Resilience

The resilience of a health system can be defined as its ability to absorb the impact of brutal disruptions and “bounce back”, i.e. to adapt to them in a way that allows it to keep providing necessary services at an acceptable level, to resume optimal performance as quickly as possible, to strengthen its structure and functioning and, potentially, to become less vulnerable to similar disruptions in the future.

Health system resilience is a relatively new concept which had already been gathering increasing interest for a number of years before the COVID-19 pandemic brutally highlighted its crucial importance for countries everywhere in the world.

This major health crisis provided researchers with a (hopefully) unique opportunity to assess the resilience of the Belgian health system through a number of indicators reflecting the way it absorbed and responded to the COVID-19 crisis.

The indicators in this section relate to three main aspects of resilience:

• The ability to ensure and maintain an adequate workforce, in a context where healthcare workers were exposed to an increased workload, a high risk of getting sick themselves and considerable psychological strain (R-1, R-2, R-3, R-12);
• The ability to maintain essential health services and routine public health services, in a context where a number of hospital and other health-related activities had to be put on hold either to free up equipment and staff for the treatment of COVID-19 patients or to limit contacts and potential exposure to the virus for patients and professionals (R-4 and R-5);
• The ability to scale up existing capacity (e.g. in intensive care units which had to deal with increased demand due to the pandemic) and implement new health services like telemedicine services, contact tracing, large-scale testing and large-scale vaccination once the necessary vaccines became available (R-6 to R-11).

Indicators on health system resilience

ID	Indicator	Score	Period
R-1 NEW	Healthcare professionals who consider leaving the profession (% of respondents, Power to Care)		04/2020 - 09/2021
R-2 NEW	Share of hospital beds that are closed due to absence of staff or force majeure		22/11/2021 - 31/12/2022
R-3	Number of nursing vacancies in hospitals		31/12/2019 - 31/12/2021
R-4 NEW	Number of essential surgical hospital acts		03/2020 - 01/2023
R-5 NEW	Number of new invasive cancer diagnoses		01/2020 - 12/2021
R-6 NEW	Hospitals with occupancy rate for COVID-19 patients in ICU licensed beds above 60% (% of general hospitals with ICU licensed beds)		20/03/2020 - 31/12/2022
R-7 NEW	Number of contacts (including teleconsultations) with a GP (as a percentage of the total number of contacts with a GP in the pre-COVID period)		01/2020 - 12/2021
R-8 NEW	Average duration between COVID-19 sampling and test result (days)		31/08/2020 - 21/02/2022
R-9 NEW	Average duration between positive COVID‑19 test result and contact tracing initiation (days)		31/08/2020 - 11/10/2021
R-10 NEW	COVID vaccination in the last six months (at least one dose, % of the population)		28/12/2020 - 19/11/2022
R-11 NEW	COVID vaccination in the last six months (at least one dose, % of the population 65+)		28/12/2020 - 19/11/2022

Link to the full synoptic table and the report

Two more indicators, relating to the perception of the government’s preparedness to tackle the next pandemic (R-13) and to the country’s preparedness to deal with public health risks (R-14), have been assessed in complementary analyses. They can be found below.

Ensuring and maintaining an adequate workforce (R-1, R-2, R-3, R-12)

A major health crisis like the COVID-19 pandemic inevitably puts healthcare professionals under considerable strain and stress because of an increased workload, potentially made worse by a higher-than-normal rate of absenteeism as healthcare workers themselves become physically ill or psychologically drained. To enable them to keep fulfilling their crucial role in the fight against coronavirus and in maintaining an acceptable level of service in other areas, countries everywhere in the world have taken measures to provide them with additional practical, financial and psychological support.

In this section, we attempt to assess how successful Belgium was in keeping its healthcare workers motivated as well as physically and psychologically healthy under those difficult circumstances, through indicators relating to the percentage of healthcare staff who considered leaving their profession (R-1), to the number of hospital beds that had to be closed due to lack of staff or force majeure (R-2) and to the number of vacancies for nurses (R-3). We have also added an indicator about the number of people registered on a “workforce reserve” list (R-12).

