How Hospitals Overcame Disruptions in the Early Stages of the COVID-19 Pandemic: A Case Study from Tokyo, Japan

ABSTRACT The COVID-19 pandemic has caused serious disruptions to health systems across the world. While the pandemic has not ended, it is important to better understand the resilience of health systems by looking at the response to COVID-19 by hospitals and hospital staff. Part of a multi-country study, this study looks at the first and second waves of the pandemic in Japan and examines disruptions experienced by hospitals because of COVID-19 and the processes through which they overcame those disruptions. A holistic multiple case study design was employed, and two public hospitals were selected for the study. A total of 57 interviews were undertaken with purposively selected participants. A thematic approach was used in the analysis. The study found that in the early stages of the pandemic, faced with a previously unknown infectious disease, to facilitate the delivery of care to COVID-19 patients while also providing limited non-COVID-19 health care services, the case study hospitals undertook absorptive, adaptive, and transformative actions in the areas of hospital governance, human resources, nosocomial infection control, space and infrastructure management, and management of supplies. The process of overcoming the disruptions caused by the pandemic was complex, and the solution to one issue often caused other problems. To inform preparations for future health shocks and promote resilience, it is imperative to further investigate both organizational and broader health system factors that build absorptive, adaptive, and transformative capacity in hospitals.


Introduction
The COVID-19 pandemic has caused serious health systems disruptions worldwide, regardless of the socioeconomic status or geographical location of the impacted population.The early days of the pandemic, in particular, brought enormous challenges for health care providers responding to the unexpected spread of an unknown virus while also delivering health care services to the population. 1 The response to the pandemic by many health systems was constrained by inherent issues, such as health workforce shortages, fragmented health care service delivery, broken supply chains for medical goods and equipment, and inadequate health information systems. 2 Although the importance of resilience has been argued since the occurrence of the global financial crisis and the Ebola outbreak, [3][4][5] the COVID-19 pandemic highlighted its importance in health systems, 6 where resilience refers to their ability to absorb, adapt, and transform in response to the disruptions caused by a crisis and to maintain their essential health care function during the crisis. 7Hospitals, as key health system actors in service provision, have played a major role in providing care to COVID-19 patients while also maintaining routine service delivery. 1,8,9There are a considerable number of studies looking at health system resilience in response to COVID-19, [10][11][12] but studies comprehensively examining hospital responses to the pandemic are limited. 6,13Consequently, it is important to examine how hospitals and health professionals involved in frontline testing and treatment adapted, responded to, and transformed the trajectory of COVID-19.
While the pandemic has not ended, this study focuses on the experience and responses to COVID-19 by hospitals and health care providers in the first and second waves in Japan, i.e., from March 2020 to September 2020.Japan's first COVID-19 case was identified on January 15, 2020, in a returnee from Wuhan, China. 14In February 2020, an outbreak of COVID-19 occurred on a cruise ship anchored off Yokohama.Of the 1,068 crew members and 2,645 passengers, 712 were confirmed to have COVID-19, of whom 14 died. 15 surge of cases began in March 2020.In response to the first wave of infections (March-May 2020), a state of emergency was declared between April 16 and May 25, 2020.While more cases were recorded during the second wave (July-September 2020) than in the first, a nationwide state of emergency was not declared due to a reduction in severe cases and a lower death rate. 14ather, the government introduced policies to revitalize the economy while easing restrictions on domestic travel and mass gatherings. 16The second wave was characterized by increasing social transmission of infection (through activities such as restaurant dining or visits to red-light districts).By December 2022, more than 27 million people in Japan had been infected and approximately 54,000 had died from COVID-19. 17he Japanese case study was undertaken as part of an overarching multi-country study on public health and hospital resilience in the COVID-19 pandemic, the HoSPiCOVID project, which examined the processes by which hospitals responded to the COVID-19 pandemic. 18That broader study aims to understand how health care provider resilience to external shocks that disrupt health systems, such as COVID-19, can be strengthened.The present paper specifically examines disruptions experienced by hospitals as a result of the COVID-19 pandemic in Japan and the processes through which they adapted and responded to the disruptions and transformed health services delivery during the early stages of the pandemic.

