key: cord-1047707-igohmd37
authors: Wyper, G. M.; Assuncao, R. M.; Cuschieri, S.; Devleeschauwer, B.; Fletcher, E.; Haagsma, J. A.; Hilderink, H.; Idavain, J.; Lesnik, T.; Von der Lippe, E.; Majdan, M.; Milicevic, M. S.; Pallari, E.; Penalvo, J. L.; Pires, S. M.; Plass, D.; Santos, J. V.; Thomsen, S. T.; Stockton, D. L.; Grant, I.
title: Population vulnerability to COVID-19 in Europe: a burden of disease analysis
date: 2020-05-04
journal: nan
DOI: 10.1101/2020.04.29.20064279
sha: d02aa105faf7db52d00cf2b9eff6f12b3c13da0e
doc_id: 1047707
cord_uid: igohmd37

Background: Evidence has emerged showing that elderly people and those with pre-existing chronic health conditions may be at higher risk of developing severe health consequences from COVID-19. In Europe, this is of particular relevance with ageing populations living with non-communicable diseases, multi-morbidity and frailty. Published estimates of Years Lived with Disability (YLD) from the Global Burden of Disease (GBD) study help to characterise the extent of these effects. Our aim was to identify the countries across Europe that have populations at highest risk from COVID-19 by using estimates of population age structure and YLD for health conditions linked to severe illness from COVID-19. Methods: Population and YLD estimates from GBD 2017 were extracted for 45 countries in Europe. YLD was restricted to a list of specific health conditions associated with being at risk of developing severe consequences from COVID-19 based on guidance from the United Kingdom Government. This guidance also identified individuals aged 70 years and above as being at higher risk of developing severe health consequences. Study outcomes were defined as: (i) proportion of population aged 70 years and above; and (ii) rate of YLD for COVID-19 for vulnerable health conditions across all ages. Bivariate groupings were established for each outcome and combined to establish overall population-level vulnerability. Results: Countries with the highest proportions of elderly residents were Italy, Greece, Germany, Portugal and Finland. When assessments of population-level YLD rates for COVID-19 vulnerable health conditions were made the highest rates were observed for Bulgaria, Czech Republic, Croatia, Hungary and Bosnia and Herzegovina. A bivariate analysis indicated that the countries at high-risk across both measures of vulnerability were: Bulgaria; Portugal; Latvia; Lithuania; Greece; Germany; Estonia; and Sweden. Conclusion: Routine estimates of population structures and non-fatal burden of disease measures can be usefully combined to create composite indicators of vulnerability for rapid assessments, in this case to severe health consequences from COVID-19. Countries with available results for sub-national regions within their country, or national burden of disease studies that also use sub-national levels for burden quantifications, should consider using non-fatal burden of disease estimates to estimate geographical vulnerability to COVID-19.

In burden of disease studies estimates of disability-adjusted life years (DALYs) are commonly used to assess the leading causes of burden amongst populations [1] . DALYs are composed of estimates of population health loss due to living with the consequences of morbidity and premature mortality. Years Lived with Disability (YLD) capture the morbidity (both the prevalence and severity of the disease) component of DALYs by estimating the number of years lost due to conditions diminishing the overall health status, and are a useful indicator to assess how impaired populations are due to living with the consequences of disease and injury [2] .

Internationally, countries have reacted to the COVID-19 outbreak by introducing key public health non-pharmaceutical interventions (otherwise known as physical, or social, distancing) to protect vulnerable population groups [3] . Evidence has emerged to show that elderly people and those with pre-existing multi-morbid conditions may be at higher risk of developing severe health consequences from COVID-19 [4] . In Europe, 31% of the population are estimated to have a condition that is on the Government of the United Kingdom's (UK) list of conditions at increased risk of severe health consequences from COVID-19 disease [5] . There is currently a disparity of comparable information across countries to objectively assess country-level vulnerability to COVID-19. However, there is a wealth of data on population structure, health status and causes of health loss in countries, which can be obtained from the Global Burden of Disease (GBD) study [6] . These data can be used to approximate how vulnerable populations are, particularly by focusing on the population share of elderly residents and the YLD for health conditions that have been identified as potentially linked to severe illness from COVID-19. This is of particular relevance for European countries, as increases in lifespan have resulted in increasingly ageing populations living with effects of non-communicable diseases, multi-morbidity and frailty [7] .

