Introduction

Mass casualty incidents (MCI) and other major incidents (MI) occur repeatedly and demand special healthcare system preparedness. Worldwide, there are global trends of increased MCIs following terrorism, changing security situation and natural disasters [1]. To effectively evaluate preparedness before an MI occurs is essential to achieve optimal resilience. Simulation exercises and repeated training are important factors for success.

Swedish healthcare has historically been pioneers in disaster medicine, but in recent MIs, insufficient preparedness and need of improvement has been seen [2,3,4]. In the past 15 years, political decisions have resulted in reduced resources for preparedness [5]. Recent global challenges such as the pandemic, terrorism and the invasion of Ukraine has increased focus on these issues, and initiatives to improve preparedness has gained urgency. In Sweden, the Health and Medical Services Act (HSL) [6] and the National Board of Health and Welfare (NBHW) regulate preparedness work through guidelines and legislation [7]. For main laws and guidelines, see Fig. 1.

Fig. 1
figure 1

Swedish laws and guidelines on disaster medicine, main paragraphs. Freely translated

Full size image

An investigation after the terror attack in Stockholm 2017 stated that the medical effort generally worked well but emphasizes great needs for review of contingency plans, communication systems and routines for calling in staff [8]. These factors are some of the cornerstones for disaster preparedness according to previous studies [1, 8,9,10]. The WHO's checklist for all-hazards approach in disaster preparedness includes, among others, surge capacity, triage and supply management [9,10,11]. Contingency plans including different types of disasters, such as epidemics and network disruptions, are recommended [11, 12]. It is also well-established that well-functioning contingency plans must be accessible, concise and well known amongst concerned staff [1, 8].

Each of Sweden’s 21 healthcare regions should have a regional contingency plan, and each hospital a local contingency plan, according to the HSL and the NBHW’s guidelines [6, 7, 15, 16]. How the plans are organized, practiced and known amongst staff is not known. A systematic method for regular control and follow-up on how hospital preparedness regulations and guidelines are implemented and followed is needed.

Aims

To identify areas in need of improvement in preparedness for incidents and disasters at Swedish emergency hospitals, with focus on mass casualty incidents, using a web-based survey, and to investigate how well current legislation and guidelines on preparedness are complied with.

Materials and methods

Study design

A descriptive cross-sectional study was conducted through a web-based survey. 87 hospitals were invited to participate with primary focus on participation from the 49 Swedish emergency hospitals (i.e. hospitals with a full-scale emergency department receiving both medical and trauma patients) [13]. Non-emergency hospitals (hospitals with an emergency department with for instance limited opening hours, only for medical patients, etc.) can also be involved in MIs, especially in the many rural parts of Sweden where distance to an emergency hospital is large. It was, therefore, of interest to investigate preparedness on some of these.

The survey (appendix 1) was designed based on aspects identified to be of greatest importance for adequate preparedness, including the WHO's checklist [12]. It consisted of 67 questions formulated in general terms covering following sections:

  • Background information

  • Contingency plans

  • Major incident and disaster exercises

  • Hospital Command Group, communication and triage

  • Surge capacity and regional coordination

  • Real-life events

  • Supply management, mobilization at the hospital, review of key areas

The survey was reviewed and deemed suitable for the aim with regards to content by Sten Lennquist, professor emeritus in Disaster Medicine. Due to confidentiality reasons, several surveyed hospitals did not wish to provide a copy of the contingency plan and comparison between provided answers and actual contingency plan was therefore not conducted.

The survey was conducted through a software program well suited for the purpose and was open between September 6th and November 1st, 2021. Respondents could pause the survey and, if so whished, choose not to disclose information requested. An initial reminder was sent after 3.5 weeks, after which weekly reminders were sent until last response date (total 4 reminders).

Data analysis

Analysed data were mainly descriptive and extracted to Microsoft Excel (version 16.43, 2020) for analysis. For descriptive data, averages, medians and percentages were used. For continuous data, standard deviation was calculated. This was performed both for collected data and divided according to hospital types. Qualitative data were analysed through content analysis where main themes for each question were identified.