Healthcare professionals who consider leaving the profession (% of respondents, Power to Care) (R-1)

This indicator is based on the Power to Care survey, carried out in 8 rounds between April 2020 and September 2021 by Sciensano and LIGB – KU Leuven. Respondents were asked to self-assess their acute and chronic stress symptoms (e.a. anxiety, sleeplessness, difficulty relaxing or concentrating, etc.), and various other personal and work-related issues (depression, loneliness, insecurity...). We have focused primarily on the percentage of respondents who considered leaving the profession (i.e. those who attributed a score of 7 or higher to this item in the survey); other issues have been examined as secondary indicators.

Results

• The percentage of healthcare workers who considered leaving the profession increased from 15.4% in April 2020 to 27.6% in September 2021, with ups and downs corresponding to peaks and dips in the number of COVID-19 cases. It tended to be higher in Wallonia and lower in Flanders, except during the last round of the survey. However, results for Wallonia were only available for the last 4 rounds.
• This percentage peaked at 35.4% in December 2020 among nurses and nursing assistants working in hospitals. It remained very high at 32.5% during the last round of the survey, in September 2021.
• Based on another survey carried out between December 2021 and February 2022, 43.9% of intensive care nursing staff considered leaving their job and 26.5% considered leaving the nursing profession altogether. This last percentage was higher in Wallonia (37.5%) and Brussels (34.9%) than in Flanders (17.4%).
• The prevalence of chronic stress symptoms remained consistently high between April 2020 and September 2021, particularly for tiredness and feeling under pressure. The prevalence of acute stress symptoms, on the other hand, decreased over the course of the pandemic.

Link to the technical sheet and detailed results

Number of hospital bed closures due to staff absenteeism or force majeure (R-2)

Staff shortages are nothing new, particularly when it comes to hospital nurses. However, this issue might have been amplified during the pandemic: hospitals needed their healthcare workers more than ever to deal with the inflow of COVID patients, but were often left even more short-handed because some professionals caught the virus themselves (and became ill or had to self-isolate as a result) or simply couldn’t take the additional stress.

In Belgium, absenteeism of healthcare professionals has been studied in some contexts, but no comprehensive data are available. This is why we chose to focus instead on the number of hospital beds which had to be temporarily closed due to staff absenteeism or for other uncontrollable reasons (“force majeure”), which was registered between November 22, 2021 and December 31, 2022.

Results

• The overall percentage of hospitals beds closed due to staff absenteeism or force majeure ranged between 5.4% and 10.1% over the studied period. At its peak, in December 2021-January 2022, this issue affected about 4000 beds.
• Over most of the studied period, ICU beds were proportionally more affected than others. The percentage of closed ICU beds ranged between 7.1% and 11.0%. In September 2022, over 10% of the licensed ICU beds were closed, vs 8% of hospital beds overall.
• Similar patterns were observed in all Belgian regions for the percentage of closed hospital beds (all types).
• The percentage of closed ICU beds was highest in Brussels (14.1%-22.7% over the studied period) and lowest in Flanders (4.8%-10.5%), except during the summer of 2022. In Wallonia, it ranged between 5.2% and 12.9%.
• There were substantial differences between provinces both for hospitals beds in general and for ICU beds. In the province of Luxembourg, an average of 30.2% of hospital beds and 17.9% of ICU beds were closed over the studied period.

Link to the technical sheet and detailed results

Number of nursing vacancies in hospitals (R-3)

Nurses play a key role in delivering healthcare, both in hospitals and other healthcare institutions and in primary care and home care services. In most countries, they also make up the largest group among all healthcare professionals, which means a shortage of nurses can easily lead to accessibility issues or have a negative impact on quality of care.

Unfortunately, such shortages have long been an issue in most industrialised countries, and the considerable strain the COVID-19 pandemic put on those professionals has not improved matters (see also R-1). In Belgium too, recruitment is becoming increasingly difficult, especially in hospitals.

The number of unfilled vacancies for nurses, which reflects how bad this shortage is and how difficult it is to find suitable candidates, gives us valuable information on the ability of the healthcare system to maintain or quickly resume operations during a crisis (resilience) and to keep running smoothly in the long run (sustainability).

