Management of the emergency response to the SARS-CoV-2 (COVID-19) outbreak in Stockholm, Sweden, and winter preparations

KEYwords: COVID-19; SARS COV-2; Emergency response; Primary Healthcare; Integrated health system; Implementation

To reduce the spread of the infection in early March 2020, the Swedish government passed laws to prohibit gatherings over 50 people, promoted border controls and limited visitors to care homes, among a number of other requirements. The Swedish constitution legally protects freedom of movement of its citizens and does not allow a ‘lockdown’ during peace time.¹ The Swedish public are expected to follow a series of ‘non-voluntary’ recommendations from the government agency responsible for public health, but sanctions are limited.²

The contrast of this approach in Sweden to approaches taken by other countries is reflected in the Oxford ‘COVID-19 government response stringency index’.³ This scores Sweden as 46/100 from 4 April 2020, compared to the scores for New Zealand of 100/100 and the UK of 76/100 on 26 March 2020, as shown in Figure 1. Despite the differences between Sweden and the UK, daily confirmed COVID-19 mortality per million and trends are almost identical for UK and Sweden, as shown in Figure 2.⁴ There are many possible explanations, but one is that fast and effective coordinated primary and community health-care responses in Sweden may have played some part.

Figure 1. COVID-19 government response stringency index for January–July 2020 (Hale et al.³ and Our world in data⁴).

Figure 2. COVID-19 deaths per million population for March–July 2020 (Our world in data⁴).

Approximately 2.3 million people live in Stockholm, which has a tax-funded mixed public and private health and social care system. Approximately 60% of primary care facilities are private, as well as one acute hospital and some geriatric hospitals. Many non-medical but health-related services are provided or purchased by the 26 municipalities in the region, including non-medical services for older people, such as private home visiting services and care homes for ~15,000 people.

Stockholm is unique in that all public non-acute health-care services are owned by the region and managed within the ‘Stockholm health-care’ organisation.⁵ This primary and community health-care organisation includes a diverse range of service delivery units and medical home care services, including specialist ‘hospital at home’ and palliative care; 70 primary care centres; services for psychiatry, addiction, rehabilitation; and some community-based geriatric hospitals.

This article describes the emergency response of the public primary and community health-care division within the Region Stockholm health system and lessons from this response for the next months and for future outbreaks. The basis for the description is the experience of two members of the emergency management team (MO and UL) and a related rapid implementation research project (principle investigator JØ). The latter was started in mid-March 2020 by a combined team of Karolinska Institutet medical university researchers and researchers and administrators employed by Region Stockholm who worked in the SLSO primary and community health-care organisation.⁶ To document and evaluate the emergency response, we used data from a weekly survey of all management heads of SLSO primary and community health-care services (n = 110); minutes of management meetings; interviews with the emergency management team; and documentation from the emergency management system and emergency planning documents.⁷

Staff practising at three levels of the SLSO organisation had different responses to the emergency over the period from March to July 2020. The three levels were: SLSO senior management and headquarters support staff; heads of service delivery units; and clinical and other staff serving patients in these units. The response is summarised in Table 1.

Table 1. Timeline of main events and responses by SLSO public primary and community health-care organisation

The emergency management team for SLSO was formed and met twice daily from the 1 March 2020 to perform the 10 functions of the NATO emergency management model.^7,8 The main objectives of the emergency system were to establish and operate an emergency management organisation structure and systems; communicate information required by staff and patients (internal and external); and mitigate and stop the spread of infection.

In the first 2 weeks of March, we documented rapid and important changes to SLSO operations at each of the three levels of services. Each change would previously have taken months or years to make. Some of these changes were common elsewhere, such as rapid scale-up of telephone and video consultations, and the provision of e-health information to the public and staff. Two strategies were used to protect staff health and reduce the risks of infection spread. Staff thought to be at high risk of serious illness if infected were deployed to non-patient contact and lower-risk work such as digital service responses. Other staff were encouraged to report sick if symptomatic and did not need certification for payment. The human resources department was able to work constructively with staff unions and established arrangements that made it possible to redeploy staff to services where they were most needed. Certain primary care units were designated as ‘node units’, dedicated to receive patients with suspected COVID-19 infection and not in need of immediate hospitalisation. At this time, these units provided clinical examinations and advice to patients, but not tests. In addition, an information system was established to provide the emergency management team with timely information, as described later in this article.