Ethical considerations

The nature of the study entails a certain risk of accumulation of potentially sensitive data that could constitute a security threat to individual hospitals, regions or Swedish healthcare in general. To reduce the risk, questions asked were kept at a general level, minimizing the amount of sensitive data. Data were handled with confidentiality according to regulations at Karolinska Institute. Results were reported so that individual hospitals or regions were not disclosed. The benefit of an overall study of Swedish disaster preparedness was assessed greater than potential security risks. As the study was conducted at the organizational level, there is no risk to autonomy or integrity of individuals. The study was approved by the Swedish Ethical Review Authority, no. 2021–02865.

Results

39 out of a total of 87 invited hospitals answered the survey (45%). 34 of responding hospitals were from the 49 emergency hospitals, giving a response rate of 69.4% among these hospitals. The other five were non-emergency hospitals that either had an emergency room with limited opening hours or a 24-h local emergency unit. However, some of the respondents indicated that more than one hospital was part in the contingency organization they represented, leading to a higher actual number of respondents than the factual one. For simplicity and security in the statistics, the numbers presented below are 39 answers of the survey that has been provided, although the results can be representative for more hospitals than the 39 answered surveys. 7 hospitals chose not to participate due to confidentiality. 18/21 (86%) regions were represented, with a good geographical spread. All hospital types were represented; university hospitals in the form of regional trauma centres (RTC), non-trauma centre university hospitals (NTCUH), and smaller hospitals including county hospitals (CH), county district hospitals (CDH) and local hospitals (LH) (Table 1). The majority had a trauma admission area of 100,000–500,000 people. The five non-emergency hospitals were LH or CDH.

Table 1 Number of respondents per hospital type
Full size table

A clear majority (34/39) of the hospitals’ respondents were contingency coordinators (CC). One response per hospital was requested. Three hospitals answered twice, in which cases only the response from the CC was analysed as the survey was mainly aimed at CCs. Results are presented by the surveys main areas.

Information about the hospital and key coordinating functions

All but one hospital had a nurse as CC. Two hospitals (1 CH, 1 CDH) had no function responsible for disaster preparedness. At UH, the CC in 4/6 hospitals (67%) had basic nurse training. At CH and CDH, 15/31 of CCs were specialist nurses, most commonly specialized in anaesthesia (n = 8), intensive care (n = 2), or emergency/prehospital care (n = 6). Regarding the CCs formal training in disaster medicine, higher levels were seen at smaller hospitals (CH, CDH, LH) compared with all UH. Only one UH (17%) had a CC with formal education, whereas for CH it was 8/15 (53%), CDH 9/14 (64%) and LH 2/2 (100%) (Fig. 2).

Fig. 2
figure 2

Requirements for formal education in disaster medicine for hospital key functions. The figure regards to the questions “What type of education in disaster medicine does the hospitals contingency coordinator/physician have?” and “Do you have any requirements for a formal education in disaster medicine for those in your disaster committee/Hospital Command Group?”. n = 39

Full size image

4/6 (67%) of UH had a contingency physician. In smaller hospitals, such function was significantly less common; 6/16 (37%) and 3/15 (20%) respectively at CH and CDH. The proportion requiring formal training in disaster medicine for contingency physicians was higher at CH and CDH compared with UH (3/4 (75%) of RTCs "do not know", 1/4 (25%) stated "no formal education". 3/6 (50%) of CH, 3/3 (100%) of CDH had educational requirements (Fig. 2)).

A “Disaster Committee” for preparedness was present at all UH and most of CH and CDH (21/31, 68%). No UH required formal training to join the committee while 5/10 (50%) CH and 4/11 (36%) CDH required this (Fig. 2). One CH did not have either a contingency coordinator, contingency physician or disaster committee.