Results

• In December 2021, 2572 nurses vacancies (in FTE) were reported in Belgian hospitals, most of them in Flanders (1293.6 FTE), followed by Wallonia (704.6 FTE) and Brussels (573.9 FTE). Most of them advertised for nurses with a bachelor’s degree (A1).
• The number of nurses vacancies in hospitals sharply increased during the COVID-19 pandemic, with peaks in September 2021 for Wallonia (+100.3% compared to September 2019) and Flanders (+62.4%), and in December 2021 for Brussels (+68.3% compared to December 2019). It should be noted, however, that figures for 2020 are not available and might have been even worse.
• Likewise, more recent data is not (yet) available, which makes it impossible to assess whether nurse vacancies have since returned to their pre-pandemic level.

Link to the technical sheet and detailed results

Number of healthcare professionals registered in a workforce reserve (R-12)

For many countries, a possible shortage of healthcare professionals was a cause for concern throughout the COVID-19 pandemic. One of the strategies put in place to prepare for this potential issue was to set up reserve lists of volunteers with a medical background (e.g. recently retired doctors or nurses, medical students, part-time workers, etc.), who could be called upon to help out should the need arise. In Belgium, such lists were for example created in Flanders, Wallonia, Brussels and the German-speaking community. Volunteers could register through dedicated platforms.

Results

• In total, the various platforms examined for this indicator registered just under 19 000 health professionals, 12 779 in Flanders, 5865 in Wallonia, 133 in Brussels and 37 in the German-speaking Community. Those figures are, however, difficult to compare because data collection was not the same for all platforms (different timeframes, different definitions of included health professionals, etc.). For the same reason, the greatest caution is required to draw general conclusions for Belgium – and all the more so because our data do not include all existing initiatives (e.g. reserve lists of workers from the public sector with a medical background).

Link to the technical sheet and detailed results

Maintaining essential health services and routine public health services (R-4, R-5)

Aside from its impact on the physical and psychological well-being of healthcare professionals, the COVID-19 pandemic also caused significant disruption in routine healthcare services, including in those that could be considered essential. This section assesses the ability of the system to keep providing routine care in times of crisis, through indicators comparing figures for several hospital procedures (R-4) and for new invasive cancer diagnoses (R-5) during the pandemic and just before.

Regular essential surgical hospital activities (R-4)

From mid-March to early May 2020 and during subsequent pandemic peaks, Belgian hospitals were asked to stop or limit elective activities (i.e. all planned consultations, investigations and procedures that could be rescheduled at a later date) in order to free up equipment, nurses and physicians for the treatment of COVID-19 patients. “Essential” activities were still allowed but, as the distinction between essential and non-essential activities was never formally defined, they seem to also have been affected, albeit to a lesser degree.

This indicator examines whether the level of essential regular surgical activities was lower than could be expected based on figures from previous years (2018-2019) (R-4). We also looked at the impact of the pandemic on various other factors like non-essential surgical activities, certain urgent and non-urgent hospital activities, the number of admissions in hospitals, emergency rooms and psychiatric wards (R-4, secondary indicators).

Results

• In April 2020, Belgian hospitals only maintained 56.9% of surgical activities that were considered essential, as well as 29.8% of their mixed activities (essential or not depending on context) and 5.7% of their non-essential activities. Subsequent drops were more limited for all categories.
• A considerable but temporary drop in expenditures was observed in April 2020 compared to a year earlier for a selection of urgent hospital care activities. An even stronger decrease was observed both in April and November 2020 for non-urgent activities.
• Clear disruptions were observed in 2020 both for ER admissions and psychiatric admissions.

Link to the technical sheet and detailed results

Number of new invasive cancer diagnoses (R-5)

The interruption of elective activities during the pandemic also impacted screening activities, including population screening programmes (e.a. for breast, cervix and colorectal cancer). As a result, there were substantially less new cancer diagnoses than expected during the first wave of the pandemic, as more tumours remained undetected. Since treatment is more likely to be successful when cancer is discovered at an early stage, this delay in diagnosis is a cause for concern, which makes it a useful indicator to assess the ability of the system to maintain services that are strictly speaking neither urgent nor essential, but can nonetheless have far-reaching consequences.