Study Design and Settings
The study employed a holistic multiple case study design, which involves two cases or more and a single unit of analysis 19 ; here, the cases are two public hospitals that played a key role in COVID-19 testing and treatment, and the unit of analysis is the process of overcoming disruptions caused by the COVID-19 pandemic.Under Japanese law, certain hospitals are assigned to treat patients with "designated infectious diseases," with exceptions determined by prefectural health administrations. 20At the beginning of the pandemic, COVID-19 was categorized as a designated infectious disease, so one hospital (Hospital 1) assigned to treat infectious diseases was selected for the present study.However, due to the magnitude of the pandemic, health facilities other than those assigned to treat designated infectious diseases also provided care to COVID-19 patients, so the study also selected another hospital (Hospital 2) that was not specifically assigned to treat COVID-19 patients but accepted them as patients.The two hospitals are in the same district but had different functions in the early stages of the pandemic: a tertiary hospital that mainly provided care to severe COVID-19 cases, and a secondary hospital mainly responsible for treating patients with moderate symptoms.Table 1 presents the hospitals' key characteristics.

Data Collection
Data were collected between September and December 2020 from key informant interviews and documentation review.The HoSPiCOVID project 18 analytical framework was used to examine disruptions experienced by health care providers due to COVID-19 and the processes by which hospitals absorbed, adapted, and transformed in response.The analytical framework was used to prepare a semi-structured interview guide with five categories of questions: (1) respondent characteristics; (2) health care service delivery organization in their department during the early stages of the pandemic; (3) challenges caused by the pandemic and how the department responded; (4) factors that facilitated or constrained their response; and (5) overall lessons learned from their experience in the early stages of the pandemic.Three versions of the questionnaire were prepared, adapted with slightly different wording for the different professional groups.The interview guide was pilot tested with two hospital staff members to ensure clarity, after which minor changes were made to the wording of some questions.
Fifty-seven (57) individual interviews were undertaken.Participants were purposively recruited to reflect variations in: (1) job category (physicians, nurses, laboratory technicians, pharmacists, radiologists, clinical engineers, hospital administrators); (2) roles within the hospital and its departments (administration, middle managers, frontline workers); and (3) level of involvement in COVID-19-related services provision.After finalizing the list of participants in consultation with the study hospitals, the study team issued a letter of invitation through the relevant departments.One potential participant did not respond, and another could not participate due to a COVID-19 infection surge in December 2020.Table 2 shows the numbers of participants by professional groups.The number of interviews varied between the two hospitals due to their different sizes.In addition to staff at the study hospitals, individual interviews were undertaken with local health administrators.
Interviews were conducted face-to-face, in Japanese.Before starting, the study's objectives and contents were explained and participants' consent was obtained.To ensure consistency, one team member led most of the interviews, while other members attended and asked supplementary questions.To avoid interviewer influence on participant responses, one team member took notes throughout the interview and monitored the interactions between participant, main interviewer, and team members.
To triangulate the information obtained from interviews, documentation was also reviewed, including hospital records and documents, published material on the COVID-19 pandemic in Japan, academic publications, and policy documentation for both national and prefectural governments.

Data Analysis
With participants' consent, interviews were audiorecorded and transcribed verbatim for analysis.NVivo software was used to facilitate qualitative data analysis and manage interview data and analytical codes.The analysis used a thematic approach, in which pre-coded themes were derived from the analytical framework.Themes not predicted from the framework were coded as they arose during data analysis.
Resilience refers to health systems' capacity to maintain essential functions during a crisis or "health shock."It involves a process of absorption, adaptation, and transformation. 7The analysis investigated the processes of absorption (understanding/persistence), adaptation (incremental change), and transformation (longerlasting systemic change) 21 undertaken by hospitals contending with disruptions to normal work practices as a result of the COVID-19 pandemic.
The analytical framework was used to present systematically the process by which hospitals overcame disruptions, specifically: (1) disruptions experienced by health care providers ("effects"); (2) the processes of absorption, adaptation, and transformation providers undertook to deal with disruptions ("strategies"); and (3) consequences, positive and negative, of the strategies ("impacts"). 18f the dimensions indicated in the conceptual framework for the multi-country study, the present case study identified five key areas (hospital dimensions) impacted by the COVID-19 pandemic: (1) governance arrangements for accommodating COVID-19 patients; (2) adequacy of human resources to ensure safe delivery of COVID-19 health care services; (3) containment of nosocomial infection risk; (4) reorganization of space and infrastructure to create sufficient bed capacity for COVID-19 patients; and (5) management of the supply of medicines and consumables to prevent shortages.In the Results section, these five dimensions are used to describe the disruptions experienced by health care providers ("effects"), their responses to the disruptions ("strategies"), and the consequences of those responses ("impacts").
To ensure objectivity in the qualitative analysis, the study employed diversity in information sources (i.e., different professional groups), data analysis by multiple researchers with different professional backgrounds, and a process to confirm information obtained from interviews by comparison with published literature.