Contingency plans

All hospitals except one CDH had a hospital-wide contingency plan. In addition, 5/6 (83%) UH had synchronized department-specific plans, for CH and CDH 8/16 (50%) and 5/15 (33%) had this. Great variation in plan length was seen (median 34 pages, min–max; 13–80). LH generally had shorter contingency plans (median 17.5). The majority updated their contingency plan in the past two years, except 5 CH and 6 CDH. All but three had updated the MCI plan 2018 and onwards. Among those who had updated all parts of the plan, CH and CDH dominated; 11/16 (69%) and 11/15 (73%) respectively compared with UH (1/6, 17%).

Large differences were seen in described structure of contingency plans, complicating structure analysis. 11/31 CH and CDH only use the regional contingency plan, in 7/11 cases with local action cards. Regarding which employees read the contingency plan, an even distribution was seen between "new employees", "up to each employee" and "according to guideline". Seven hospitals wrote that the responsibility had been delegated to the units and four hospitals that “The guideline is that all new employees should read it. This is probably not the case”.

Regarding separate plans for different types of MIs, most notably was, that not all hospitals had a plan for network disruptions. All participating UH, but only 13/16 (81%) of CH and 14/15 (93%) of CDH had this. In general, UH used action cards for more different functions. One CH did not use action cards at all.

Major incident and disaster exercises

One CH and two CDH stated that they never conduct disaster drills. Among others, the just over half of the respondents said either that all trauma-receiving employees or a few per year according to a rolling schedule participate in disaster drills. Five hospitals stated that only HCG practices. The 26/39 (67%) practice different scenarios. One CDH did not evaluate exercises. Most evaluate either through review in the disaster committee (8/39, 20%) or through evaluation of staff knowledge development (11/39, 28%). A clear majority (30/39, 77%) used the results to update and improve contingency plans.

Hospital Command Group

All UH and 25/31 (81%) of CH and CDH had a multidisciplinary and trained HCG. At four CH and CDH, HCG consisted of regular managers or regional management. The included functions were fairly even at different hospitals, apart from nursing professions, where no RTC had such representation (for others; 33–100%, 50% median). Three CH/CDH stated that it varied depending on the incident. HCG at all but one CH and one CDH worked according to a special structure, most often the NATO model for staff methodology. All respondents said that HCG is trained, 20/39 (51%) trained MCI 2019. Regarding requirements for training in disaster medicine, 4/6 (67%) of UH, 11/16 (69%) of CH, 12/15 (80%) CDH and 2/2 (100%) of LH had such requirements (Fig. 2).

Communication

Regarding disaster communication systems, the 21/39 (54%) planned to use regular systems with pagers, mobile phones and landlines, in several cases with addition of radio communication. 3/6 (50%) of UH, 9/16 (56%) of CH, 8/15 (53%) of CDH and 2/2 (100%) of LH had a tested backup system for communication. One CH had no backup system at all. 30/39 (77%) had a designated person responsible for updating hospital staff contact lists. 2/6 (33%) of UH, 3/16 (19%) of CH and 1/15 (7%) of CDH had no system to ensure updated staff contact lists.

Triage

All hospitals had a plan for where primary triage in an MI would take place. For secondary triage, one quarter stated that trauma teams decide where it would take place, the rest had specified the location in advance. Regarding triage method, large variations were seen (Table 2). Two hospitals responded that they did not know which method was used.

Table 2 Triage methods in the event of a disaster
Full size table

Surge capacity and regional coordination

3/4 (75%) RTC had evaluated the hospital surge capacity. For CH, this was 6/16 (38%) and for CDH 3/15 (20%), both NTCUH replied "do not know". No surge capacity evaluation had been done at 7/16 (44%) of CH and 9/15 (60%) of CDH. Two CH described difficulties evaluating capacity due to large variations in available resources, one of which said that capacity revaluation is done continuously due to "significantly reduced margins". Three hospitals (2 RTC, 1 CH) had evaluated surge capacity in connection with a disaster drill, one hospital in connection with the COVID-19 pandemic and one said that it “occurs at every specific incident”. Note that not all hospitals specified mode of surge capacity evaluation.

