The target period of the study was between January 20, 2020, when the nation’s first COVID-19 case was identified, and the end of April 2020, when the first wave of the pandemic stabilized. The city of Daegu and the KCDC announced daily counts of COVID-19 cases, deaths, and the number of people released from self-quarantine [14,15]. We examined both official databases to understand the pattern of the pandemic, and graphically represented the daily and cumulative incidence.

To consider the timeliness of Daegu’s public health emergency response, we classified the target period into 4 phases: the preparatory phase and phases 1–3. Agreement was reached to use the trends in daily and cumulative incidence and major changes in the city’s response strategies (e.g., the introduction of residential treatment centers and the conversion of the quarantine from city-led to citizen participation–based) as criteria for classification by phases. The adequacy of the classification criteria was subsequently reviewed with external specialists.

Our target period was equivalent to stages 1 and 2 of the KCDC report [16], which organized the domestic situation into 5 stages. Stage 1 in the report was from January 20 to February 17, 2020, when sporadic cases were coming in from abroad. Stage 2 was from February 18 to May 5, 2020, when the outbreak in a religious group was predominant in Daegu, coinciding with the target period of our study.

We referred to the emergency response models produced by the WHO, the EU, and the United States [10–13]. Of these, the Public Health Emergency Preparedness and Response Capabilities [10], established by the Centers for Disease Control and Prevention (CDC) in the United States, had more subdivided definitions for capabilities that we could use to inspect the city’s response in detail. The CDC model, announced in 2011 and updated in 2018, consists of 6 domains and 15 capabilities (Table 1). Each capability consists of (1) its definition, (2) the functions critical for accomplishing that capability’s definition, (3) the tasks necessary to achieve 1 or more functions, and (4) the resources necessary to perform the tasks [10]. Since Daegu’s response was not based on the CDC response model, it was decided not to evaluate their response based on the completion of a function or a task. Taking this practical aspect into account, we adopted the capability definitions as criteria for assessing how Daegu reacted to the first wave of the COVID-19 pandemic in Korea. In other words, we investigated whether the city’s responses fulfilled each of the capability definitions, as presented in Supplemental Material 1.

The city’s responses to COVID-19 were summarized by collecting the officially announced data and internal data of the city of Daegu, such as daily records of the COVID-19 response and meeting notes. The responses were sorted according to the capability definitions of the CDC model. Next, preventive medicine specialists who engaged directly in Daegu’s COVID-19 response evaluated whether the city’s response achieved the capability definitions, followed by a secondary evaluation by the entire research team. A final review was done during 5 advisory meetings.

This study was based on data open to the public. Ethical approval by an institutional review board and written informed consent were not required.

Figure 1 shows the status of both the daily COVID-19 cases and the daily number of people released from self-quarantine between February 18 and April 30, 2020. On February 27, 10 days after the first COVID-19 case was identified in Daegu, the cumulative number of confirmed cases exceeded 1000. On February 29, the number of daily confirmed cases peaked at 741. After the number of cumulative cases reached over 5000 on March 7, the number of daily cases was approximately 100, decreasing to less than 100 on March 11 and remaining in single digits after April 8. The number of people released from self-quarantine steadily increased and became greater than the number of daily confirmed cases on March 11.

Figure 2 shows the daily death counts and the case fatality rates. On March 7, approximately 2 weeks after the first COVID- 19 death was identified, the number of confirmed deaths reached 33. The highest number of daily deaths was 7 on March 10, and the number of cumulative deaths until March 31 was 111. With 1 death to 5 deaths occurring every day in April, the cumulative deaths by April 30 was 169. The case fatality rate was 2.43%.

Daegu’s response was divided into the preparation phase and phases 1–3, according to COVID-19 status and the timing of the introduction of quarantine policies. Table 2 shows the major events during the early pandemic in Daegu and the responses by authorities, as well as the national crisis alert level corresponding to each phase.

During the preparation phase from January 20, 2020 (when the nation’s first case was confirmed) until February 16, 2020, Daegu prepared for the occurrence of cases within the community. As the crisis alert level was raised to Level 2 (caution) on January 20, 2020 and again to Level 3 (alert) on January 27, 2020 the city expanded the countermeasure team.