Results

• In 2020, the number of new invasive cancer diagnoses in Belgium was 19.3% lower in March, 39.0% lower in April and 21,7% lower in May compared to the 2017-2019 average for the corresponding months. This comes down to 1222, 2181 and 1366 “missing” cancer diagnoses respectively. At its peak, in April, this decrease was clearly stronger in age groups over 40.
• The number of new diagnoses was back to its expected level by June, but a substantial increase (+19.9%) was observed in September 2020 compared to the 2017-2019 average. Figures then remained fairly stable – if slightly higher than expected on average – until December 2021.
• Trends were similar everywhere in Belgium, though decreases tended to be stronger in Brussels, without larger subsequent increases.
• All types of cancer diagnoses followed roughly similar trends, but with variations in magnitude.

Link to the technical sheet and detailed results

Scaling up capacity and implementing new services (R-6 to R-11)

Aside from its potential impact on the healthcare workforce and on regular services, it goes without saying that a major health crisis like a pandemic is also an issue in and of itself, which requires immediate action to contain, monitor, treat and hopefully put a stop to the disease. In the case of COVID-19, this meant scaling up hospital capacity in relevant services to deal with the inflow of severe cases, issuing recommendations to restrict direct interactions and limit contamination, developing and implementing testing and contact tracing strategies and, later on, organising vaccinations on a hitherto unknown scale.

However, the need to achieve these goals also gave a boost to fairly novel approaches which might, otherwise, have taken much longer to be widely adopted, telemedicine being a case in point. This ability to quickly turn to new solutions in response to a new situation or new needs is part of what makes a system resilient.

This section contains indicators relating to the occupation of ICU beds by COVID-19 patients (R-6), to the number contacts with GPs during the pandemic (including teleconsultations) (R-7) and to COVID-19 testing (R-8), contact tracing (R-9) and vaccination (R-10, R-11).

Hospitals with occupancy rate for COVID-19 patients in ICU licensed beds above 60% (% of general hospitals with ICU licensed beds) (R-6)

Intensive care resources faced enormous pressure during the pandemic and, in some places, demand even ended up exceeding the normal number of available ICU beds. In response, many countries took measures to create extra bed capacity, known as “surge capacity”. In Belgium, ICU bed capacity was promptly increased by a whopping 60%.

However, beds are one thing, qualified staff to tend to the patients in those beds is quite another, and it quickly became apparent that most hospitals did not have enough nurses with specific ICU expertise to staff these “surge” beds. Therefore, relying on surge beds to handle an excess of COVID-19 patients (i.e. an occupation rate above 60% of the licensed ICU beds by COVID-19 patients, referred to as “ICU overflow”), might have contributed to in-hospital mortality rates in this population. After a few months, hospitals were therefore encouraged to aim for a better distribution of COVID-19 patients between hospitals (with the help of an overarching coordination centre) instead of using extra ICU bed capacity.

As available data doesn’t allow us to ascertain whether some hospitals were overloaded while others still had free capacity, this indicator focuses on the percentage of hospitals with high occupancy rates by COVID-19 patients in licensed ICU beds (i.e. their “normal” ICU beds, excluding surge capacity). This gives us an idea of the variation between hospitals and provides indirect insights on whether patients were distributed in a way that reduced risks associated with overflow.

Results

• During the first wave of the pandemic, up to 69.7% of licensed ICU beds in Belgium were occupied by COVID-19 patients. This rate reached its maximum of 76.7% during the second wave. It was consistently higher in Brussels than in the other regions, even between waves.
• During the first wave, a maximum of 70.4% of hospitals in Belgium faced an ICU overflow. This percentage reached 80.6% during the second wave. In April 2021, while the overall occupancy rate for COVID 19 patients in licensed ICU beds was still close to 50%, only 30% of hospitals faced an ICU overflow. In December 2021, when the occupancy rate for COVID-19 patients in licensed ICU beds was above 40% at the national level, only 10% of hospitals faced an ICU overflow.
• Brussels had the highest percentage of hospitals faced with ICU overflow during the first wave, Wallonia during the second. In April 2021, while the occupancy rate for COVID-19 patients in licensed ICU beds was still close to 50% overall, only 21.6% of Flemish hospitals faced ICU overflow, vs 41.7% in Wallonia and 54.5% in Brussels. These regional differences disappeared during later waves.
• It is worth noting the occupancy rate of ICU beds available specifically for COVID 19 patients (including surge capacity if applicable) never reached 100%, even though it came close (up to 92.4% in December 2021). This indicates Belgium was able to adapt its COVID-19 ICU capacity to the pandemic waves.