Regarding regional coordination, 1/4 (25%) RTCs had a plan for when and how patients will be sent to other hospitals, while this for others was 1/2 (50%) NTCUH, 10/16 (62%) CH, 8/15 (54%) CDH and 1/2 (50%) LH. 10 hospitals had no plan on how to send patients. 3 hospitals stated that they do not have disaster drills alongside the region. Among others, 4 answered "do not know”. 28 hospitals stated that exercises with the region had been carried out.

Real-life events

Last time the contingency plan was activated for MCI was for UH in half of the cases 2–5 years ago, while a majority of CH and CDH activated the plan more than 10 years ago (17/31, 55%), but spread was large. Just over half of CDH had never activated the disaster plan for MCI.

Supply management, mobilization at the hospital, view on key areas

In this section, many chose "cannot answer". Response base is thus limited. Of those responding, 9 CH, 11 CDH and 1 LH (21/39 in total) had no system for controlling material supply, 3/39 hospitals (1 CH, 1 CDH, 1 LH) had such system. 16 hospitals had disaster stores at the hospital, 12/39 lacked such stores (Table 3). Nine hospitals updated the disaster store annually and two less than every two years.

Table 3 Does your hospital keep emergency storage at the hospital in case of a disaster?
Full size table

Among RTC, it varied greatly whether they had a system to increase capacity for patient beds or not; 1/4 (25%) had such plan, 1/4 did not, 1/4 did not know and 1/4 chose not to answer due to confidentiality. Among others, a majority had a system to increase capacity; 2/2 (100%) NTCUH, 13/16 (81%) CH, 8/15 (53%) CDH and 1/2 (50%) LH. Regarding how to distribute patients in an MI, 16/39 (41%) stated that they would be placed in the same ward and a relatively large proportion of mainly smaller hospitals that the patients should be placed in different wards (13/39, 33%; for smaller hospitals 12/31, 39%).

Respondents were asked to rate how important specific areas were considered to be for improving the hospitals disaster preparedness, on a scale of 1–10, 10 being highest. In all areas, respondents for RTC chose a lower rating compared to other hospital types (in terms of mean and median). International cooperation was ranked lowest (mean 6.47 from all respondents) followed by increased military cooperation (mean 8.35). Most important were exercise and simulation (average 9.64), and disaster management and leadership (average 9.06).

Finally, respondents were asked to openly describe what they found important for increasing hospital disaster preparedness. Three themes recurred among all hospital types; increased national collaboration and consistency of methods and contingency plans, national guidelines, and increased training in disaster medicine.

Discussion

Emergency hospitals play a key role in society's ability to handle major incidents such as mass casualty incidents and must be prepared for such crises. Main findings of this survey study include several possible areas for improvement in disaster preparedness at Swedish hospitals. All-though response rate was lower than hoped, most of the emergency hospitals were represented (69.4%), the largest hospital in most regions participated, and almost all regions were represented. General conclusions can thereby be drawn. Legislation and guidelines were largely complied with and results pointed to major differences in how preparedness work is conducted, for example regarding contingency plans, exercises, and triage methods. A clear majority of hospitals cover most key areas described by the WHO [11], but no hospital covers everything.

Results indicated that this method could possibly be used as a quality indicator to monitor that hospitals maintain similar and adequate levels of preparedness. However, the survey was only tested once in this study and should undergo further validation and testing before clear conclusions on this can be drawn. In combination with specified requirements on what functions and levels of education should be available for each hospital type, a survey like the presented, addressed by accredited organizations or authorities, could make hospitals and regions uniformly prepared and contribute to making hospitals better equipped for MIs. If results are openly presented and carried out at regular intervals, it could function as a trigger for improvements [1, 11].

Large differences in the hospitals’ contingency plans were seen, even though this is known to worsen conditions for coordinating efforts in MIs [1]. In this study, implementation of contingency plans among health-care staff also varied. The fact that contingency plans at several hospitals was very long increases the risk of staff not reading it [1, 11]. The importance of increased coherence and collaboration is emphasized in recently completed studies on Swedish health care's disaster preparedness [14].