Phase 1, from February 17 (when the city’s first case was detected) until March 1, was the period when the number of confirmed patients surged in a certain religious group. The city responded by forming a public-private joint task force, and the government aided by forming a pan-government support team. More hospitals were designated for COVID-19 treatment, and doctors monitored patients waiting to be hospitalized by telephone. As a full-scale investigation of the religious group was conducted, the increasing demand for tests was handled by arranging drive-through testing and increasing the number of screening centers. On February 23, the crisis alert level was raised to Level 4 (severe), the highest level, as preemptive action against nationwide spread of the virus.

Phase 2 ran from March 2 (when residential treatment centers began to operate) until April 6. The implementation of designated COVID-19 hospitals and residential treatment centers allowed for a more stable treatment environment. However, the response concentrated on fighting the mass infections in convalescent hospitals and psychiatric hospitals. Although there were attempts to transfer groups of confirmed patients out of these facilities, it was difficult to secure beds because of limited nursing services or the difficulties in managing patients with mental disorders. This resulted in additional cases. By allocating one public official to each hospital and organizing a joint task force with the designated COVID-19 hospitals, it was possible to share information on inpatient status and facilitate patient transfers.

Phase 3 was between April 7 and the end of April. Despite stabilization of the early pandemic, long-term responses to COVID-19 appeared necessary. To relieve the physical and emotional exhaustion caused by prolonged high-intensity social distancing, the city formed a pan-citizen committee composed of 200 civic representatives. By holding regular videoconferences between the city and the committee to review government proposals, the 7 basic living rules for Daegu residents and quarantine rules for 9 areas including the economy, medical care, and education were established. While the crisis alert level remained at Level 4, the government announced that it would extend its social distancing rules for 16 days beginning April 20, but would relax some of the rules.

The evaluation of the response to the early pandemic in Daegu is outlined in Table 3 according to the CDC model capabilities.

First, the fundamental problem of the community preparedness capability (categorized under the community resilience domain), was that organized preparation for a public health emergency had not been developed in advance. In Korea, the Framework Act on the Management of Disasters and Safety, which stipulates preparation and restoration measures for social and natural disasters, was enacted in 2004, but was partially amended every 1–2 years [17]. Whenever a crisis occurred, the act was supplemented to correct shortcomings [18], leading the public to feel that the performance of both central and local governments was not effective [19]. The CDC, after defining infectious disease outbreaks, biochemical accidents, and natural disasters as public health emergencies and developing a response model, encouraged local governments to use the model in preparation for and response to emergencies [10,20]. The Korean government also needs to establish an integrated and systematic guideline for local governments and communities to prepare for and respond to any kind of emergency. In addition, the community recovery capability within the same domain should be taken into account while responding to a crisis, not after the crisis ends. Considering the economic and psychological impact of repeated emerging infectious diseases such as the H1N1 pandemic, the MERS outbreak, and the COVID-19 pandemic on the community, true recovery of the community can be achieved only by accepting a life that exists with these crises, not expecting or waiting for them to end. This is in line with the CDC’s definition of community resilience, which is the process of adapting to a crisis beyond resistance to it [10]. Furthermore, unlike recovery, which focuses on restoring normal conditions [21], resilience means strengthening the ability of a community to do what it can do during a crisis [10]. Thus, when recovery is implemented in terms of resilience, communities can continue their activities without having to wait for a crisis to end. This is especially important in a continually changing crisis like the COVID-19 pandemic [21].

Second, regarding the emergency operations coordination capability (under the incident management domain), in the early stage of an emerging infectious disease like COVID-19 when its characteristics are not fully known [22], governance must be flexible so that decision-making and rapid cooperation are feasible. Moreover, a single management system linked to health care and all other public and private sectors must be developed for emergency response. In Korea, where private medical services predominate, a close relationship with private medical institutions is critical, and prior cooperative governance between the Ministry of Health and Welfare, the Ministry of the Interior and Safety, the national government, and local governments must be established [17]. In Daegu’s response, it was desirable that public officials and preventive medicine specialists, both of whom could make major decisions, shared an office from the beginning to overcome any physical barriers to communication and to enable rapid cooperation.