Link to the technical sheet and detailed results

Number of contacts with a GP (including teleconsultations) (R-7)

While consultations through phone, videocall or other remote means of communication were formerly met with limited enthusiasm in Belgium, when the pandemic hit, they suddenly appeared as a welcome solution to a completely new problem: how to maintain access to essential care without putting patients and doctors at risk of contamination? As soon as the first wave of COVID-19 hit, in March 2020, temporary emergency measures were put in place to allow and fully reimburse teleconsultations in primary care in Belgium. As of August 2022, this temporary system was replaced by a permanent one requiring patients to make a small co-payment.

This will and ability to quickly make a new option available to respond to a new need is a valuable indicator of the health system’s resilience. Measuring its uptake and effectiveness as a replacement for in-person consultations was no easy task, however, as the total number of contacts with a GP could also have been affected by the pandemic. We therefore chose to focus on the total number of contacts with GPs (including teleconsultations) during the pandemic as a primary indicator; in our secondary indicators, we also examined the number of in-person contacts during the same period and the share of teleconsultations in the total number of contacts, then compared the number of contacts during and before the pandemic, including or excluding teleconsultations.

Results

• During the first wave of COVID-19, all three Belgian regions experienced a large drop of face-to-face contacts with GPs, which was largely compensated by a rise in teleconsultations.
• A limited disruption was observed in April and May 2020, where the number of contacts with GPs (including teleconsultations) fell to 92% and 85% respectively compared to the corresponding months in 2019. As of June 2020, however, contacts with GPs (including teleconsultations) remained close to or above 2019 levels.
• The evolution of the overall number of contacts with GPs was fairly similar in all regions, but the share of teleconsultations was consistently higher in Wallonia and Brussels than in Flanders.

Link to the technical sheet and detailed results

Average duration between sampling of a COVID-19 test and test result (R-8)

Identifying COVID-19 cases through testing is essential to know when it is necessary to isolate someone and alert their recent contacts to help prevent and control outbreaks, and the testing capacity in Belgium was promptly scaled up during the first few weeks of the pandemic. However, for this strategy to be really effective, it is important to have the results as quickly as possible after the sampling is performed: if they take too long to arrive, potentially infected individuals might not be willing or able to self-isolate “just in case” and their contacts might not be reached before they have infected others.

EU authorities recommend trying to keep the time between the moment the test is requested and the communication of the results under 24 hours. As multiple factors can influence the time it takes to actually perform a test, we have chosen to focus instead on the time between the moment the sampling is performed and the moment the results are communicated, which is a better indicator of the Belgian testing capacity. Data were available as of September 2020.

Results

• The average time between test and result was highest in September-October 2020, peaking at 1.5 days (1.7 in Brussels and Wallonia). At the time, the percentage of people receiving their test result within 24 hours was 50-60%.
• From November 2020 onwards, the average time between sampling and result remained below one day in all three regions, even when a very large number of tests were performed. Almost all test results were communicated within 24 hours.

Link to the technical sheet and detailed results

Average duration between positive COVID-19 test result and contact tracing initiation (R-9)

Along with testing, contact tracing – identifying and notifying all recent contacts of a person with a confirmed infection (the “index patient”) – can be a valuable tool to control the spread of COVID-19… if it can be organised very soon after the index patient has caught the virus (ideally within 24 hours of diagnosis) and if enough people actually participate in the system. Belgian data were available as of September 2020.

Results

• During most of the study period (September 2020-October 2021), the average time between test result and start of contact tracing was less than 1 day in all three regions of Belgium.
• This time was longest during the second COVID-19 wave (September-October 2020), when a high number of persons needed to be to be contacted.
• Overall, the percentage of people contacted within 24 or 48 hours after a positive test result was high (above 80% and close to 100% respectively) for most of the study period.
• In all regions, less than 50% of contacted people (who agreed to answer some additional questions) had installed the Coronalert app, and even less (7-17%) actively used it to alert their contacts after a contamination. This is not enough for such a system to be effective.
• Data from October 2021 onwards (during the circulation of the delta and omicron variants) could not be analysed because the number of index cases that needed to be contacted increased dramatically, which resulted in a very high percentage of index cases that could not be reached (up to 67% during the week of November 15th, 2022 vs 8-9% on average during the study period).

