Since the last report on the epidemiology of the Korean outbreak1 on March 2, 2020, when the number of fatal cases was 22 patients, the mortality number has increased to 54 cases as of March 10, 2020. Fig. 1 shows the timeline for the occurrence of fatal cases in Korea.

The first fatal case (national patient number 104) occurred on February 19 and was officially reported on February 20 by Korea Centers for Disease Control and Prevention (KCDC).2 National patient number 443 (mortality patient number 3) was the first patient who died suddenly without an epidemiological link and was diagnosed with COVID-19 after death on February 21.3 Since the patient (national patient number 31) was identified and related to the religious group was reported on February 18, there was a surge of the confirmed cases with hundreds of new cases per day, mainly in Daegu and Gyeongsangbuk-do area.1 With a significant surge of new cases, there was the first fatal case (national patient number 1,443; mortality patient number 13) that the patient died during home isolation after having been confirmed with COVID-19 on February 27, 2020.4 Since March 1, referrals from Daegu area to hospitals in Seoul (the capital city of the nation with more tertiary hospitals) and other parts of the area in Korea were started for serious patients who needed intensive care. On March 2, patients in mild conditions were relocated to “life treatment centers” that were temporarily transformed for housing the mild confirmed patients in individual rooms (Supplementary Fig. 1).5 These facilities were initially been used as accommodation facilities for human resource training centers of the local government, universities, or corporations.6 On March 4, there was the first case of mortality in a patient (mortality patient number 32, the national patient number was not available) without significant underlying illness. As of 0 am, March 10, 2020, there were 54 out of 7,513 confirmed patients with COVID-19 died in Korea.7

It is of importance to expect what would happen during the outbreak as the epidemic evolves with increasing numbers. Fig. 2 shows the mortality related milestones, the cumulative number of confirmed cases, number of daily new cases, and cumulative number of cases that have been released from isolation in Korea, by March 10, 2020.189 Therefore, reviewing these data is important to assess the impact on the patient-care capacities of the healthcare facilities, especially during a surge of an outbreak.

The first death due to COVID-19 occurred on February 19 and was reported on February 20. This patient was the national patient number 104.2 The first sudden-death case with postmortem diagnosis of COVID-19 was observed when the total number of patients reached 204 on February 21.3 The first death on home isolation due to COVID-19 occurred when the number of daily new patients reached 505 with a total number of over 1,766 on February 27.4 First death without underlying illness occurred when the total number of patients reached 5,328 on March 4.10 Of note, although without significant illness, this was a 72-year-old woman who developed symptoms on February 23 and was diagnosed with COVID-19 on February 26 after she became hospitalized to intensive care unit and died on March 4, 2020.10

It is of note that even when the total confirmed cases reached 7,513, only 247 patients were released from isolation after treatment on March 10, 2020 (Fig. 2).

Fig. 3A shows the daily number of deaths, daily case fatality rate (CFR) (number of new deaths/number of total patients on each day), and daily cumulative CFR (number of cumulative deaths/number of total patients on each day).8 During the first week after the first morality case, the daily CFR was 1.22%. However, it soon decreased to 0.04% on March 10, 2020. The cumulative CFR of COVID-19 in Korea was 0.7% (54 deaths of total 7,513 confirmed cases) as of 0 am, March 10, 2020.

Fig. 3B shows the mortality distribution in the nation as of March 10, 2020.7 As expected, Daegu and Gyeongsangbuk-do province had most of the mortality cases. One patient from Gangwon-do died at a hospital in Gyeongsangbuk-do and had a funeral in Gangwon-do. One patient from Gyeongsangbuk-do died at National Medical Center in Seoul, and the other patient (A Mongolian patient) died in Gyeonggi-do.

Table 1 shows the baseline characteristic of fatal cases in Korea. The median age at death was 75.5 years old (interquartile range [IQR], 66–80; range, 35–93 years old). Men was 61.1% (33/54). Forty-nine patients (90.7%) had an underlying illness. The most common underlying illness was cardiovascular disease such as hypertension and heart diseases followed by diabetes, and neurologic diseases (mostly dementia and stroke).

CFR of men was higher than that of women (men, 33/2,852 [1.16%] vs. women, 21/4,661 [0.45%]; P < 0.001; χ² test). Gender difference on CRF was more evident in the older age group in patients 60 years or older (men 8/592, 4.73% vs. women 19/1,013, 1.88%; P = 0.002).

Among the age groups, CFR was higher in older age group (< 20 years, 0/460 [0%] vs. 20–50 years, 2/4,032 [0.05%] vs. ≥ 50 years, 52/3,021 [1.72%]; P < 0.001; Fisher's exact test). Similar results were observed when different age groups were compared (< 50 years, 2/4,492 [0.04%] vs. 50–70 years, 18/2,345 [0.77%] vs. ≥ 70 years, 34/676 [5.30%]; P < 0.001; all comparison and all pairwise comparisons).

There were eight immunocompromised patients (8/54 [14.8%], seven cancer patients, and one kidney transplant recipient). There were seven patients with the psychologic disease. This was mainly related to the cluster of COVID-19 transmission among hospitalized psychologic patients in a mental ward of Daenam Hospital where the religious groups' leader had his family funeral.1 This religious group was later found to be the main cause of COVID-19 surge in Daegu and Gyeongsangbuk-do area.

We tried to analyze survival duration from symptom onset to death in 31 patients (57.4%) of whom the symptom onset dates were available (Table 1 and Fig. 4A). Among the 31 patients in whom the dates of symptom onset were available, the median days from symptom onset to death was 10 days (IQR, 5–12 days; range, 35–93)

Duration from symptom onset to death was not different between men and women (data not shown, P = 0.677) or the two age groups (< 70 years old [n = 9] vs. ≥ 70 years old [n = 22]; P = 0.086) (Fig. 4B). It also did not differ between the patients with underlying illness (n = 28) and without underlying illness (n = 3) (P = 0.499, log-rank test) (Fig. 4C). However, it is of note that all three patients without underlying illness were alive until day 10 from the symptom onset.

Fig. 5 shows the distribution of CFR in some selected countries, including Korea.11 CFRs of the three countries (Italy, Iran, and Korea) with the highest numbers of confirmed cases next to China, were 5.0%, 3.0%, and 0.7%, respectively. In addition, CFRs in the US and Italy were similar (5.2% and 5.0%), although the numbers of confirmed cases were different.

Fig. 5 shows the age distribution of the general population in the background, the proportion of age groups among confirmed cases, and CFRs in each group in Korea (as of March 10), China (as of February 21), and Italy (as of March 9). Age distribution of the Korean outbreak shows M shape with peak ages in the 20s and 50s (Fig. 6A)712 while Chinese data shows bell shape with a peak age in the 50s (Fig. 6B).1314 In Italy, it appears that more individuals in older ages were infected compared to Korea or China (Fig. 6C).1516

CFRs were compared among the three countries. Although there seemed a trend, when adjusted for age (< 50, 50–70, ≥ 70 years old) effect the CFR of Korea was not significantly lower than that of China (Korea vs. China, P = 0.054) or that of Italy (Korea vs. Italy, P = 0.124). When adjusted for country effect, CFR of older age group was higher than those of younger age group (< 50 years old vs. ≥ 70 years old, P < 0.001; 50–70 years old vs. ≥ 70 years old, P < 0.001, Poisson regression).