The aim of this study was to identify the countries across Europe that have populations at highest risk for severe disease progression after COVID-19 infection by using estimates of population structure and YLD for health conditions linked to severe illness from COVID-19. This study was carried out using data from GBD 2017 for the reference year 2017, considering two measures of vulnerability: (i) rate of elderly population; and (ii) rate of YLD for health conditions identified at risk of severe health consequences from COVID-19.

The GBD Results Tool [8] was used to extract Years Lived with Disability (YLD) estimates for both sexes, age-groups (all ages; 70 years and above; and 80 years and above) and GBD 2017 level 3 cause [9] for each country defined as residing in Central, Eastern and Western Europe (N=45 countries). Estimates were considered for the constituent nations of the United Kingdom (UK): England; Northern Ireland; Scotland; and Wales, rather than the UK as a whole. In this study, hereafter, the elderly population denotes the age-group 70 years and above.

Data were retained for specific causes based on guidance from the UK Government (as at 30 th March 2020) on those health conditions that indicated a risk of severe health consequences from COVID-19 [10] . Two groups were defined: individuals aged 70 years and above, and those under 70 years that have one or more pre-existing underlying health condition. The guidance provided by the UK Government is outlined in the Supplementary Appendix and the list of pre-existing conditions were mapped to the GBD 2017 cause list (Table 1) .

Some emerging evidence has considered obesity and hypertension as independent risk factors for severe health consequences from COVID-19 [11, 12] . However, we do not consider them separately in this study as the vast majority of disease outcomes associated with these risk factors are included in the mapping to the GBD cause list (Table 1) . For hypertension, all disease outcomes linked to high systolic blood pressure risk factor are included (cardiovascular diseases and chronic kidney disease . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 4, 2020. . https://doi.org/10.1101/2020.04.29.20064279 doi: medRxiv preprint in Table 1 ). In addition, GBD include separate estimates for hypertensive disease and these are included within cardiovascular diseases. The disease outcomes associated with obesity are also all covered in the mapping to the GBD cause list with the exception of Gout. Gout accounted for only 0.2% (95% uncertainty interval: 0.15-0.25%) of total YLD in GBD European Region in 2017 [8]. A permalink to the GBD Results Tool [8] query that were used to generate the data used in this study are outlined in the Data Availability section. Additionally, data on the total 2017 resident populations and population aged 70 years and above for each country were sourced from the Global Health Data Exchange (GHDx) [13] . These population denominators that were used in the production of GBD 2017 estimates.

Descriptive summaries were calculated for the proportion of elderly population, and YLD for COVID-19 vulnerable health conditions were described using crude rates per 100,000 population. The numerators for the population proportion calculations were based on elderly populations, whereas the YLD rate calculation numerators were based on population totals. Denominators were based on the all ages population data sourced from GHDx [13] .

Each measure was divided into tertiles (three binned categories: low; mid; and high). These categories were calculated to determine three equal size groups of vulnerability. Bivariate groupings were established by considering overlapping of the measures and were depicted in a scatter plot to . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 4, 2020. . https://doi.org/10.1101/2020.04.29.20064279 doi: medRxiv preprint identify groups of countries, both in terms of the proportion of elderly population and the rate of YLD for conditions associated with worse COVID-19 prognosis. Spearman's rank correlation coefficient (ρ) was used to describe the correlation between the percentage of elderly population and the rate of YLD for COVID-19 vulnerable health conditions.

The five countries with the highest proportions of elderly residents (aged 70 years and above) were: Italy (16.4%); Greece (16.2%); Germany (15.4%); Portugal (15.3%); and Finland (14.7%) ( Table 2) . Conversely, the countries with the lowest proportions of elderly population were Israel (7.7%); Moldova (8.2%); Russian Federation (8.8%); Macedonia (8.8%); and Albania (9.0%). The ratio of the country with the highest (Italy) and lowest (Israel) proportion of elderly residents was 2.14, indicating over a two-fold difference between the countries.

