Analysis of the COVID-19 pandemic: lessons towards a more effective response to public health emergencies – Globalization and Health – Globalization…

Posted: February 5, 2022 at 5:02 am

Heterogeneity of COVID-19 cases and deaths around the world: what can explain it?

Table1 indicates that the pandemic of COVID-19 is heterogeneous around regions of the world. Figure1 also shows that there is a strong and significant correlation between HDI and globalisation (with an increase in trade and tourism as proxy indicators) and a corresponding strong and significant correlation with COVID-19 burden.

Human development index and its correlates associated with COVID-19 in 189 countries*

Globalisation and pandemics interact in various ways, including through international trade and mobility, which can lead to multiple waves of infections [11]. In at least the first waves of the pandemic, countries with high import and export of consumer goods, food products and tourism have high number of cases, severe cases, deaths and CFRs. Countries with high HDI are at a higher risk of importing (and exporting) COVID-19 due to high mobility linked to trade and tourism, which are drivers of the economy. These may have led to multiple introductions of COVID-19 into these countries before border closures.

The COVID-19 pandemic was first identified in China, which is central to the global network of trade, from where it spread to all parts of the world, especially those countries with strong links with China [12]. The epidemic then spread to Europe. There is very strong regional dimension to manufacturing and trading, which could be facilitate the spread of the virus. China is the heart of Factory Asia; Italy is in the heart of Factory Europe; the United States is the heart of Factory North America; and Brazil is the heart of Factory Latin America [13]. These are the countries most affected by COVID-19 during the first wave of the pandemic [2, 3, 14].

It is also important to note that two-third of the countries currently reporting more than a million cases are middle-income countries (MICs), which are not only major emerging market economies but also regional political powers, including the BRICS countries (Brazil, Russia, India and South Africa) [3, 15]. These countries participate in the global economy, with business travellers and tourists. They also have good domestic transportation networks that facilitate the internal spread of the virus. The strategies that helped these countries to become emerging markets also put them at greater risk for importing and spreading COVID-19 due to their connectivity to the rest of the world.

In addition, countries with high HDI may be more significantly impacted by COVID-19 due to the higher proportion of the elderly and higher rates of non-communicable diseases. Figure 1 shows that there is a strong and significant correlation between HDI and demographic transition (high proportion of old-age population) and epidemiologic transition (high proportion of the population with non-communicable diseases). Countries with a higher proportion of people older than 65years and NCDs (compared to communicable diseases) have higher burden of COVID-19 [16,17,18,19,20]. Evidence has consistently shown a higher risk of severe COVID-19 in older individuals and those with underlying health conditions [21,22,23,24,25]. CFR is age-dependent; it is highest in persons aged 85years (10 to 27%), followed by those among persons aged 6584years (3 to 11%), and those among persons aged 55-64years (1 to 3%) [26].

On the other hand, regions and countries with low HDI have, to date, experienced less severe epidemics. For instance, as of January 12, 2022, the African region has recorded about 10.3 million cases and 233,000 deaths far lower than other regions of the world (Table 1) [27]. These might be due to lower testing rates in Africa, where only 6.5% of the population has been tested for the virus [14, 28], and a greater proportion of infections may remain asymptomatic [29]. Indeed, the results from sero-surveys in Africa show that more than 80% of people infected with the virus were asymptomatic compared to an estimated 40-50% asymptomatic infections in HICs [30, 31]. Moreover, there is a weak vital registration system in the region indicating that reports might be underestimating and underreporting the disease burden [32]. However, does this fully explain the differences observed between Africa and Europe or the Americas?

Other possible factors that may explain the lower rates of cases and deaths in Africa include: (1) Africa is less internationally connected than other regions; (2) the imposition of early strict lockdowns in many African countries, at a time when case numbers were relatively small, limited the number of imported cases further [2, 33, 34]; (3) relatively poor road network has also limited the transmission of the virus to and in rural areas [35]; (4) a significant proportion of the population resides in rural areas while those in urban areas spend a lot of their time mostly outdoors; (5) only about 3% of Africans are over the age of 65 (so only a small proportion are at risk of severe COVID-19) [36]; (6) lower prevalence of NCDs, as disease burden in Africa comes from infectious causes, including coronaviruses, which may also have cross-immunity that may reduce the risk of developing symptomatic cases [37]; and (7) relative high temperature (a major source of vitamin D which influences COVID-19 infection and mortality) in the region may limit the spread of the virus [38, 39]. We argue that a combination of all these factors might explain the lower COVID-19 burden in Africa.

