Patterns of COVID-19 vaccination rates around the world date: 2022-02-14 journal: nan DOI: nan sha: 395d1fea0dbcfd942e82b3d6b60d9e840ec6b98f doc_id: 555910 cord_uid: 3bzmwkzw We estimate patterns of covariation between COVID-19 vaccination rates and a set of widely used indicators of human, social, and economic capital across 146 countries in July 2021 and February 2022. About 70% of the variability in COVID-19 vaccination rates worldwide can be explained by differences in the Human Development Index (HDI) and, specifically, in life expectancy at birth, one year after the campaign debut. Trust in doctors and nurses adds predictive value beyond the HDI, clarifying controversial discrepancies between vaccination rates in countries with similar levels of human development and vaccine availability. Cardiovascular disease deaths, an indicator of general health system effectiveness, and infant measles immunization coverage, an indicator of country-level immunization effectiveness, are also significant, though weaker, predictors of COVID-19 vaccination success. The metrics of economic inequality, perceived corruption, poverty, and inputs into the health system have strong bivariate correlations with COVID-19 vaccination but no longer remain statistically significant when controlling for the HDI. Our analysis identified the contours of a social structure that sustains life and is reproduced through this process. COVID-19 vaccines have proven to be part of the Matthew effect of accumulating advantages and aggravating disadvantages that the pandemic inflicted on societies and communities across the world. At the same time, the remaining variability in vaccination success that cannot be pinned down through these sets of metrics points to a considerable scope for collective and individual agency in a time of crisis. The mobilization and coordination in the vaccination campaigns of citizens, medical professionals, scientists, journalists, and politicians, among others, account for at least some of this variability in overcoming vaccine hesitancy and inequity. Comedian Dave Barry recalled his mother telling him, "Son, it is better to be rich and healthy than poor and sick" 1 . This still holds when examining COVID-19 vaccination patterns worldwide. In this paper, we discuss the relative contribution to predicting COVID-19 and measles vaccination rates of a set of widely used, publicly available indicators of human, social, and economic capital. There has been a significant increase in life expectancy over the last two hundred years in many societies. Humankind has become more adept, collectively, to sustain life for its members, although externalities, in terms of climate impact, have begun to raise doubt on the longer-term perspectives of this accomplishment. Life expectancy serves as a synthetic measure of the capacity of society to prevent death in a certain period. Given that the avoidance of death is one of humankind's major goals, life expectancy is, therefore, a useful metric to capture the effectiveness of social organization for public health at a certain time and place. Vaccination has played a considerable role in reducing the mortality inflicted by preventable diseases 2 over the last two centuries. Vaccines have been, therefore, an important cause of the recent increase in life expectancy across the world. This also holds true for the COVID-19 pandemic, which has visibly lowered life expectancy in most countries 3 , 4 . There is convincing evidence that vaccination against COVID-19 has prevented numerous deaths globally 5 . At the same time, rates of vaccination have varied widely during the pandemic. Societal resources shape a collectivity's ability to immunize its members against infection through vaccination 6 . However, COVID-19 vaccination has been unevenly implemented because of differences in availability of vaccines, uneven logistics of vaccine distribution, and people's variable trust in vaccines and mainstream science and expertise 7 , 8 , 9 , 10 , 11 . In this paper, we explore and discuss the correlation between the success of vaccination campaigns against COVID-19 in mid-2021 and early 2022 and pre-pandemic life expectancy (estimated in 2019), alongside other measures of human, social, and economic capital, at country-level. Our study aimed to answer an essential question: What can such broad patterns of co-variation in vaccination success tell us about the social structures and forms of agency that keep people alive? Human, social, and economic resources have been of utmost importance in COVID-19 vaccination. They have facilitated earlier access to vaccines and powered the required logistics of a large-scale vaccination campaign. Several studies signaled a positive association between coverage of COVID-19 vaccination, the Human Development Index (HDI), and gross domestic product (GDP) per capita 12 , 13 , 14 . Education