When looking at stratified differences within the elderly age-group, the five countries with the highest percentage of population aged 80 years and above were: France (44.5%); Spain (44.4%); Greece (43.2%); Belgium (42.2%) and Italy (41.7%). The five countries with the lowest percentage of population aged 80 years and above were: Albania (30.8%); Bosnia and Herzegovina (31.4%); Czechia (31.5%); Macedonia (32.4%); and Montenegro (32.5%). Between the country with the highest (France) percentage of population aged 80 years and above and lowest (Albania), there was an absolute difference of 13.7%.

When the rate of YLD for health conditions associated with higher COVID-19 vulnerability was assessed for all ages, the five countries with the highest rates per 100,000 population were: Bulgaria (3,760); Czechia (3,690); Croatia (3,651), Hungary (3,613); and Bosnia and Herzegovina (3,553) ( Table 2 ). The five countries with the lowest rates were: Israel (1,870); Cyprus (2,272); Moldova (2,448); Albania (2,460); and France (2,505). There was a rate ratio of 2.01 between the country with the highest rate (Bulgaria) and the country with the lowest rate (Israel).

Insights into rates of YLD for health conditions indicating higher COVID-19 vulnerability for those under 70 years and elderly residents were that there were four countries that were common amongst the leading five countries in both age-groups. These countries were: Czechia, Croatia, Hungary and Bosnia and Herzegovina. Of the five countries with the lowest rates in the under 70 years and elderly age-groups, there were three countries that were common: Israel, France and Spain.

There was a moderate association (ρ=0.54) between the percentage of elderly population and the rate of YLD for COVID-19 vulnerable health conditions. A bivariate analysis indicated that the countries which had high proportions of elderly population and high rates of YLD for COVID-19 vulnerable health conditions were: Bulgaria; Portugal; Latvia; Lithuania; Greece; Germany; Estonia; and Sweden. Conversely, the countries with the lowest proportions of elderly population and lowest rates of YLD for COVID-19 vulnerable health conditions were: Israel; Cyprus; Moldova; Albania; Andorra; Ireland; Russian Federation; Iceland; and Belarus. Bosnia and Herzegovina had a high rate of YLD for COVID-19 vulnerable health conditions, but a relatively low proportion of elderly population. On the other hand, Spain, France and Austria all had high proportions of elderly population but a relatively low rate of YLD for COVID-19 vulnerable health conditions (Figure 1 ).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 4, 2020. . https://doi.org/10.1101/2020.04.29.20064279 doi: medRxiv preprint

All rates described are crude rates per 100,000 population. White vertical and horizontal gridlines indicate the tertile dividing lines for the measures: percentage of elderly population; and rate of YLD for COVID-19 vulnerable health conditions, respectively.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 4, 2020. 

This study set out to establish which countries across Europe had populations that were most likely to be vulnerable to severe health consequences as a result of COVID-19 infection. This assessment was made using data on population age structure, and data on YLD for health conditions identified as increasing the risk of COVID-19 severity, the latter illustrating the extent to which populations are vulnerable through living with ill-health due to causes of disease.

Estimates of vulnerability to COVID-19 using elderly population share indicated that the countries with the highest proportions of elderly residents were Italy, Greece, Germany, Portugal and Finland. When assessments of population-level YLD rates for COVID-19 vulnerable health conditions were made the highest rates were observed for Bulgaria, Czechia, Croatia, Hungary and Bosnia and Herzegovina. Our bivariate analysis indicated that the countries which had high rates across both measures of vulnerability were: Bulgaria; Portugal; Latvia; Lithuania; Greece; Germany; Estonia; and Sweden.

