Cause-specific mortality in Korea during the first year of the COVID-19 pandemic

OBJECTIVES This study aimed to examine the trends in total mortality between 1998 and 2020 and to compare the changes in a wide range of detailed causes of death between 2020 (i.e., during the coronavirus disease 2019 [COVID-19] pandemic) and the previous year in Korea. METHODS We used registered population and mortality data for the years 1998–2020 obtained from Statistics Korea. The age-standardized all-cause mortality rate and the annual percent change between 1998 and 2020 were determined. The rate ratio and rate difference of the age-standardized mortality rate between 2019 and 2020 were calculated. RESULTS The age-standardized all-cause mortality rate in Korea has been on a downward trend since 1998, and the decline continued in 2020. In 2020, 950 people died from COVID-19, accounting for 0.3% of all deaths. Mortality decreased for most causes of death; however, the number of deaths attributed to sepsis and aspiration pneumonia increased between 2019 and 2020 for both men and women. Age-specific mortality rates decreased or remained stable between 2019 and 2020 for all age groups, except women aged 25–29. This increase was mainly attributed to a higher number of suicide deaths. CONCLUSIONS This study shed light on the issues of sepsis and aspiration pneumonia despite the successful response to COVID-19 in Korea in 2020. Cases of death from sepsis and aspiration pneumonia should be identified and monitored. In addition, it is necessary to develop a proactive policy to address suicide among young people, especially young women.


INTRODUCTION
began in February 2021 [7]. This successful response has led to a low number of confirmed cases and deaths, and several studies have found that there were no excess deaths in Korea during the COVID-19 pandemic in 2020 [8][9][10][11].
Despite the excellent outcome of Korea's quarantine measures, there have been reports of increased mortality due to specific causes or delayed treatment during the COVID-19 pandemic [12][13][14][15]. When mortality rates were analyzed by separating them by deaths in medical facilities and deaths outside of medical facilities, a significant increase in deaths outside of medical facilities was observed in 2020. This indicates potential changes in healthcare-seeking behaviors and the accessibility of emergency medical services during the epidemic [12]. A study conducted in Taiwan, where there were no excess deaths, reported fewer deaths from pneumonia and influenza in 2020 [16]. In contrast, mortality from pneumonia, cardiovascular disease, and diabetes increased between January and March 2020 in Wuhan, a location that reported excess deaths [17]. However, studies have had limitations in analyzing a small number of causes of death or using a broader classification of causes of death. This study aimed to examine the trends in total mortality between 1998 and 2020 and compare the changes in a wide range of detailed causes of death between 2020 during the COVID-19 pandemic and the previous year in Korea.

Data
We used registered population and mortality data for the years 1998-2020. Both data sets are publicly accessible by visiting a website administered by Statistics Korea. The number of deaths was obtained from death certificate data provided by the MicroData Integrated Service (https://mdis.kostat.go.kr/) of Statistics Korea [18]. Mortality data included gender, age, and cause of death. Causes of death were coded according to the International Classification of Diseases, 10th revision (ICD-10). The mid-year population data based on information from the resident registration system of the Korean government were obtained from the Korean Statistical Information Service (http://kosis.kr/) of Statistics Korea [19]. The registered population data provide the number of the population according to gender and one-year age groups. Supplementary Material 1 presents the population size and number of deaths during the study period.
We categorized the causes of death into 17 broad classes and 48 specific groups (Supplementary Material 2) based on the categorization of Statistics Korea. Furthermore, sub-specific causes (i.e., alcohol-specific disorders and poisonings, smoking-related causes, and avoidable causes including amenable and preventable causes) were categorized separately. We obtained the list of causes of death associated with smoking from a 2019 United States study [20]. The list of causes of death associated with alcohol-specific disorders and poisonings was obtained from Organization for Economic Cooperation and Development (OECD)/Eurostat lists [21]. OECD/Eurostat lists were also used to categorize the causes of death that were considered amenable and preventable (Supplementary Material 3).

Statistical analysis
The rate of age-standardized all-cause mortality (per 100,000 people) and the annual percent change between 1998 and 2020 were determined. Age-standardized mortality rates and their 95% confidence intervals (CIs) according to the cause of death were calculated using direct standardization. The 2015 Korean population was used as the standard population. The rate ratio (RR) and rate difference (RD) of the age-standardized mortality rate were calculated as relative and absolute measures for differences between 2019 and 2020, respectively. Age-standardized mortality rates, RR, RD, and their 95% CIs were calculated using the PROC STDRATE procedure in SAS based on a normal distribution without bootstrap. We presented the age-standardized mortality rates by specific causes of death using 2015-2019 combined data and compared the age-standardized mortality rate between 2015-2019 and 2020 in Supplementary Materrials 3-5. All analyses were conducted separately for men and women using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA)

Ethics statement
This study was approved by the Institutional Review Board of the Ewha Womans University College of Medicine, Seoul, Korea (SEUMC 2022-08-044).