and GDP per capita also contribute to the speed of the COVID-19 vaccination campaign 15 . A positive correlation between measles vaccination and HDI has also been noted 16 . Trust in the state and in the health system has been associated with greater compliance with COVID-19 restrictions in Europe 17 . Trust in medical and scientific experts has been a strong correlate of pro-vaccination attitudes in general 18 , 19 , 20 , 21 and of the declared intention to receive a COVID-19 vaccine internationally 22 , 23 , 24 . Social and economic inequality has been associated with lower COVID-19 vaccination rates aggregated at country-level 25 . Indicators of corruption in the public sector are significant predictors of COVID-19 vaccination in August 2021 when controlling for GDP per capita and strength of the health system 26 , without controlling for life expectancy or education. Perceived corruption is associated with decreased vaccination coverage globally 27 and it also affects trust in mainstream health policy, exacerbating vaccination hesitancy 28 . We accessed publicly available data on COVID vaccination rates and other country-level indicators of human, social, and economic capital from the datasets of Our World in Data (OWID) 29 , the metrics included in the 2020 Human Development Report (HDR) of the United Nations Development Programme 30 , the Corruption Perception Index computed by Transparency International 31 , and World Bank data on poverty rates 32 . We included in the study all countries and territories with a population larger than 1 million and available information for vaccination rates, according to OWID data, resulting in 146 units of analysis 1 . Our first dependent variable of interest was the rate of fully vaccinated people, per hundred, measured at two points in time: July 31, 2021 (or the closest day to July 31, 2021) and February 4, 2022 (or the closest day to February 4, 2022) . The second dependent variable, included for comparison purposes, is the rate of infants immunized against measles at one year of age, in 2019, as reported by the HDR. The descriptive statistics and sources for the predictors included in the analysis are presented in Table 1 . The control variable for partial correlations was the HDI, which aggregates three dimensions: 1) life expectancy at birth; 2) an education index composed of mean years of schooling and expected years of schooling; and 3) gross national income per capita (GNI) 30 . An exploration of bivariate correlations indicated a strong relationship between COVID-19 vaccination rates and the HDI (bivariate r = 0.826 in February 2022, p = 0.000). The relationship changed from an exponential to a linear shape during the vaccination campaign from July 2021 ( Fig. 1) to February 2022 (Fig. 2) . In mid-2021, there was a much more abrupt co-variation of vaccination success with HDI, compared with the later stage, when access to vaccines was more widespread and countries' own resources for large-scale collective action became more relevant. Therefore, an exponential regression model (R 2 = 66.7%) is better fitted for the observed data in July than a linear regression model (R 2 = 48.3%). For February 2022, a linear model is better suited to model the relationship between HDI and vaccination rate (R 2 = 68.0%) than an exponential model (R 2 = 62.5%). A logarithmic model is marginally less fitted (R 2 = 66%) than a linear one, anticipating a turn toward a logarithmic-shaped relationship as more countries on the HDI continuum evolve toward the plateau of high vaccination rates. A bivariate analysis of vaccination rates and multiple indicators of human, social, and economic capital indicates a broad pattern of covariation (Table 2) . Vaccination rates are higher, on average, in countries with better outcomes in health and education, higher inputs into the health system, lower inequality, lower poverty rates, lower perceived corruption, and higher trust rates. The indicators that stand out in this pattern through their relative predictive power ( The three components of the HDI have differential predictive power for COVID-19 vaccination success (Table 3) . A multiple regression model of the vaccination rate in February 2022 on the three dimensions of HDI (Model 1 includes the mean years of schooling, and Model 2 includes the expected years of schooling) indicates that, when controlling for the other dimensions, the strongest predictor remains life expectancy. The model, including all three HDI dimensions, does not lead to a substantial increase in predictive power. This is due to the fact that life expectancy, GNI per capita, and the mean and expected years of schooling are strongly intercorrelated and the latter do not contribute much in terms of additional explanatory power. The same holds if we include expected years of schooling instead. While a wide variety of indicators of human, social, and