The early and timely efforts by African leaders should not be underestimated. The African Union, African CDC, and WHO convened an emergency meeting of all African ministers of health to establish an African taskforce to develop and implement a coordinated continent-wide strategy focusing on: laboratory; surveillance; infection prevention and control; clinical treatment of people with severe COVID-19; risk communication; and supply chain management [40]. In April 2021, African Union and Africa CDC launched the Partnerships for African Vaccine Manufacturing (PAVM), framework to expanding Africas vaccine manufacturing capacity for health security [41].

Figures 2 and 3 illustrate the variability of cases and deaths due to the COVID-19 pandemic across high-income countries (HICs). Contrary to the overall positive correlation between high HDI and cases, deaths and fatality rates due to COVID-19, there are outlier HICs, which have been able to control the epidemic. Several HICs, such as New Zealand, Australia, South Korea, Japan, Denmark, Iceland, and Norway, managed to contain their epidemics (Figs. 2 and 3) [15, 42, 43]. It is important to note that most of these countries (especially the island states) have far less cross-border mobility than other HICs.

Scatter plot of COVID-19 cases per million population in countries with high human development index (>0.70)

Scatter plot of COVID-19 deaths per million population in countries with high human development index (>0.70)

HICs that have been successful at controlling their epidemics have similar characteristics, which are related to governance of the response [44], synergy between UHC and GHS, and existing relative socio-economic equity in the country. Governance and leadership is a crucial factor to explain the heterogeneity of the epidemic among countries with high HDI [45]. There has been substantial variation in the nature and timing of the public health responses implemented [46]. Adaptable and agile governments seem better able to respond to their epidemics [47, 48]. Countries that have fared the best are the ones with good governance and public support [49]. Countries with an absence of coherent leadership and social trust have worse outcomes than countries with collective action, whether in a democracy or autocracy, and rapid mobilisation of resources [50]. The erosion of trust in the United States government has hurt the countrys ability to respond to the COVID-19 crisis [51, 52]. The editors of the New England Journal of Medicine argued that the COVID-19 crisis has produced a test of leadership; but, the leaders in the United States had failed that test [47].

COVID-19 has exposed the fragility of health systems, not only in the public health and primary care, but also in acute and long-term care systems [49]. Fragmentation of health systems, defined here to mean inadequate synergy and/ or integration between GHS and UHC, is typical of countries most affected by the COVID-19 pandemic. Even though GHS and UHC agendas are convergent and interdependent, they tend to have different policies and practices [53]. The United States has the highest index for GHS preparedness; however, it has reported the worlds highest number of COVID-19 cases and deaths due to its greatly fragmented health system [54, 55]. Countries with health systems and policies that are able to integrate International Health Regulations (IHR) core capacities with primary health care (PHC) services have been effective at mitigating the effects of COVID-19 [50, 53]. Australia has been able to control its COVID-19 epidemic through a comprehensive primary care response, including protection of vulnerable people, provision of treatment and support services to affected people, continuity of regular healthcare services, protection and support of PHC workers and primary care services, and provision of mental health services to the community and the primary healthcare workforce [56]. Strict implementation of public health and social intervention together with UHC systems have ensured swift control of the epidemics in Singapore, South Korea, and Thailand [57].

The heterogeneity of cases and deaths, due to COVID-19, is also explained by differences in levels of socio-economic inequalities, which increase susceptibility to acquiring the infection and disease progression as well as worsening of health outcomes [58]. COVID-19 has been a stress test for public services and social protection systems. There is a higher burden of COVID-19 in Black, Asian and Minority Ethnic individuals due to socio-economic inequities in HICs [59, 60]. Poor people are more likely to live in overcrowded accommodation, are more likely to have unstable work conditions and incomes, have comorbidities associated with poverty and precarious living conditions, and reduced access to health care [59].