RESULTS
In Korea, age-standardized all-cause mortality decreased from 1998-2020 ( Figure 1). This trend was sustained in the first year of the COVID-19 pandemic. The age-standardized mortality rate per 100,000 people was 475.6 (95% CI, 473.8 to 477.3) in 2019 and 467.4 (95% CI, 465.7 to 469.1) in 2020. The decrease in mortality was greater among men than among women (Supplementary Material 1). Tables 1-3 show the cause-specific mortality rates and their differences between 2019 and 2020. In 2020, there were 950 (483 men and 467 women) deaths due to COVID-19. COVID-19 deaths accounted for 0.3% of all-cause mortality in 2020. Mortality decreased for most causes of death; however, it increased for sepsis, pneumonitis due to solids and liquids, and senility between 2019 and 2020 (Figure 2A-C). Among them, except for senility, the mortality rates for sepsis and pneumonitis due to solids and liquids were higher than the average mortality rate in 2015-2019 (Supplementary . This pattern was observed for both men and women. The number of deaths attributed to sepsis was 1,183 (540 men and 643 women), and the number of deaths attributed to pneumonitis due to solids and liquids was 751 (381 men and 370 women). Among men, the mortality rate decreased the most for cancer, followed by respiratory disease and external death between 2019 and 2020. Respiratory disease, cardiovascular disease, and cancer showed the highest decrease in the mortality rate among women. Among the detailed causes, pneumonia decreased by 911 cases in 2020 compared to 2019, resulting in an RD of -3.4% (95% CI, -4.0 to -2.7), and chronic lower respiratory disease decreased by 508 people, resulting in an RD of -1.3% (95% CI, -1.6 to -1.0) (Supplementary Material 3). However, pneumonia showed a decreasing trend in 2019 and 2020 compared to 2018, and chronic lower res-     piratory disease also showed a decreasing pattern after 2002 (Figure 2D and E). Although the number of deaths attributed to smoking among those aged 35 and over increased, the age-standardized mortality rate decreased between 2019 and 2020. Mortality due to alcohol-specific disorders and poisonings increased for both men and women. Avoidable deaths, including those that were amenable and preventable, significantly decreased among men only.
Age-specific mortality rates were decreased or stable between 2019 and 2020 for all age groups, except for women aged 25-29 (Table 4 and Supplementary Material 6). The increase in mortality among women aged 25-29 was attributed to an increase in external causes of deaths; in particular, an increase in suicide deaths. The suicide rate among women aged 25-29 increased by 2.88 from 16.55 (per 100,000) in 2019 to 19.43 in 2020 (Figure 3 and Supplementary Material 7). Since 2017, there has been an increasing trend in the suicide rate for women aged 25-29 ( Figure 2F).