Effects
In February and March 2020, as cases increased in Japan, hospitals became concerned about preparedness to accept COVID-19 patients and undertake infection control measures while continuing to provide non-COVID-19 services.While prefectural governments coordinated local health care providers and endeavored to increase bed numbers, not all hospitals initially accepted COVID-19 patients. 16Consequently, hospitals treating COVID-19 patients bore the burden of increased workload and COVID-19 infection risk.When the COVID-19 surge commenced in March 2020, the workload of those providing care to COVID-19 patients increased significantly, and hospital staff were very anxious about the "unknown" virus.
The number of patients rapidly increased before we fully realized what was happening.We still needed to do our jobs and provide care to [COVID-19] patients.While we were always anxious, there was no time for us to give any thought to, or fully understand, what the anxiety was all about (Nurse, Hospital 1).

Strategies
The study hospitals accepted COVID-19 patients mainly in line with their designated roles: one is designated to treat infectious diseases, and the other is a public hospital with a strong community health focus.After informing the staff of their decisions, hospital managers created task forces, in collaboration with hospital infection control teams (ICT), as platforms to discuss operations during the COVID-19 pandemic.
The task forces introduced structures for close collaboration between hospital management and infection control experts, with representation from several hospital departments, establishing communication platforms to coordinate and facilitate timely collective action.
To create consistency and coherence in the response to the pandemic, we organized a task force in such a way that those with expertise in infection control worked with hospital managers to form the core of the task force and representatives from all hospital departments were included in the task force to cover all groups working in the hospital (Manager, Hospital 1).
In one hospital, in consultation with the ICT, the task force facilitated the development of "business continuation plans" (BCPs) for each clinical department to describe the department's function and operational procedures while the hospital accommodated COVID-19 patients.
To address the maldistribution of burden and risk among health facilities in a district, one hospital took the initiative to coordinate local health facilities, the local health administration, and the local government to create a temporary referral system for allocating COVID-19 patients to different hospitals in the district.
Things moved quickly . . . .Discussions were held with the directors of the eight hospitals in the district, with which we closely collaborate, and we also talked to the district medical association.Then, together with the head of the district medical association, we asked the district head to create a model for delivery of PCR testing and COVID-19-related care . . . .Bed control and the expansion of PCR testing were a big challenge at that time . . . .Based on the discussions, we designed a blueprint, which was again discussed with the district government . . . .The head of the district government agreed with our ideas and to use the district budget to implement the proposed model (Physician, Hospital 1).

Impacts
The study hospitals established general arrangements for treating COVID-19 patients, while also providing limited non-COVID-19 health services.Task forces helped strengthen coordination between clinical departments, accountability mechanisms for explaining hospital managers' decisions, and communication platforms to share information about the situation in the hospital.
Information on COVID-19 was emailed regularly, which helped update our knowledge on how many COVID-19 patients were currently in our hospital (Nurse, Hospital 2).
The regional referral model facilitated collaboration between health facilities in the same district and enabled facilities to share roles and responsibilities for COVID-19 testing and treatment in the district.
[Due to the district-wide model of collaboration for COVID-19 testing and care], the district took responsibility for running COVID-19 outpatient wards . . . .The medical resources in the district were redistributed and other hospitals also decided to accommodate COVID-19 patients . . . .The situation in which only infectious disease certified hospitals provided COVID-19-related care was finally over (Physician, Hospital 1).