The epidemiology of COVID-19 is also variable across MICs, with HDI between 0.70 and 0.85, around the world. Overall, the epidemic in MICs is exacerbated by the rapid demographic and epidemiologic transitions as well as high prevalence of obesity. While India and Brazil witnessed rapidly increasing rates of cases and deaths, China, Thailand, Vietnam have experienced a relatively lower disease burden [15]. This heterogeneity may be attributed to a number of factors, including governance, communication and service delivery. Thailand, China and Vietnam have implemented a national harmonized strategic response with decentralized implementation through provincial and district authorities [61]. Thailand increased its testing capacity from two to over 200 certified facilities that could process between 10,000 to 100,000 tests per day; moreover, over a million village health volunteers in Thailand supported primary health services [62, 63]. Chinas swift and decisive actions enabled the country to contain its epidemic though there was an initial delay in detecting the disease. China has been able to contain its epidemic through community-based measures, very high public cooperation and social mobilization, strategic lockdown and isolation, multi-sector action [64]. Overall, multi-level governance (effective and decisive leadership and accountability) of the response, together with coordination of public health and socio-economic services, and high levels of citizen adherence to personal protection, have enabled these countries to successfully contain their epidemics [61, 65, 66].

On the other hand, the Brazilian leadership was denounced for its failure to establish a national surveillance network early in the pandemic. In March 2020, the health minister was reported to have stated that mass testing was a waste of public funding, and to have advised against it [67]. This was considered as a sign of a collapse of public health leadership, characterized by ignorance, neoliberal authoritarianism [68]. There were also gaps in the public health capacity in different municipalities, which varied greatly, with a considerable number of Brazilian regions receiving less funding from the federal government due to political tension [69]. The epidemic has a disproportionate adverse burden on states and municipalities with high socio-economic vulnerability, exacerbated by the deep social and economic inequalities in Brazil [70].

India is another middle-income country with a high burden of COVID-19. It was one of the countries to institute strict measures in the early phase of the pandemic [71, 72]. However, the government eased restrictions after the claim that India had beaten the pandemic, which lead to a rapid increase in disease incidence. Indeed, on 12 January 2022, India reported 36 million cumulative cases and almost 485,000 total deaths [15]. The second wave of the epidemic in India exposed weaknesses in governance and inadequacies in the countrys health and other social systems [73]. The nature of the Indian federation, which is highly centripetal, has prevented state and local governments from tailoring a policy response to suit local needs. A centralized one-size-fits-all strategy has been imposed despite high variations in resources, health systems capacity, and COVID-19 epidemics across states [74]. There were also loose social distancing and mask wearing, mass political rallies and religious events [75]. Rapid community transmission driven by high population density and multigenerational households has been a feature of the current wave in India [76]. In addition, several new variants of the virus, including the UK (B.1.1.7), the South Africa (20H/501Y or B.1.351), and Brazil (P.1), alongside a newly identified Indian variant (B.1.617), are circulating in India and have been implicated as factors in the second wave of the pandemic [75, 76].

The pandemic is characterized by variable CFRs across regions and countries that are negatively associated with HDI (Fig.1). The results presented in Fig.4 show that the proportion of elderly population and rate of obesity are important factors which are positively associated with CFR. On the other hand, UHC, IHR capacity and other indicators of health systems capacity (health workforce density and hospital beds) are negatively associated with the CFR (Figs. 1 and 4).

Correlates of COVID-19 cases, deaths and case-fatality rates in 189 countries

The evidence from several research indicates that heterogeneity can be explained by several factors, including differences in age-pyramid, socio-economic status, access to health services, or rates of undiagnosed infections. Differences in age-pyramid may explain some of the observed variation in epidemic severity and CFR between countries [77]. CFRs across countries look similar when taking age into account [78]. The elderly and other vulnerable populations in Africa and Asia are at a similar risk as populations in Europe and Americas [79]. Data from European countries suggest that as high as 57% of all deaths have happened in care homes and many deaths in the US have also occurred in nursing homes. On the other hand, in countries such as Mexico and India, individuals <65years contributed the majority of deaths [80].

