Years of life lost by COVID-19 in Portugal and comparison with other European countries in 2020

In 2020, the COVID-19 pandemic spread around the world, causing serious effects on populations [1]. It is a serious threat due to its high transmissibility and overall lethality (1.7% in Portugal, 2.1% in Europe and 2.2% worldwide as of 27 December 2020, not considering under-ascertainment) [1, 2]. This is only comparable to the Spanish flu in the year 1918, for which worldwide lethality is estimated to have been above 2.5% [3].

COVID-19 affects elderly people more severely [4]. In Portugal, there have been relatively few deaths in patients under the age of 50, accounting for only 1.2% of the total number of recorded deaths. On 27 December 2020, around 67.7% of COVID-19 deaths were recorded among people aged 80+ years and 31.2% among those aged 50 to 79 years of age [2].

The Centers for Disease Control and Prevention refers to excess deaths (or excess mortality) as “the difference between the observed numbers of deaths in specific time periods and expected numbers of deaths in the same time periods” [5]. EUROMO counted excess mortality across 26 European countries in 2020, including that directly attributed to COVID-19 [6]. However, not all excess deaths were directly caused by COVID-19. Because healthcare systems had to adapt to the overwhelming demands of the pandemic, patients with other common and serious diseases may have not received the level and quality of care that they would have received under normal circumstances [7]. By late May 2020, England and Wales, Italy, and Spain had already reported that 28.8, 32.5 and 61.8% of excess mortality, respectively, was not due to COVID-19, often referring to it as non-COVID-19 excess mortality or collateral mortality [8]. In Portugal, the percentage of non-COVID-19 excess mortality has fluctuated over time, from 51% between March and mid-April, to 92% in July and finally to no excess registered in December 2020 [8‐10].

The number of deaths is an imperfect measure of mortality, as it does not provide insight into the age distribution of deaths or how risk levels vary by age. It is equally important to measure premature deaths, in terms of the years expected to live in relation to the person’s life expectancy (LE), and to account for the number of years living with a disability. These two dimensions of the burden of disease can be expressed in terms of potential years of life lost (YLL) due to premature death and to years lived with disability (YLD) [11]. The disability-adjusted life year (DALY) is a known measure of the population’s health that results from their sum [12]. This calculation takes into consideration the age at which a certain death or health condition occurred, which can be estimated for a certain disease or group of diseases, using LE as the reference for the years the person is expected to live. A higher YLL can be due to larger numbers of death, deaths in younger ages, or some combination of the two [13].

Cancer (CA), cardiovascular diseases (CVDs) and diseases of the respiratory system (DRSs) are the most common causes of death in Portugal and tend to kill those affected at younger ages [14]. In 2017, CVDs and CA were the major causes of natural deaths worldwide for those aged 15 to 49 years, accounting for 1.26 million deaths and 1.06 million deaths, respectively [15]. The DRSs were the third cause of death for people aged 50 to 69 years and people aged 70+ years, accounting for 1 million deaths and 3.96 million deaths registered, respectively [15, 16].

The estimation of the number of YLL due to COVID-19 and due to other causes adds to the calculation of excess mortality. Several authors [17, 18] have developed preliminary models for estimating YLL to COVID-19, namely in Italy, where they estimated that the number of YLL per patient due to COVID-19 was about 13 years per male patient and 11 years per female patient.

This study aims to:

This is an observational, cross-sectional study. Our analysis rests on deaths registered by the Eurostat database for all-cause mortality, from the Institut National d’Études Démographiques [19] (INED) and from the Northern Ireland Statistics and Research Agency [20] (NISRA). The calculation method for each objective is explained below:

In this paper, we estimated that Italy, the United Kingdom, and Spain had the greatest amount of YLL directly due to COVID-19 and that Spain, Portugal, and the Netherlands had excess YLL due to non-COVID-19 causes.

Few studies have tried to estimate the YLL to COVID-19 and to non-COVID-19 causes in 2020 [18, 21, 27]. This paper does not calculate the YLD and, consequently, the DALYs [12] for COVID-19 because no firm data exists (yet) for a deeper understanding of the long-term effects of thisdisease, even though some COVID-19 survivors may report histological, imagiological, and epidemiological evidence that suggests they may produce relevant burden of disease in futureyears [28‐30]. The most severe forms of acute disease seem to affect mainly older people, many of whom already have significant comorbidities and disabilities, who end up dying in a few weeks. The increased disability during those weeks adds little to what is learned by calculating YLL. Thus, as the DALYs were not calculated, any value was not discounted or weighted for YLL as referenced in other methods [11].

In a recent study [18], it was found that men had about 45% more YLL to COVID-19 than women. This is close to our findings, where we state that men had around 60% of YLL to COVID-19 in the selected European countries. A higher average age-at-death for women and a higher number of affected men with COVID-19 are factors that may account for these results [18].

In the United State, it was estimated that, by early July, there were already 1.2 million YLL to COVID-19 and that by October that number already exceeded 2.5 million [31], 76.16 YLL/10000 inhabitants. This rate is much higher than that observed in this study, where the highest rate was registered in the United Kingdom (62.00 YLL/10000 inhabitants).

The assumption that the mortality by age and cause registered between 2012 and 2016 would apply to 2020 if COVID-19 had not occurred is fair. CA, CVDs, and DRSs are chronic and degenerative issues that do not change over short periods of time, except in cases of major events that drastically change the environmental, social, and economic determinants of health, as happened at the time of the breakup of the Soviet Union [32]. However, it is certainly an area to be investigated in the future.