Effects
During the first wave, there were insufficient clinical staff to care for COVID-19 patients due to the rapid increase in patient numbers, infection control treatment taking longer than with usual patients, and hospital staff being required to take on additional tasks.Moreover, when in-hospital infection occurred, staff who had been in close contact with those infected had to quarantine, further constricting staff availability.Organizational routines were significantly adversely affected by the inadequate numbers of clinical and specialist staff in intensive care units (ICUs) to provide necessary care to severe COVID-19 cases.
Throughout the first and second waves, it was challenging to maintain motivation and allay fears of infection in clinical staff.The underlying causes of frontline providers' stress included: maldistributed, heavy workloads and responsibilities; an unfamiliar method of providing care due to infection control measures; fear of infection through work; and continuing pressure for individual providers to remain uninfected.Also, early in the pandemic, misconceptions about COVID-19 and discrimination by the general population against health care providers added to their stress.
In fact, the number of nurses resigning from the hospital has increased . . .and there is an increasing number of nurses taking leave due to psychological pressure and stress (Physician, Hospital 2).
Early in the pandemic, hospital management and frontline providers had different understandings of the abrupt challenges created by COVID-19, including the extraordinary work overload among those treating COVID-19 patients, which caused stress in those at the frontline of health care service delivery, particularly as the pandemic continued.

Strategies
To increase the clinical staff available for COVID-19 patients, hospitals used temporary staff from other hospitals particularly to care for severe COVID-19 cases early in the first wave, mobilized clinical staff from multiple hospital departments to provide care to COVID-19 patients (task-sharing between departments), and shifted tasks to administrative staff and nurses, particularly when external, non-clinical services were suspended.
To address fear, stress, and motivation issues, middle managers at department levels increased communication and dialogue with their staff and catered to individuals' needs and concerns.
Our department holds a daily morning meeting, and there I spend lots of time sharing information about COVID-19 and what needs to be done to prevent infection, as I think the right information about control measures reduces fear in my departmental staff (Physician, Hospital 1).
Innovative approaches were taken to motivate staff.In one of the study hospitals, a "total pain approach" was introduced for the care of COVID-19 patients, which helped revive professionalism in nurses.
[With heavy workloads and many additional tasks] a number of nurses began to wonder what nursing was all about and were losing meaning in nursing care . . .and so to buoy up their motivation as nurses, I gave some thought to what could be done to engage with patients from a nursing perspective.This is something we're continuing now.We realized that COVID-19 patients often face psychological and societal "pain."I'm a nurse certified in palliative care and thought we could apply the "total pain approach" [used in palliative care] to the care of COVID-19 patients and ask patients what they were struggling with . . .this way, patients could also develop self-care ability after being discharged from the hospital (Nurse, Hospital 2).

Impacts
The hospitals applied strategies to increase human resources available to provide care to COVID-19 patients, which included creating multi-departmental care teams.
In our hospital, there was a shortage of intensive care beds.At the beginning [of the pandemic], we received help from outside the hospital . . .but eventually, other departments within the hospital took on some tasks, so some cases were jointly treated by the Department of Respiratory Medicine, and some by the Intensive Care Department, and also, in cases requiring ECMO, by the Emergency Department . . . .Indeed, it was not possible for only the Infectious Diseases Department to provide treatment [during the rapid rise in cases] and collaboration with other departments was necessary (Physician, Hospital 1).
However, due to different clinical backgrounds, team care involving different departments sometimes resulted in disagreements about the approach to care.
[Clinicians from different clinical departments] sometimes look at problems from different directions.At the beginning we discussed [care strategies for a case], but [the discussion] could actually create some stress (Physician, Hospital 1).
Task-shifting also created additional work for hospital staff, especially nurses and administrative staff, whose workloads increased significantly.
We [administrative staff] all assisted in the outpatient ward reception, created by the district government within the hospital, and worked with PPE for almost the whole day, and after returning to the office we also needed to complete our usual work, as the hospital continued to run normal tasks . . .In March and April [2020], we felt that no matter how much time and how many staff we had, it was never enough (Hospital administrator, Hospital 1).

Effects
At the pandemic onset, given the unknown nature of the virus, there were no clear guidelines for containment of risks associated with COVID-19 hospital infection.Methods for infection control while continuing normal service delivery were unclear, raising fear and concern among hospital staff.Furthermore, it was difficult to avoid contact between patients infected with COVID-19 and other patients.In fact, on several occasions, emergency patients and/or inpatients with other ailments tested positive for COVID-19 after attending hospital.The hospitals found it difficult to strictly impose infection prevention measures, such as mask wearing, on patients.There were patients who did not wear masks at that time [onset of the first wave] . . .even when we gave them paper towels to cover their mouths [as they did not have masks] . . .There were some who did not follow instructions (Nurse, Hospital 2).
It was also necessary to prevent COVID-19 from entering the hospital via non-patient visitors, such as suppliers of materials and equipment, external service contractors, and participants in training activities.