Third, regarding the emergency public information and warning capability (under the information management domain), the city held regular briefings for the media and informed citizens of the pandemic status through various channels. Risk communication is the process of informing the public of an unexpected crisis, and the person in charge should carefully deliberate how to alert the public to the crisis. Appropriate content, format, and timing of risk communication can improve the situation, whereas inappropriate risk communication may worsen the situation [23]. The CDC suggests 6 principles of risk communication: be first, be right, be credible, express empathy, promote action, and show respect [23]. In an infectious disease crisis like the COVID-19 pandemic, each citizen is not simply an object of crisis management, but rather an agent for overcoming the crisis who can promote an early end to the crisis by complying with quarantine rules and trusting in the authorities. Therefore, it is vitally important to provide accurate information to citizens in a timely manner. Regarding the information sharing capability of the same domain, information was not shared in real time between the health and medical resource management systems operated individually by ministries [24], creating an obstacle to rapidly responding to the surge in COVID-19 patients. Furthermore, even though an integrated management system for diseases and health care was developed by the KCDC, its operation was unsystematic and it was inconvenient to use [24]. Recognizing these problems, it is urgently necessary to incorporate management systems that are spread across ministries or to introduce another, more efficient system that would enable local governments to share information and identify the status of health and medical resources as they respond to the public health emergency.

Fourth, regarding the medical countermeasure dispensing and administration capability (under the countermeasures and mitigation domain), the Central Clinical Committee proposed an initial patient classification system. Because this involved checking the vital signs of those waiting for hospitalization at home, it could not be applied on-site. Therefore, with the help of the Daegu Medical Association, a more practical patient severity classification system was introduced using the available epidemiological data and the National Health Insurance Service data on underlying diseases, allowing doctors to monitor patients by telephone [7,25]. Regarding the medical material management and distribution capability of the same domain, medical supplies must be managed and dispensed according to the guidelines once they are supplied. Next, regarding the non-pharmaceutical interventions capability, citizens voluntarily engaged in social distancing and, as a result, helped to control the early COVID-19 pandemic and prevent a nationwide spread. Regarding the responder safety and health capability, protection systems were not adequate for either public officials or medical staff. The safety of responders must be assured by identifying risks and providing both physical and psychological protection [26–29].

Fifth, regarding the fatality management capability (under the surge management domain), the government quickly distributed guidance for the funerals of those who died with COVID-19 [30]. In Phase 1 of Daegu’s response, deaths occurred in severely ill patients or patients whose symptoms worsened while waiting for hospitalization. Delays were due to limited availability of the negative-pressure rooms being used for confirmed cases. In Phase 2, many deaths occurred due to the mass infections in convalescent hospitals for elderly patients. Daegu quickly formed a response team based on the government’s funeral guidelines that had been distributed 6 days after the city’s first case was confirmed. Regarding the mass care capability of the same domain, mass infections in convalescent hospitals and psychiatric hospitals resulted in many COVID-19 cases. Public officials were assigned to those hospitals that lacked infectious disease specialists and were in charge of overall management, including the isolation of close contacts and patient transfers. Next, regarding the medical surge capability, the rapid increase in the demand for inpatient treatment was resolved with the prompt help of private hospitals. Confronting a public health emergency like an emerging infectious disease, a response organization should be able to determine the level of crisis; predict medical needs; establish plans to secure additional personnel, facilities, and equipment; and request necessary support from the private sector [10]. Regarding the volunteer management capability, there was dissatisfaction among responders over the volunteers’ wage differences. This issue impeded the emergency response process and needs to be improved.

Last, regarding the public health laboratory testing capability (under the biosurveillance domain), to meet the increasing demand for tests in Phase 1, Kyungpook National University Chilgok Hospital implemented drive-through testing based on a large-scale screening center model suggested by Kim Jin Yong at the Incheon Medical Center [8]. Safe and fast tests were performed, reducing contact with others and saving time because there was no need to leave the vehicle for sample collection [8]. This strengthened testing capabilities, which enabled comprehensive surveillance of the high-risk facilities, including the religious group and convalescent hospitals, and served as a key component in Daegu’s response. Regarding the public health surveillance and epidemiological investigation capability of the same domain, the 2-stage investigation, basic and in-depth, was challenging due to the rapid increase in confirmed cases in the early pandemic. The investigation method was adjusted by introducing phone surveys with simplified forms. Nonetheless, it was difficult to conduct an accurate investigation, including determination of close contacts and estimation of the onset of symptoms, because the personnel performing the surveys had insufficient pre-training.