economic capital are correlated with vaccination success, both in July 2021 and February 2022, their predictive relevance is, most often, overlapping with the HDI. As we can see in Table 2 , partial correlations when controlling for the HDI are, as a rule, statistically insignificant. Two indicators of social capital contribute to predicting vaccination success beyond the HDI: the share of people who trust doctors and nurses and the share of people who trust their national government. Trust seems to play a significant role in the country-level success of the COVID-19 vaccination campaign. Indicators of health system effectiveness retain statistically significant partial correlations with the vaccination rate in February 2022 when controlling for the HDI. Cardiovascular (CVD) death rate has a partial correlation of -0.300 (p = 0.000), and the proportion of infants immunized for measles before one year of age has a partial correlation of 0.231 (p = 0.006). CVD are the leading cause of death globally. While their prevalence is higher in more developed countries, the associated mortality is higher in less developed countries. This makes this indicator a powerful proxy to capture the effectiveness of a country's medical system and overall social organization in increasing lifespan. The proportion of infants immunized for measles is a more specific indicator, pointing to a country's performance in its vaccination infrastructure. The prevalence of diabetes is not correlated with the COVID vaccination rate when controlling for the HDI, despite diabetes being a risk factor for severe COVID infections, which was associated with priority in the early vaccination campaigns. COVID-19 vaccination. The strongest bivariate predictors are life expectancy and the HDI. When controlling for the HDI, trust in the national government and trust in doctors and nurses remain statistically significant, but others indicators do not -except national poverty rates, which are relevant for measles but not for COVID-19 vaccination. Conversely, the CVD death rate remains significant for COVID-19 vaccination when controlling for the HDI, but not for measles. In Table 4 vaccination. As discussed before, a similar understanding holds for measles vaccination (Table 4 , Model 5). The HDI is also the strongest predictor of the rate of infants immunized for measles. The lower beta coefficient also reflects the nonlinear relationship, which is better approximated by a logarithmic curve, because of the vaccination plateau (Fig. 3) . Therefore, the predictive relevance of the HDI goes beyond COVID-19 vaccination, covering previous, better institutionalized vaccines as well. The rate of trust in doctors and nurses is also a significant predictor of measles vaccination. The CVD rate does not add a statistically significant predictive power for measles vaccination. Neither does the national poverty rate, despite having a significant partial correlation when controlling for the HDI. The relationship between COVID-19 vaccination rates and trust in doctors and nurses, while controlling for HDI and other country-level health outcomes, is useful to clarify divergences that rank prominently in public debate. COVID-19 vaccination trajectories among countries in the same HDI categories have been quite different. While the United States and Israel were initially champions due to securing early access, by July 2021 they had started losing ground compared with other high HDI countries, which benefit from very high levels of trust in their medical systems 33 , 34 (Fig. 4) . Another contrasting story concerns Romania and Spain. Romania has had a temporary advantage in the early days of the vaccination campaign, which was lost in June, when Spain started gaining ground, reaching one of the top global vaccination rates in February 2022. Since both are members of the European Union, this difference reflects vaccine hesitancy more than vaccine inequity. Trust in the healthcare system has been invoked to account for differences in vaccine hesitancy. Spain is credited with a high level of trust in vaccines and in its medical system 34 , 33 . On the other hand, less than 40% of the Romanian public trusts public hospitals 35 , following a decade-long struggle with corruption 36 . India, which has a very high rate of trust in doctors and nurses 33 , has reached higher vaccination rates than Romania, despite its considerable challenges in vaccine availability. The low levels of public trust in the medical system, associated with a longstanding crisis 37 , 34 , also seem to account, in part, for Bulgaria's low rate of vaccination despite the high availability of vaccines typical of EU countries (Fig. 5) . This supports the argument that vaccination success is less a matter of overcoming deficits in scientific literacy, and more a matter of establishing public trust in a health system and science