Strategies
The ICTs provided regular advice on infection control measures, daily supervision of all departments to ensure adherence with measures, and on-demand advice for hospital staff.ICTs created COVID-19 hospital infection control manuals, which included guidelines for preventing nosocomial infection.
Because circulation of staff and patients increases infection risk, care delivery procedures were modified to minimize staff movement.Hospitals also introduced thermographic screening of those entering the hospital, enforced hand sanitizer use, and restricted visitors by not allowing family and friends to visit in-hospital patients.
Patients suspected of having COVID-19 are the most difficult . . .for CT scans, we allocated two staff members-one who was in contact with the patients for the scan and the other who operated the computer.That way we could separate the space between the "clean area" and the "risk area" and also use only one set of PPE . . .Also, for other tests, we divided the time [that testing facilities and equipment were used] . . .When we had to do tests for COVID-19 patients during the day, we stopped tests for non-COVID-19 patients and did their [non-COVID-19] tests in the evening so we could avoid contact between COVID-19 and non-COVID-19 patients (Radiologist, Hospital 1).
In addition, all suppliers were prohibited from entering hospital buildings and all training courses involving external participants were postponed.Furthermore, restrictive measures were placed on hospital staff to reduce the chances of infection outside the hospital.These included prohibiting physicians from working outside the hospital, restricting work trips, and prohibiting multiple people from dining out together.

Impacts
In general, the strategies implemented helped minimize nosocomial infection risk.Regular communication and supervision of hospital staff by ICT members strengthened knowledge of hospital infection risk and prevention and contributed to infection risk minimization practices in hospital staff.
Hand washing, finger disinfection, mask wearing-ICT requires us to undertake preventive actions almost 100% of the time, so our awareness of infection control absolutely increased (Paramedics, Hospital 1).
Conversely, restrictions on physical movement and face-to-face communication within the hospital made it difficult to provide certain types of services, such as team-based medicine involving frequent face-to-face interaction between clinical staff from multiple departments and nursing care involving close communication with patients.Also, the restriction on daily activities outside the hospital, such as dining out, removed a means of stress relief for hospital staff.

Effects
Japanese law stipulates that designated infectious diseases be treated at specific hospitals.However, in emergencies and/or unavoidable circumstances, prefectural governments are legislated to decide whether other hospitals can provide care for designated infectious diseases.Local health administrators struggled to find beds to accommodate symptomatic COVID-19 cases, particularly in areas with high infection rates. 22The unexpected rapid increase in COVID-19 patients in March-April 2020 resulted in a frequent lack of hospital beds.
We thought, in Tokyo, if they [COVID- 19 patients]  were well distributed between hospitals, the situation would be manageable . . .Well, eventually we managed, but initially many large hospitals did not accept them and it was a near thing, but eventually, little by little, those hospitals started to accept COVID-19 patients and the situation improved [in terms of beds available for COVID-19 patients].However, a negative consequence was on ambulance transport, where time was spent seeking beds to accommodate COVID-19 patients rather than finding beds within a hospital [overflow of patients] (Physician, Hospital 1).
Due to the pre-COVID-19 pandemic disease structure in Japan, many health facilities were not structurally designed or adequately equipped to accommodate large numbers of infectious disease patients.Typically, Japanese hospital buildings are structured to accommodate different clinical departments on different floors, or sections of floors, and usually have one main entrance and another for emergency cases, making it difficult to avoid contact between COVID-19 patients and other patients.Hospitals had only a limited number of the negative pressure rooms designed to prevent airborne disease transmission.

Strategies
The case study hospitals created designated ward and bed allocations for COVID-19 patients by relocating existing patients to other wards and/or hospitals.During the first wave, all available ICU beds were used for severe COVID-19 cases, while non-COVID-19 patients requiring intensive care were relocated to high care units (HCUs).To secure beds for COVID-19 patients, hospitals limited non-COVID-19 services to urgent care and procedures.Specifically, hospitals discontinued elective surgery requiring endotracheal intubation for general anesthesia.
The hospitals also created designated outpatient wards for those with COVID-19-related symptoms.At one of the study hospitals, a designated outpatient ward was further developed to be a district-wide PCR testing site for COVID-19, and the hospital parking area was used for this purpose.Early in the pandemic, due to the urgent nature of the problem, simple, quick approaches were used, rather than substantial changes to hospital infrastructure, to create negative pressure environments in the rooms for COVID-19 patients.