Due to its contagiousness and seriousness, COVID-19 exerts strong pressure on the human and technical resources of the entire health system [33]. A relevant part of the pre-hospital care facilities, consultations, diagnostic facilities, general inpatient beds, and intensive care beds are occupied or reserved for patients with COVID-19 [9]. As it is impossible to increase the supply of these resources in proportion to the growth in demand, patients with other pathologies, such as CA, CVDs, DRSs, end up having their consultations, exams, hospitalizations, and scheduled surgeries postponed, sometimes sine die. It is reasonable to assume that the delay or lack of care may have had an impact on the health and survival of these patients.

The capacity of each country’s healthcare system to manage COVID-19 and non-COVID-19 patients may partly explain the difference in excess YLL to COVID-19 and other causes. Also, a lower intensity flu season in the analysed countries may be in part responsible for the lower mortality found in some nations [34]. In Portugal, as a result of the COVID-19 pandemic, there was a radical change in the availability and use of health services [9], because a significant part of the capacity of the national healthcare system was reallocated for the treatment of patients with COVID-19. In a previous report [10], we estimated an overall excess of 11,736 deaths between March and December 2020 in Portugal. This number is above the estimated excess YLL (10115) because, although they are related, they represent different outcomes. YLL only accounts for deaths below the age of 80 and increases with deaths that occur at younger ages. In this case, this is in line with the fact that most excess deaths were mainly seen at advanced ages.

Of all countries, Portugal stands out as having the highest ratio of excess YLL to non-COVID-19/COVID-19 YLL. This ratio can be a proxy for the countries capacity in coping with increased pressure on the healthcare system regarding management of non-COVID-19 disease. Some characteristics, such as the age distribution of the population, lack adherence to the policy measures during different phases of the pandemic, and lack of preparedness on behalf of healthcare systems may have contributed to these results in Portugal.

The increase of incidence of COVID-19 may end up with YLL beyond the expected values, even beyond the direct burden of this disease, which parallel to control the pandemic. Although no correlation was found in our study (probably due to organizational differences between healthcare systems between countries), countries need to be aware that exc-YLL-non-Cvd can seriously increase if there are no efforts to care for the short- and long-term management of other diseases. This is probably related to the lack of clinical follow-up, screenings, postponing surgeries, diagnostics, and healthcare-seeking hesitancy.

These results of this paper compare well with other studies, because they used the conventional methods of calculating YLL. This study focused on only eight European countries, because it was difficult to get the number of COVID-19 deaths by age and gender for many other countries. Our estimates of YLL may have some degree of error because mortality data by age and cause is available only up to 2016 in many countries and is referred to as provisory for some countries in 2020. Also, estimates of YLL are conservative, because we used 80 years of age as the upper limit for the range class [21]. These results may be even more conservative due to an expected decrease in the proportion of deaths by other non-natural causes, such as traffic accidents, increasing the proportion of deaths by other natural diseases. In fact, in a brief analysis, we found that the proportions of deaths by non-natural causes in Portugal were about 1.7, 1.6, and 1.5% in the years 2017, 2018, and 2019, respectively, and 1.3% in 2020 [35].

Results of this study must be interpreted cautiously. At first glance, it may seem obvious that countries with a higher LE are more penalized in the calculations of YLL. However, some of the characteristics that cause countries to obtain a higher average LE are also factors that contribute to better control of deaths at younger ages, such as a possible better organization and responsiveness of the National Health Systems, socioeconomic factors, and/or better health literacy levels, which can assist in the adoption of more preventive and responsible behaviours by their populations in controlling the dissemination of the virus [36, 37]. YLL is a different measure and may be very different when comparing it to overall mortality. It is dependent on the LE for each country and only for people below that value.

Other limitations are that the calculations of YLL were not adjusted by age, as some countries did not have these distributions available. As the data obtained are grouped, some analyses were not possible to be carried out, such as the calculation of confidence intervals. Also, some under-ascertainment in certification of COVID-19 deaths across Europe is not ruled out [38].

This study needs to be replicated later, as since January 2021, the number of cases and deaths of COVID-19 have greatly increased all across Europe [39].

COVID-19 has had a major impact on mortality rates in Portugal, as in other European countries, both in terms of the number of deaths and number of YLL. The relative impact of COVID-19 on the number of deaths has been greater than that on the number of YLL, because COVID-19 affects older people disproportionately. COVID-19 accounts for the overwhelming number of excess deaths and excess YLL in Portugal, more so than in other the European countries considered in this study. All the countries considered experienced overall excess YLL during the first year of the pandemic. Spain, Portugal, and the Netherlands experienced significant excess YLL to non-COVID-19 causes. COVID-19 accounted for far more YLL than the excess YLL caused by CA, CVDs, and DRSs combined.

The relative impact of COVID-19 on the number of deaths has been greater than that on the number of YLL, because COVID-19 affects older people disproportionately. Estimating the number of YLL to COVID-19 is an important method to measure the burden of this disease per country. Not all countries that had more YLL to COVID-19 had a greater excess YLL to other causes. Differences between countries in the number of excess YLL to non-COVID-19 causes are related to the capacity of healthcare systems to adapt and respond to COVID-19 and non-COVID-19 diseases.