Over the past 20 years, the number of hospital beds in Sweden has gradually decreased to the lowest level in the OECD [15]. In the same period, occupancy rate has increased. As this is strongly linked to hospital capacity [1], trends of increased overcrowding can be seen as a sign that disaster preparedness has deteriorated. Reduced number of hospital beds and ventilators combined with a lack of material supply also constitutes three of five main limiting factors for hospitals ability to handle MCIs found in a study from 2020 [15] of Swedish surgical surge capacity, where large regional differences were seen [15]. In March 2020, the Canadian Association of Emergency Physicians (CAEP) stated, that Canadian healthcare, contrary to the opinion of many authorities, was dangerously unprepared for disasters, and active efforts were required to fill identified gaps [16]. In line with Blimark et al. [15], this study has indicated a need for increased hospital surge capacity. The trend of declining numbers of hospital beds must be reversed to create conditions for improved disaster preparedness. In combination with increasingly subspecialized and fragmented healthcare, which has a negative impact on trauma care [16,17,18], reduced margins could be seen as one of the biggest challenges in maintaining sufficient capacity in Swedish hospitals.

According to the Health and Medical Services Act [6], a function or group at hospital level responsible for preparedness is recommended to best mobilize resources [1, 19]. A clear majority of those answering the survey have coordinating functions, but the fact that the figure does not reach 100% leaves room for improvement. Without a function responsible, there is a risk that review of contingency plans become less frequent and less structured, exercises fewer and general interest in disaster preparedness at the hospital risks diminishing [9, 11, 12].

Specific training requirements for coordinating functions varied greatly between hospital types. In Sweden, disaster medicine is no longer part of basic medical education for future doctors and is limited to 1–2 weeks during basic education for nurses. Formal education in disaster medicine is conducted in the form of separate courses through different universities, organisations (for example the MRMI courses) and the Centre for Disaster Medicine and Traumatology in Linköping, Sweden. Smaller hospitals generally place higher demands for formal education in disaster medicine for coordinating functions compared with university hospitals, which was surprising. Larger hospitals may generally have more experience in trauma care and could consider themselves to have sufficient practical experience. However, experience from trauma in everyday life does not reflect on ability to adapt to special working methods required in an MCI [1, 15]. In addition, trauma patient volume at Scandinavian hospitals has been shown to be insufficient for maintaining competence [17]. Formal education in disaster medicine could work as a quality assurance for competence and should be a basic requirement for those responsible for the hospitals’ disaster preparedness.

Most hospitals have a multidisciplinary Hospital Command Group (HCG), but several smaller hospitals only have regional management or decide who should be included in connection with the major incident. This despite the fact that contingency plans, according to national guidelines, clearly must state who has the initial management responsibility in an MI [7] (Fig. 1). The fact that HCG has not been defined in advance poses a great risk that establishment of management and thus the entire effort is delayed. Different working methods in hospitals’ HCG can lead to difficulties with coordination between regions. Several hospitals do not require leadership training for those in the HCG, which goes against research and recommendations [1]. Improved management structure was the key area for improving preparedness ranked second highest by respondents.

Uncertainty in surge capacity risks leading to hospital overloading in an MCI. Overloading could cause normal quality requirements not being maintained despite adequate measures [1]. This affects patients through reduced quality of care despite sufficient capacity being available at other, less affected, hospitals in the immediate vicinity. Distributing patients in a disaster is a major logistical challenge, but knowledge on capacity and limit for each hospital can facilitate the task and be seen as a prerequisite. The fact that many hospitals lacked a plan for when and how patients are to be transported to different hospitals complicates the task even more. This study does not test capacity, but results indicate that capacity constraints highlighted by the NBHW [20] 4 years ago have not sufficiently been taken into account in the hospitals' disaster planning.

Regarding supply management in an MI, areas for improvement are seen at most hospitals. Many regions worked according to the "just-in-time" principle, which for instance led to lack of personal protective equipment when demand increased at the start of the COVID-19 pandemic 2020 [5]. In a sudden MI such as an MCI, there is a risk of rapid material consumption, and it is reasonable that each hospital has stockpile to manage for the first day. In addition, more hospitals should include material supply in surge capacity evaluations and contingency plans.