Impacts
Hospitals were able to treat COVID-19 patients with both moderate and severe symptoms and reduce risk of infection between patients by minimizing patient movement between beds and wards.To secure designated space for COVID-19 patients, hospital managers had to coordinate internally with hospital departments and externally with other hospitals.The designated space required more nurses and physicians and redistribution of clinical staff between departments.Limiting non-COVID-19 services resulted in decreased hospital revenue and maldistribution of workload among hospital staff.
Regarding staff workloads [as hospitals limited non-COVID-19 services], there was a divide between the staff whose workloads were considerably reduced, and those who experienced huge increases in workload (Physician, Hospital 2).

Effects
At the beginning of the pandemic, hospital staff feared shortages of disinfectants, masks, and other personal protective equipment (PPE).In March and April 2020, supplies of masks and disinfectant declined in the community, partly due to panic buying by the general population.Their market price increased, exacerbating the fear of shortages among hospital staff.
Before COVID-19, Japan relied largely on imports of masks, disinfectants, etc., and many of the raw materials used to produce medicines.The quality of masks, PPE, and other infection control supplies was negatively affected by limited availability of imported raw materials, which inhibited the production of infection control supplies in Japan.Some medicines became less available due to the reduced supply of imported materials.
Surgical masks have changed to ones made from cheaper materials.They used to be better . . .Now these, they were replaced by cheaper ones (Physician, Hospital 1).

Strategies
Hospitals changed infection control supply management practices as a result of the pandemic.For example, one of the study hospitals introduced centralized, rather than clinical department-based, management of supplies.Also, hospital staff minimized unnecessary use of supplies and used their own resources (such as goggles rather than face shields, when safe to do so) and/or hand-made materials (such as isolation gowns).Local governments distributed supplies to hospitals providing COVID-19 care, and infection control materials were donated by individuals and organizations.To address shortages of certain medicines, at pharmacists' request, physicians changed prescription practices to use available medicines.
Propofol, used for induction of anaesthesia, is usually imported by air freight from Germany, but we were informed it would soon be unavailable . . .So we needed to ask physicians to replace it with an alternative medicine . . .We had to discuss the matter with physicians (Pharmacist, Hospital 2).

Impacts
Infection control supplies did not run out during the first two waves but, at the beginning of the pandemic, their quality deteriorated.The steady supply of materials helped alleviate fear and stress in hospital staff, who ultimately felt the infection risk was well managed.
Thanks to the donation [of infection control supplies], as well as the careful use of the stock at the hospital, we managed to secure sufficient infection control related supplies, which resulted in a sense of security among hospital staff (Physician, Hospital 1).