Whilst these findings indicate population-level vulnerability due to health loss suffered, they do not take into account other important factors such as country and sub-national responses to the COVID-19 outbreak through public health non-pharmaceutical interventions. Neither do they take into account factors such as: population density, the capacity or ease of access to health and social care services and the disruption to existing services due to the COVID-19 crisis, all of which will have a significant impact on the extent to which vulnerable populations are adequately protected from harm. This may partly explain why countries identified in this analysis with high and low vulnerability to severe health consequences from COVID-19 do not always correspond with those countries in Europe with the highest and lowest case fatality ratios due to COVID-19 [14] . For example, within the Baltic states Latvia and Estonia have high vulnerability as measured on both indicators. However, Latvia responded to the crisis quickly by closing their borders and implementing restrictive measures much faster than Estonia, and case fatality rates are higher in Estonia [14] [15] [16] . This example highlights that a number of additional factors could contribute to differences between vulnerability and extent of adverse consequences, including: care identification and under-reporting, the speed at which countries introduced restrictive measures, and restrictions on air travel. The use of summary health indicator such as YLD to identify severe health consequences from COVID-19 infections should be regarded as just one of the elements that need to be taken into account in a complete risk assessment of vulnerability.

The study was carried out using estimates from GBD 2017, which is a widely used and wellestablished mechanism that has methodological consistency when producing estimates for individual countries [6] . The use of GBD 2017 is advantageous as estimates are publically accessible, which allows for the rapid assessments of the impact in response to public health emergency scenarios, such as the COVID-19 outbreak. Our findings are comparable on a like-for-like basis across countries. However, data sources that are fed into the modelling process for country-level estimates can vary based on location, therefore there is a risk that some of the differences which we observe may be attributed to the use, or omission, of high quality data sources [17] . We have opted not to include estimates of uncertainty in our estimates. Uncertainty intervals in the GBD study can often be wide, representing large degrees of uncertainty, so users of these results must bear in mind that these findings relate to the best available point-estimate. To retain consistency with estimates of YLD from GBD 2017, data on population size and structure was obtained from GHDx [13] which may differ from nationally produced estimates.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 4, 2020. Previous research has suggested that the assumption of fixed severity distributions across countries may be unreasonable [18] . In our study of COVID-19 related vulnerable conditions, we did not include some of the leading causes of YLD, such as major depressive disorders and substance use disorders, which are thought to be the most likely to be affected by this assumption. Thus, our COVID-19 vulnerable conditions analysis may be less affected by this assumption [19] . Additionally, our study has assumed that the extent of vulnerability to COVID-19 can be determined by disability weights. For example, on average a greater weight would be given to those suffering from chronic obstructive pulmonary disease than to ischaemic heart disease [20] . This assumption may be problematic if the risk of COVID-19 associated with each health condition is not representative of relative differences in disability weight between causes. Also, particular combinations of disease may result in higher risks of consequences of COVID-19, while all combinations are in this approach assumed to have a similar effect.

We have used YLD as a proxy for the severity of the selected vulnerable health conditions as YLD includes a weighting of the severity of diseases stages i.e. a weighted prevalence. We have chosen to explore the aim of the YLD summary measure to combine all conditions, rather than examine the impact of individual causes. We acknowledge that using disease prevalence data from GHDx could add further insight into quantifying the disease specific implications of severe health consequences from COVID-19. However, since prevalence gives equal weighting to each condition, we did not consider prevalence as useful for summary analyses as YLD which allows a weighted sum of prevalence of different diseases. Further analysis has previously been carried out elsewhere to explore using prevalence to quantify the risk for severe health consequences from COVID-19 infection to enhance assessment of a health systems vulnerability to COVID-19 [5] .

Our findings have highlighted that routine data on population structure might be usefully extended by using estimates of YLD to consider how populations are impaired by living with the consequences of ill-health due to causes of disease and injury. Countries with available estimates for sub-national regions within their country, or national burden of disease studies that also estimate at sub-national levels should consider using non-fatal burden of disease estimates to estimate geographical vulnerability to COVID-19.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 4, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 4, 2020. . https://doi.org/10.1101/2020.04.29.20064279 doi: medRxiv preprint Declarations Ethics approval and consent to participate Not applicable.

Not applicable.

The Global Burden of Disease: A comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020

Contribution of Chronic Disease to the Burden of Disability

Covid-19 and community mitigation strategies in a pandemic

The relationship of COVID-19 severity with cardiovascular disease and its traditional risk factors: A systematic review and meta-analysis

How many are at increased risk of severe COVID-19 disease? Rapid global, regional and national estimates

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

Living too long: the current focus of medical research on increasing the quantity, rather than the quality, of life is damaging our health and harming the economy

GBD 2017 causes of disease and injuries

Guidance on social distancing for everyone in the UK. United Kingdom Government

Obesity in patients younger than 60 years is a risk factor for Covid-19 hospital admission

Are patients with hypertension and diabetes mellitus at increased risk for COVID-19 infection?