In March, shortages of personal protective equipment (PPE) in all patient contact services was reported in the weekly surveys of unit managers. Local links and flexibility made it possible for SLSO and Karolinska hospital to set up a ‘command centre’ for purchasing and logistics that included Scania construction company staff, using their expertise and systems to source and purchase materials internationally and distribute them to units.⁹ Part of this response was to establish a system to assess the material needed in all 700 units and establish a central store at the SLSO offices of protection material and distribute this quickly when needed.

PPE shortage was experienced in private long-term care homes contracted by the municipalities. This contributed to outbreaks in infection and mortality in some homes. Although it was not the responsibility of SLSO to do so, the above logistics centre distributed PPE to homes in need and formed five mobile teams to reinforce testing capacity in long-term care homes and our geriatric units.

In May, the weekly unit managers’ surveys indicated that staff became more practiced and less anxious and the systems were revised to become more efficient. To coordinate all private services and municipality governments, three area organisations were established. These three high-level groups were headed by SLSO and municipality directors covering social care and combined the head municipality and private health service leaders.

The disease activity curves began to level-off in May, and in June, the twice-daily meetings of the emergency group were reduced to once a day and then three a week. After 10 July the emergency management system was dis-established, and ‘normal operations’ declared on 10 July. Surveys and reports from staff in the first COVID-19 months show that, although anxiety was high, there was also a strong esprit de corps and mutual support, as well as the enjoyment of being free to focus on medical practice with less ‘form filling’. However, by May and June, there was evidence of the effects on some staff of the high and continual workload in an uncertain situation, that was compounded by staff sickness.

From data about service utilisation and financial payments to services, we observed high levels of staff sickness leave, yet for many services, similar levels of utilisation and service production were observed compared to the same period last year in this multi-service organisation. Possible explanations are that remaining staff worked harder or more efficiently and provided services in response to similar levels of demand to 2019. Others have reported significantly lower levels of health-care utilisation, but we did not observe this in all services from the available data.¹⁰ The exceptions were in primary care services where data, from March to June 2020 compared to 2019, showed that there was an overall reduction of 24% in physician visits, 22% in nurse visits and 25% in rehabilitation, including telehealth visits.¹¹

During July and August, the services catered for localised outbreaks and a growing demand from patients whose health may have deteriorated, but who may not have contacted services. As yet, there is no firm evidence of the latter demand. There are different views about a possible increase in mental health issues, as well as problems arising from health conditions that may have deteriorated due to patients avoiding care, or experiencing challenges in performing self-care and maintaining a healthy lifestyle. Two groups that are of concern are older patients living alone and family carers supporting people with health conditions such as dementia and psychiatric illness. SLSO continued to provide specialised support to patients at home with compromised immune systems and to some patients who experienced periods of serious illness. COVID-19 care is being revised for what staff are learning to be a multi-organ disease that can have variable and serious long-term effects.

Preparations are now being made for potential localised outbreaks, as well as more widespread resurgence in disease activity at any time; these may be more likely as the winter proceeds. Three possible scenarios are: (1) no change to usual demand and use of health care; (2) significant changes in demand or use from different patient groups; and (3) proactive and preventive targeted outreach to individuals or specific groups who have not presented but whose health is likely to deteriorate and are likely to need care (known patients, groups or areas at risk). These are informed by the three scenarios for the winter made by the Public Health Agency of Sweden.¹²

An inquiry is underway into the tragedy of infection and COVID-19 deaths among residents of specific care homes, and we are also aware of possible infection of staff employed by services who visit older people in their homes. Our initial survey shows no difference in mortality rates between private and public long-term care homes. There are indications, reported elsewhere, of an association between higher rates of infected cases and the size of unit in terms of number of residents.¹³

During the 2020 summer break, we are considering how to prevent outbreaks in care homes. One issue to be addressed is clarification and implementation of responsibility for preventing infection. Municipalities contract private care homes and home-visiting social care services. The health-care system contracts private general medical practice staff to serve residents. In addition, funding and support responsibilities will be reviewed, given the high human and economic costs of infection prevention failure. Related to this are methods to identify homes that might be most at risk of infection spread, both when making and reviewing contracts, and then routinely. This will be done using formal risk assessment methods¹⁴ and by seeking and using local reports and knowledge of relatives, professionals visiting the homes and care staff. It may be possible to develop systematic and effective ways to seek and act on this knowledge. The media have demonstrated that this can be done.

Outbreaks in nursing homes affected hospitals, often because PPE was not available, or infection control was poor. The emergency highlighted how actions or inactions by one sector or service affected others, often later in time and in unpredictable ways. Rapid transition to telemedicine by primary and community health services reduced emergency department attendances and possible infection of uninfected people.