Discussion
Early in the pandemic, faced with a previously unknown infectious disease, the Japanese case study hospitals applied absorptive, adaptive, and transformative strategies to deliver care to COVID-19 patients while providing limited non-COVID-19 health care services.These strategies were in response to issues associated with five hospital dimensions: governance; nosocomial infection control; human resources; space and infrastructure for COVID-19 patients; and supply of medicines and consumables.As COVID-19 cases increased, absorptive actions, such as accepting COVID-19 patients into existing hospital settings, led to adaptative strategies and subsequent incremental changes that allowed hospitals to continue to function.Where systemic changes were required, transformative strategies were also introduced.
The strategies employed by the two study hospitals to address issues arising in the five hospital dimensions were very similar, apart from a number of individual strategies specific to one hospital.The hospital-specific responses reflected the hospitals' roles and responsibilities in the health system.For example, at the beginning of the pandemic, Hospital 1 temporarily recruited staff (ICU specialists) from other hospitals, while Hospital 2 did not.Hospital 1 used this strategy because more ICUqualified/experienced staff were required as the hospital accepted a larger number of severe COVID-19 patients.
While the literature examining hospitals' responses to COVID-19 in-depth is limited, many strategies used in the Japanese case study hospitals were similar to strategies reported in other countries with similar socioeconomic settings.For example, in many countries in Europe and North America, to prevent nosocomial infection, non-urgent services were postponed during the first wave of the pandemic. 1Hospitals created designated COVID-19 units, mobilized additional clinical staff and/or reallocated existing hospital staff, and enacted temporary measures to mitigate shortages of personal protective equipment in order to treat COVID-19 patients while continuing to deliver care to other patients. 9A number of innovative solutions were reported that enabled hospitals to continue service delivery while facing various challenges caused by the pandemic, 23,24 as found in the present case study (e.g., using a palliative care approach for COVID-19 patient care to motivate staff).However, certain strategies addressed issues specific to the Japanese context.For example, due to the lack of clear referral mechanisms in the Japanese health system, temporary district-level referral mechanisms were needed to coordinate the allocation of COVID-19 patients to different hospitals in the district.
The process of overcoming disruptions caused by the pandemic was complex, and the solution to one problem (e.g., nosocomial infection control) could create problems elsewhere (e.g., heavier workload, psychological stress).However, in an adaptative response, hospitals established task forces that enabled close collaboration between managers and infection control experts, facilitated inter-departmental coordination and communication within hospitals, and supported other adaptative and transformative responses.Hospitals can better prepare for future health shocks by determining which factors enhance their capacity to absorb the situation, adapt practices, and transform operations.Several studies have examined factors that can strengthen health facilities resilience, including both "hard" (e.g., infrastructure) and "soft" (e.g., leadership and management) dimensions of organization. 6,25,26Building on the COVID-19 pandemic experience, however, it is imperative to further investigate hospitals' strategic actions for coping with health shocks, to better understand the interactions between dimensions of hospital operations, and to identify factors that support strategic action.
Broader societal issues outside the control of health care providers, such as community support, social services for working parents, people's infection prevention behaviors, etc., also affected the delivery of COVID-19 and non-COVID-19 health care during the initial stages of the pandemic.More importantly, structural and institutional issues in the Japanese health system (e.g., ratios of physicians and nurses per hospital bed; a legal framework governing bed control for infectious diseases; a high proportion of private health care providers; unclear referral systems within geographical regions) affected hospitals' ability to provide timely services to COVID-19 patients.Consequently, further study is required to identify the broader health system factors that foster health care provider resilience during health shocks.
A limitation of this study is that, while 81% of hospitals in Japan are privately owned, 27 only public hospitals were studied.Because public health facilities are mandated to provide health services for the public good, including diagnosis and treatment of infectious disease, more public than private hospitals accepted COVID-19 patients during the early stages of the pandemic in Japan.Furthermore, all designated medical institutions for treating designated infectious diseases, a category that included COVID-19 in the early stages of the pandemic, were public hospitals. 28In fact, in the district where the study was conducted, very few COVID-19 patients were treated at private hospitals.Consequently, only public hospitals were selected for the study.In Japan, due to uniform purchasing arrangements for both public and private health care providers and the financial autonomy given to public hospitals, the health care financing structure may not have been a determinant of responses to COVID-19 for hospitals of different ownership types.However, as both public and private hospitals are sub-categorized according to services provided and ownership type, further study is required to gain a more holistic view of hospitals' experiences and responses in the early stages of the COVID-19 pandemic in Japan.
Another limitation is that the study hospitals were both located in the same district in Tokyo, which precluded any examination of variations due to geographical location.Specifically, while there was considerable variation in COVID-19 infection levels across Japan, our study focused on one geographical context where a high number of infections was recorded.As such, this study was designed as an in-depth examination of hospitals with different functions under the same local health administration, and further study is needed to investigate variations in hospital responses according to factors associated with geographical context, including levels of infection.Moreover, the study used only a qualitative approach.While this was useful for exploring the process through which hospitals experienced and responded to the early stages of COVID-19, other approaches, including quantitative analysis, could be useful to further examine issues related to the study findings, including the effects of the strategies used in the study hospitals to address the disruptions caused by the pandemic.

Conclusion
In the early stages of the pandemic, faced with a previously unknown infectious disease, the Japanese case study hospitals undertook absorptive, adaptive, and transformative actions to provide care to COVID-19 patients while also providing limited non-COVID-19 health care services.The process of overcoming the disruptions caused by the pandemic was complex and the solution to one issue often caused other problems.It is imperative to further investigate the organizational factors that support the absorptive, adaptive, and transformative capacity of hospitals to inform preparations for future health shocks.In addition, further study is required to identify the broader health system factors conducive to health care provider resilience in times of health shocks.

Table 1 .
Characteristics of the study hospitals

Table 2 .
Individual interviews according to professional group