Institute for Health Metrics and Evaluation. Global Health Data Exchange (GHDx)

European Centre for Disease Prevention and Control. Data on the geographic distribution of COVID-19 cases worldwide: European Union

State Chancellery of the Republic of Latvia. Press releases

Republic of Estonia Government. The government declared an emergency situation in Estonia until 1 May

Changes in health in the countries of the UK and 150 English Local Authority areas 1990-2016: a systematic analysis for the Global Burden of Disease Study

The impact of worldwide, national and sub-national severity distributions in Burden of Disease studies: A case study of cancers in Scotland

Prioritising the development of severity distributions in burden of disease studies for countries in the European region

Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2016 (GBD 2016) Disability Weights. Seattle, United States: Institute for Health Metrics and Evaluation (IHME

We would like to acknowledge the GBD 2017 study investigators and collaborators, as without this study and their resulting estimates of population and YLD our study would not have been possible. Additionally, we would like to thank COST action CA18218 (European Burden of Disease Network) for bringing together researchers working across Europe on burden of disease activities, which has made this collaborative research possible.

The datasets used in this research study are all publically available. The permalink to data query used to obtain estimates of YLD is: http://ghdx.healthdata.org/gbd-resultstool?params=gbd-api-2017-permalink/376d9a9ad8401f49f104650fab0b9305

All other authors declare that they have no competing interests.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. If applicable, open access publications fees upon acceptance of this article in a peer-reviewed journal will be funded under the COST action CA18218 (European Burden of Disease Network).

GW and IG generated the initial idea for the study. GW carried out all analyses and visualisation of the results. GW drafted the manuscript with assistance from IG. IG coordinated and made edits to the manuscript based on responses from co-authors with assistance from GW. All other authors provided critical input into the interpretation of the results, revisions to the manuscript and approved the final draft.

Source: Public Health England. Guidance on social distancing for everyone in the UK. United Kingdom Government. https://www.gov.uk/government/publications/covid-19-guidance-onsocial-distancing-and-for-vulnerable-people/guidance-on-social-distancing-for-everyone-inthe-uk-and-protecting-older-people-and-vulnerable-adults. Accessed 10 Apr 2020.We are advising those who are at increased risk of severe illness from coronavirus (COVID-19) to be particularly stringent in following social distancing measures.This group includes those who are: aged 70 and above (regardless of medical conditions)  under 70 with an underlying health condition listed below (i.e. anyone instructed to get a flu jab as an adult each year on medical grounds):  chronic (long-term) respiratory diseases, such as asthma, chronic obstructive pulmonary disease (COPD), emphysema or bronchitis  chronic heart disease, such as heart failure  chronic kidney disease  chronic liver disease, such as hepatitis  chronic neurological conditions, such as Parkinson's disease, motor neurone disease, multiple sclerosis (MS), a learning disability or cerebral palsy  diabetes  problems with your spleenfor example, sickle cell disease or if you have had your spleen removed  a weakened immune system as the result of conditions such as HIV and AIDS, or medicines such as steroid tablets or chemotherapy  being seriously overweight (a body mass index (BMI) of 40 or above)  those who are pregnant Note: there are some clinical conditions which put people at even higher risk of severe illness from COVID-19. If you are in this category, next week the NHS in England will directly contact you with advice about the more stringent measures you should take in order to keep yourself and others safe. For now, you should rigorously follow the social distancing advice in full, outlined below.People falling into this group are those who may be at particular risk due to complex health problems such as: people who have received an organ transplant and remain on ongoing immunosuppression medication  people with cancer who are undergoing active chemotherapy or radiotherapy  people with cancers of the blood or bone marrow such as leukaemia who are at any stage of treatment  people with severe chest conditions such as cystic fibrosis or severe asthma (requiring hospital admissions or courses of steroid tablets)  people with severe diseases of body systems, such as severe kidney disease (dialysis)