with proven anterior performance in keeping people healthy and alive 10 . Our analysis also highlights the role of trust in doctors and nurses as a predictor of vaccination success. It remains statistically significant when controlling for the HDI and other generic and specific indicators of health system effectiveness (CVD mortality and measles vaccination coverage, respectively). Trust is statistically significant in partial correlation and multiple regression models of both COVID-19 and measles vaccination, while other indicators concerning economic inequality, perceived corruption, and inputs into the health system do not add predictive value beyond the HDI. National poverty rates seem to remain a relevant predictor for both types of vaccination, though statistical significance is oscillating around the 5% threshold, depending on model specification. COVID-19 vaccines prove to be part of the Matthew effect of accumulating advantages and exacerbating disadvantages that the pandemic inflicted on societies and communities across the world 38 . At the same time, the remaining 28% of variability that cannot be determined through these sets of metrics points to a considerable scope for collective and individual agency in a time of crisis. For example, countries with an HDI of approximatively 85 ranged from rates of 40% to 80% for fully vaccinated people. The mobilization and coordination in the vaccination campaigns of citizens, medical professionals, scientists, journalists, and politicians, among others, account for at least some of this variability in overcoming vaccine hesitancy and inequity. All authors made a significant contribution to the development of this manuscript and approved the final version for submission. This work was supported by the Ministry of Innovation, Scientific Research, and Digitalization, Romania, project PN-III-P4-ID-PCE-2020-1589. 20 Questions: Dave Barry Vaccine impact: Benefits for human health Quantifying impacts of the COVID-19 pandemic through life expectancy losses: A population-level study of 29 countries Effects of covid-19 pandemic on life expectancy and premature mortality in 2020: time series analysis in 37 countries The real-life impact of vaccination on COVID-19 mortality in Europe and Israel. Public Health States and vaccines in the age of Covid-19 Vaccine hesitancy: Definition, scope and determinants Understanding Vaccine Hesitancy as Extended Attitudes Social amplification of risk and ''probable vaccine damage": A typology of vaccination beliefs in 28 European countries The Social Bifurcation of Reality: Symmetrical Construction of Knowledge in Lay Science-Distrusting and Science-Trusting Discourses Vaccine hesitancy in the era of COVID-19. Public Health Determinants of access to the SARS-CoV-2 vaccine: a preliminary approach Assessing Inequities in COVID-19 Vaccine Roll-Out Strategy Programs: A Cross-Country Study Using a SARS-COV-2 Vaccines: What Indicators are Associated with the Worldwide Distribution of the First Doses How education and GDP drive the COVID-19 vaccination campaign Forecasted trends in vaccination coverage and correlations with socioeconomic factors: a global time-series analysis over 30 years How confidence in health care systems affects mobility and compliance during the COVID-19 pandemic Repertoires of Vaccine Refusal in Romania Vaccinating against COVID-19: The Correlation between Pro-Vaccination Attitudes and the Belief That Our Peers Want to Get Vaccinated No need for the needle. A qualitative analysis of the antivax movement in Romania The vaccination debate in the "post-truth" era: social media as sites of multi-layered reflexivity. Sociol Health Illn Correlates of intended COVID-19 vaccine acceptance across time and countries: results from a series of cross-sectional surveys What Accounts for the Variation in COVID-19 Vaccine Hesitancy in Eastern, Southern and Western Europe? Communicating with patients and the public about COVID-19 vaccine safety: recommendations from the Collaboration on Social Science and Immunisation Social and economic inequality in coronavirus disease 2019 vaccination coverage across Illinois counties Table 4 Regression results with the % of the population who have been fully vaccinated against COVID-19 as of 30.08.2021 and corruption as the main explanatory variable Corruption costs lives: evidence from a cross-country study Coronavirus Pandemic (COVID-19) [Internet]. Our World in Data Transparency International. Corruption Perceptions Index The World Bank Data Percentage of persons with trust in healthcare in 2021, by country Bulgaria's doctors have scant future Romania's health-care system, the EU's worst, struggles to reform [Internet]. The Economist Less Than 5% of Bulgarians Trust Healthcare System Coronavirus and Inequality: The Rich Get Richer, the Poor Get a Kick in the Behind 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.