Nevertheless, CFR also depends on the quality of hospital care, which can be used to judge the health system capacity, including the availability of healthcare workers, resources, and facilities, which affects outcomes [81]. The CFR can increase if there is a surge of infected patients, which adds to the strain on the health system [82]. COVID-19 fatality rates are affected by numerous health systems factors, including bed capacity, existence and capacity of intensive care unit (ICU), and critical care resources (such as oxygen and dexamethasone) in a hospital. Regions and countries with high HDI have a greater number of acute care facilities, ICU, and hospital bed capacities compared to lower HDI regions and countries [83]. Differences in health systems capacity could explain why North America and Europe, which have experienced much greater number of cases and deaths per million population, reported lower CFRs than the Southern American and the African regions, partly also due to limited testing capacity in these regions (Table 1) [84,85,86]. The higher CFR in Southern America can be explained by the relatively lower health systems surge capacity that could not adequately respond to the huge demand for health services [69, 86]. The COVID-19 pandemic has highlighted existing health systems weaknesses, which are not able to effectively prepare for and respond to PHEs [87]. The high CFRs in the region are also exacerbated by the high social inequalities [69].

On the other hand, countries in Asia recorded lower CFRs (~1.4%) despite sharing many common risk factors (including overcrowding and poverty, weak health system capacity etc) with Africa. The Asian region shares many similar protective factors to the African region. They have been able to minimize their CFR by suppressing the transmission of the virus and flattening the epidemic curve of COVID-19 cases and deaths. Nevertheless, the epidemic in India is likely to be different because it has exceeded the health system capacity to respond and provide basic medical care and medical supplies such as oxygen [88]. Overall, many Asian countries were able to withstand the transmission of the virus and its effect due to swift action by governments in the early days of the pandemic despite the frequency of travel between China and neighbouring countries such as Hong Kong, Taiwan and Singapore [89]. This has helped them to contain the pandemic to ensure case numbers remain within their health systems capacity. These countries have benefited from their experience in the past in the prevention and control of epidemics [90].

There are a number of issues with the use of the CFR to compare the management of the pandemic between countries and regions [91], as it does not depict the true picture of the mortality burden of the pandemic. A major challenge with accurate calculation of the CFR is the denominator on number of identified cases, as asymptomatic infections and patients with mild symptoms are frequently left untested, and therefore omitted from CFR calculations. Testing might not be widely available, and proactive contact tracing and containment might not be employed, resulting in a smaller denominator, and skewing to a higher CFR [82]. It is, therefore, far more relevant to estimate infection fatality rate (IFR), the proportion of all infected individuals who have died due to the infection [91], which is central to understanding the public health impact of the pandemic and the required policies for its prevention and control [92].

Estimates of prevalence based on sero-surveys, which includes asymptomatic and mildly symptomatic infections, can be used to estimate IFR [93]. In a systematic review of 17 studies, seroprevalence rates ranged from 0.22% in Brazil to 53% in Argentina [94]. The review also identified that the seroprevalence estimate was higher than the cumulative reported case incidence, by a factor between 1.5 times in Germany to 717 times in Iran, in all but two studies (0.56 times in Brazil and 0.88 times in Denmark) [94, 95]. The difference between seroprevalence and cumulative reported cases might be due to asymptomatic cases, atypical or pauci-symptomatic cases, or the lack of access to and uptake of testing [94]. There is only a modest gap between the estimated number of infections from seroprevalence surveys and the cumulative reported cases in regions with relatively thorough symptom-based testing. Much of the gap between reported cases and seroprevalence is likely to be due to undiagnosed symptomatic or asymptomatic infections [94].

It is important to note that the pandemic has significant collateral effects on the provision of essential health services, in addition to the direct health effects [96]. Disruptions in the provision of essential health services, due to COVID-19, were reported by nearly all countries, though it is more so in lower-income than higher-income countries [97, 98]. The biggest impact reported is on provision of day-to-day primary care to prevent and manage some of the most common health problems [99].

The causes of disruptions in service delivery were a mix of demand and supply factors [100]. Countries reported that just over one-third of services were disrupted due to health workforce-related reasons (the most common causes of service disruptions), supply chains, community mistrust and fears of becoming infected, and financial challenge s[101]. Cognizant of the disruptive effects of the pandemic, countries have reorganized their health system.

Countries with better response to COVID-19 have mobilized, trained and reallocated their health workforce in addition to hiring new staff, using volunteers and medical trainees and mobilizing retirees [102]. Several strategies have also been implemented to mitigate disruptions in service delivery and utilization, including: triaging to identify the most urgent patient needs, and postponing elective medical procedures; switching to alternative models of care, such as providing more home-based care and telemedicine [101].

Follow this link:

Analysis of the COVID-19 pandemic: lessons towards a more effective response to public health emergencies - Globalization and Health - Globalization...

Related Posts