A related systems observation are comments made by some management staff in interviews; they had noted theirs and others’ limited knowledge of the role played by different functions and services and the consequences of not organising systems as a whole. Initially, it was not clear to some managers in the emergency management team why some functions in the NATO model were needed. They reported that they learned how necessary all the functions were at the decision table, as well as the need to bring expertise, such as from pharmacists, into meetings. Implications for the future include the need to develop a systems perspective among staff. It is also necessary to ensure that there are people from different parts of the system with different perspectives at decision-making meetings, as well as effective decision processes to integrate perspectives, and a method for later reviews for fast decisions that have to be made where there is uncertainty about the effect of these decisions on different parts of the system.

A large number of staff employed by one organisation in many different services can quickly be redeployed and retrained as necessary, to balance staff availability to the rapidly changing needs of services and populations. We documented effective redeployment of high-risk staff out of harm’s way; for example, to telemedicine, but in some cases, redeployment of staff in low-workload services to areas with high workload was not achieved. This only became clear from the retrospective analysis of production and reporting data in July 2020.¹¹ One lesson is that more timely data to show comparative high and low workloads in different services are necessary to take better advantage of the benefit of a large single multi-service employer in an emergency.

During the summer transition, service staff had more time to carry out scheduled follow ups for patients with chronic and other conditions and identify patients at risk of health deterioration who had not been in contact. It is uncertain whether there will be a rise in demand for health care from patients whose physical or mental health conditions have deteriorated or what the possible health consequences of unemployment and stress in vulnerable populations will be.¹⁰ We learned to regard primary and community health care as the ‘front line’, with acute hospitals as ‘back-up’, but to be avoided if possible. Winter preparations include planning to target specific groups and services for proactive action to prevent health deterioration. This includes regular updates from research elsewhere about such demand so as to guide targeting and action. If planning is effective, the load on acute hospitals can be minimised to retain their capacity to respond to a winter COVID-19 resurgence, if this occurs.

The value of timely information to management about service use, infection rates and valid predictive models was demonstrated early, as were the limitations of national level data. Experts and analytical members of the SLSO emergency management team found a way to link local databases and present analyses regularly to the team, to enable informed decision-making. This continues with revisions to help winter deisions and developing a more robust, more accurate and efficient data analytics system for this and other health-care emergencies. Related to this was the rapid implementation research project involving a team of university researchers and SLSO staff. The researchers were able to bring relevant knowledge from different sciences to design the weekly surveys and provide knowledge support about management, organisation and implementation to SLSO. One lesson for the winter and for others is of the value of engaged rapid partnership research to provide real-time information to management.

This article drew on research by a joint university and health-care system team and thus may be biased towards positive findings about the health-care system investigated. However, there were advantages of a joint team using rapid research methods and building on existing partnerships from other ongoing research. One was rapid access to data and our ability to quickly document and make assessments. The primary conclusion was that the pandemic highlighted the strengths of the health-care system and of its staff, but also its vulnerabilities. Managers reported they were aware of many of the system weaknesses highlighted by the emergency. The likely winter demands on the system and the need for a more effective response mean careful preparations are now being made, informed by the lessons of the last months. The pandemic showed clearly the service interdependencies, but also the absence of timely reliable information and previous effective coordination arrangements between public and private services and the municipalities providing social care.

The COVID-19 crisis brought people together, put aside squabbles about budgets and responsibilities and resulted in fast and effective action. The sense of professional vocation and of common purpose will be needed in the coming months. People who came together and were tested carry this experience and the benefits they saw into the coming months, but not all will be in post or present for the next outbreak. We saw that effort and rising to the occasion is not enough and people become exhausted over time. We have learned we need to use this experience to build systems and to revise and establish the organisation and management needed to support clinicians and patients, both for emergencies and for improvements to routine care.

Research team members were employed by Stockholm Region Government and by the Karolinska Institutet medical university, and conducted the research in addition to their other funded projects and works assignments. No separate funding was provided to researchers and health service staff for this project. One research member was retired and not employed by either organisation. Possible bias could result as many project members were employed by Stockholm Region Government and were investigating how their employers implemented emergency management.

Etikprövningsmyndigheten BESLUT 2020–04–08 Document ID: B42F283821A94986986F74E4F866C7FD Title: ‘Omställning av närsjukvården som svar på covid-19-utbrottet: en studie av krisorganisationen i Stockholms läns sjukvårdsområd’ (‘Implementation of management and organisation response to the COVID-19 outbreak: a study of the crisis organization in Stockholm County's healthcare area’). Sökande forskningshuvudman Region Stockholm, Forskare som genomför projektet John Övretveit.