DISCUSSION
The age-standardized all-cause mortality in Korea has been on a downward trend since 1998 and this trend continued in 2020, the first year of the COVID-19 pandemic. In 2020, 950 people died from COVID-19, accounting for 0.3% of all deaths. Mortality was decreased for most causes of death; however, the number of deaths attributed to sepsis and pneumonitis due to solids and liquids increased between 2019 and 2020 for both men and women. Age-specific mortality rates decreased or were stable between The increase in mortality among women aged 25-29 was mainly attributed to a higher number of suicide deaths. On January 20, 2020, the first case of COVID-19 in Korea was officially confirmed by the Korea Centers for Disease Control and Prevention (hereafter "KCDC, ", the predecessor of the Korea Disease Control and Prevention Agency [KDCA], which was expanded, reorganized, and renamed on September 12, 2020) [5]. Korea had a successful COVID-19 response, including a testing-tracingtreatment (3T) strategy, swift application of focused, social distancing measures in high transmission areas, border control, and risk communication [4,6,22]. The cost of COVID-19 testing, quarantine, and treatment was covered by the Korean government and national health insurance program to reduce the economic burden [22]. Financial support for hospitalized or quarantined people was also provided [23]. These countermeasures led to a relatively low rate of confirmed cases and a low fatality rate, despite three COVID-19 waves in 2020 [24][25][26]. Based on the findings of this study, 950 deaths from COVID-19 were confirmed in 2020 death certificate data, accounting for only 0.3% of the total deaths. Along with the overall decrease in deaths attributed to other causes, the age-adjusted all-cause mortality rate in Korea decreased in 2020. This is in agreement with the findings of other studies analyzing the total mortality rate in Korea, which reported no excess deaths in 2020 [8,12].
However, a significant increase in death from sepsis may suggest the direct or indirect effects of COVID-19. The manifestations of COVID-19 range from asymptomatic to life-threatening sepsis [27], and sepsis findings were reported in all deceased patients in a previous study [28]. If death occurred after a COVID-19 polymerase chain reaction test was negative and the patient was transferred from the isolation ward to the general ward, the cause of death on the death certificate may only list sepsis without any mention of COVID-19 [29]. However, appropriate treatment for sepsis due to other causes may have been delayed or insufficient. The Korean government has prioritized medical resources for COVID-19 treatment to cope with the COVID-19 pandemic [30]. At the end of 2020, the government secured around 6,000 beds in infectious disease hospitals and supported the deployment of medical and paramedical personnel [31], which may have disrupted existing medical care. Delays in the management of patients with acute stroke or acute myocardial infarction during the pandemic were reported [14,15,32]. Additionally, longer or delayed emergency transfer times were noted [33]. In particular, in the case of fever with sepsis, emergency transportation was sometimes refused [34]. Therefore, it is necessary to monitor the occurrence of sepsis and associated death during the COVID-19 pandemic.
Cases of pneumonitis due to solids and liquids (i.e., aspiration pneumonia) also significantly increased. Approximately 5-15% of cases of community-acquired pneumonia may be attributed to aspiration pneumonia [35,36]; compared with other types of community-acquired pneumonia, aspiration pneumonia is associated with higher mortality [35]. The risk factors for aspiration pneumonia include impaired swallowing, degenerative neurologic dis-eases, impaired consciousness, poor dentition, and old age [35,37]. Aspiration pneumonia is a serious condition for those living in long-term care facilities, as these individuals are more likely to have risk factors [38,39]. Therefore, caregivers play a crucial role in preventing aspiration pneumonia [39][40][41]. As clusters of infections occurred in long-term care facilities among the elderly in the early stages of the COVID-19 pandemic, the government established guidelines for restricting visits (February 24, 2020) and eventually banned all visits (March 13, 2020) [42]. Cohort isolation was implemented when cases were confirmed in long-term care facilities for the elderly, and preventive cohort isolation was practiced in some places [43]. The quality of care for the elderly may have been reduced as the additional quarantine measures increased caregivers' work burden and prevented family visits [42,43]. Moreover, similar to cases of sepsis, treatment for patients with fever associated with aspiration pneumonia may have been delayed [43]. A more thorough investigation of aspiration pneumonia deaths, taking into account the COVID-19 pandemic in 2021-2022, is required.
In this study, only the mortality rate for women aged 25-29 increased significantly in 2020 compared with 2019, which could be explained by a considerable increase in suicide deaths. However, the increase in the suicide rate of young Korean women is a trend that has continued since 2017 rather than a specific consequence of the 2020 COVID-19 pandemic, and it is not only limited to women aged 25-29 but also common among men and women in their 20s. The suicide rate from 2017 to 2020 in Korea increased from 16.7 (per 100,000) to 19.8 for men aged 20-24 and from 25.3 to 27.5 for men aged 25-29. The suicide rate of women was higher than that of men and increased from 9.5 (per 100,000) to 19.3 for women aged 20-24, and from 13.4 to 19.4 for women aged 25-29 [44]. Unemployment, job insecurity, financial hardships, social exclusion, and social isolation have been suggested as social risk factors for suicide among young adults [45,46]. Specifically, gender inequality and conflict may influence the suicide rate of young women [45]. However, the COVID-19 pandemic could also be related to suicide, as the risk of suicidal thoughts may have increased due to quarantine, exhaustion, social distancing, unemployment, and financial hardships [47]. Therefore, it is necessary to monitor changes in suicide patterns as long as the COVID-19 pandemic persists.
Senility, pneumonia, and chronic lower respiratory disease also showed significant changes in 2020 compared to 2019. However, when all of these diseases were analyzed by extending the period, it was difficult to judge that the change was a specific change from 2019 to 2020. However, since a decrease in pneumonia and chronic lower respiratory disease during the COVID-19 period has been reported in other studies [48], it is necessary to follow up in 2021 and 2022.
The study findings should be interpreted with caution due to certain limitations. First, death certificate data may be inaccurate, which could directly affect the accuracy of the study results. All deaths in Korea must be reported to the National Statistical Office within 1 month. When reporting the cause of death, death certification by a doctor is essential. In Korea, the rate of death registrations accompanied by a doctor's medical certificate of death has continuously increased over the years, surpassing 90% since 2007, and after 2015-this study's objective period-it exceeded 99% [49]. However, issues could arise due to the doctors' lack of training and expertise in death certification [50]. A non-differential misclassification bias caused by the inaccuracies in recording the cause of death may not significantly affect the direction of annual comparison results, but it may have limitations in calculating an accurate mortality rate. Second, the projected mortality rate and the existing mortality rate may be compared to examine excess deaths; however, in this study, 2019 and 2020 were directly compared. For a more rigorous interpretation, we also analyzed whether the absolute death toll was relatively high and whether there were cases where the mortality rate exceeded the average between 2015 and 2019.
Despite its limitations, this study shed light on the issues of sepsis and aspiration pneumonia despite the successful response to COVID-19 in Korea. Cases of deaths from sepsis and aspiration pneumonia should be identified and monitored and preventive measures should be implemented. A proactive policy to address suicide among young people, especially young women, must be developed.