Triage methods used in a disaster differ greatly between hospitals and regions. Several planned on using regular triage methods, i.e. methods not adapted to disasters such as the RETTS system. The use of many different methods also poses a risk of miscommunication as patients are sent between hospitals and regions. A common triage method intended for disasters, for instance, physiological triage systems such as triage sieve and triage sort as initial triage followed by an anatomical triage, should be implemented at all hospitals in Sweden.

Swedish hospitals have been relatively spared from MCIs, making exercises essential. [1, 20]. Several hospitals rarely had drills, and two did not train at all. This highlights an obvious gap in disaster preparedness. Among RTCs, frequency of exercises was lower than expected. Contingency plans thereby risk being insufficiently tested. Exercise and simulation also create the opportunity to evaluate hospital surge capacity and has in a study [21] after the terrorist attacks in Manchester 2017, been shown valuable for staff confidence in decision-making in MIs, which is in line with several other studies on the subject [1, 8, 10, 21,22,23,24,25].

In an MCI, there are great risks that communication systems are overloaded [1, 19]. That a large proportion of hospitals plan on using regular systems in an MI creates an obvious risk of communication failure, which happened during the terrorist attack in Stockholm 2017 [8]. Communication failure could delay and impair care and is often an overlooked area [19]. Alternative communication systems in the event of disturbances must be in place. Backup systems for communication were also found in a study from 2020 [19] to be lacking in many trauma centres in Canada, Australia, New Zealand and the United Kingdom. Gabbe et al. [19] conducted a questionnaire study on all trauma centres in the above-mentioned countries, and their results are in line with those obtained through this study. The areas of improvement identified thus do not appear to be unique to Sweden but are found in several countries with similar healthcare systems.

The respondents ranked international and military cooperation to be of lowest value for improving hospital disaster preparedness. This may have since changed, as Sweden now experience a vastly different security situation and level of threat in the immediate area.

The study has a good geographical coverage with 18/21 regions represented, which means that trends seen can be considered to apply to the whole country. Whether the results are transferable to other countries is more uncertain as healthcare systems and guidelines differ. However, the method used could be used for control and follow-up in other contexts, as the areas explored are fundamental to all healthcare systems.

Limitations

The study would clearly benefit from a larger response rate as issues about national representability can be raised. However, as declared above, almost all regions were represented and a good geographical spread was achieved. In addition, the response rate among emergency hospitals, which were the main target of the study, was considerably higher than among all invited hospitals. Surveys carry inherent limitations and risks for misinterpretation of questions. Results should be interpreted with care. There is a risk, that respondents wanted to give a better picture of preparedness at the hospital they represented than the actual, and pressure from hospital leaders to paint a better picture than the real is always a risk. Measures were taken to overcome this by anonymity. Many of the questions are easy to validate. In addition, answers partly reflect a perceived description of reality. As MIs are unusual, it is difficult to test how well the perceived image of hospital preparedness matches reality. To obtain this, realistic simulation exercises and surge capacity tests, are needed. [26, 27] As the survey was aimed at and mainly answered by contingency coordinators, the prerequisite for correct answers has been optimized. Future studies validating this method using this or similar surveys, are needed.

Conclusions

Swedish hospitals' disaster preparedness generally covers many key areas, but no responding hospital seems to cover all. Several areas for improvement were found, especially within implementation and overall extent of contingency plans. Frequency of exercises differed greatly and almost half of hospitals did not perform surge capacity evaluation. University hospitals have lower requirements for formal education in disaster medicine for key functions. Large differences between hospitals identified in this study offers difficulty in achieving national equality regarding preparedness. Several local contingency coordinators requested national consistency and guidelines for contingency plans. Future in-depth studies and simulation exercises using existing resources are needed to test several of specific areas discussed. The survey used in this study was found useful as an overview of hospital disaster preparedness and should be further evaluated and conducted at regular intervals.