Link to the technical sheet and detailed results

COVID vaccination in the last six months (at least one dose, % of the population) (R-10, R-11)

In the context of the COVID-19 pandemic, new vaccines were developed and approved faster than ever before and administered through large-scale public immunisation campaigns in the hope of halting the virus or at least mitigating its devastating effects. In Belgium, the first vaccination campaign was launched at the end of 2020, starting with residents and staff of community care institutions, then healthcare professionals, people aged 65 or more, younger adults with underlying health conditions and finally the rest of the population.

The percentage of vaccinated individuals – the so-called “vaccination coverage” – has been widely monitored in the weeks and months after the immunisation campaigns started. In this section, we will focus on the evolution of the percentage of the total population and of the population aged 65 or more who received at least one vaccine dose in the last six months. This will give us an overview of the early stages of deployment of the vaccine, but also of the participation in later booster campaigns.

Results

• In Belgium, over 97% of people aged 65 or more and nearly 90% of the overall adult population received their initial (primary) vaccination. For the adult population, the coverage rate was above the EU-27 (77.0%) and the EU-14 average (88.5%).
• Over 90% or people aged 65 or more and 76.3% of adults also received their first booster vaccination. For the adult population, this was also above the EU-27 (65.4%) and the EU-14 average (73.1%).
• The share of the population who received at least one dose in the last six months decreased after the first booster campaign. This is perfectly normal, as subsequent booster campaigns were spaced more than 6 months apart. However, this percentage remained far below initial levels even during new campaigns (peaking at 70.1% for the 85+, 68.1% for the 65+ and 38.0% for the general adult population), reflecting a lower uptake of booster vaccinations.
• The second and third booster campaigns were more successful in Flanders, possibly reflecting different invitation strategies in different regions of the country.
• At the end of 2022, 37.6% of adults and 65.8% of older adults (65+) in Belgium had received at least one dose during the last six months, but this share remained substantially higher in Flanders.
• 60% of healthcare professionals were vaccinated by the end of May 2021 and 90% by mid-November 2021. The share of healthcare professionals who also received a first booster peaked at 83.4% (vs 76.3% for the general adult population). At the end of 2022, 48.2% of them had received at least one dose in the last six months.

Link to the technical sheet and detailed results

People who perceived that the government is likely to be prepared for the next pandemic (% of respondents, Trust) (R-13)

Recent studies have shown that trust in the government is extremely important in the context of a health crisis like the COVID-19 pandemic: when people trust the officials making decisions, they are also more likely to comply with instructions (e.g. those relating to social distancing, mask wearing or vaccination). This can, in turn, result in lower infection rates and lower mortality.

Trust is closely related to the way the public perceives the ability of the government to learn the lessons from a health crisis like the COVID-19 pandemic, and to be better prepared for similar events in the future. This preparedness is part of what makes the health system resilient.

Results

• In 2021, approximately 37% of Belgian respondents expressed confidence that the government would be prepared for the next pandemic. This is lower than the EU-14 (49%) and EU-27 (48%) averages.

Link to the technical sheet and detailed results

Country preparedness to public health risks and acute events (average score on a 1-100 scale) (R-14)

This indicator is based on the all-capacity (overall) average International Health Regulations (IHR) score, which provides information about a country’s preparedness to public health risks and acute events using a self-assessment tool.

The IHR result from an agreement between all World Health Organization (WHO) Member States to work together for global health security. They are a binding set of legal instruments designed to ensure and improve countries’ capacity to prevent, detect, notify and respond to public health risks and acute events of national and international concern, while avoiding unnecessary interference with international traffic and trade. They cover a total of 13 core capacities (e.g. surveillance, response, preparedness, coordination, risk communication, human resources…) which are rated on a scale from 1 (no capacity) to 5 (sustainable capacity).

Results

• In Belgium, the overall average IHR score remained stable at Level 5 (range 81-100%) between 2010 and 2020, but decreased to Level 4 (range 61-80%) in 2021 and 2022.
• In 2022, Belgium’s overall score (63%) was lower than both the EU-14 (77%) and EU-27 (76%) averages.

Link to the technical sheet and detailed results