The Lombardy Emergency Medical System Faced with COVID-19: The Impact of Out-of-Hospital Outbreak

Abstract

Objective

The Lombardy region was among the areas most affected by COVID-19 infection worldwide; the Lombardy Emergency Medical System (EMS) responded immediately to this emergency. We analyzed several critical aspects to understand what occurred in that region.

Methods

This retrospective study compares the events managed by the dispatch center and the characteristics of the patients transported to the hospital -age, sex, SpO₂, deaths- managed by the EMS in Brescia and Bergamo provinces between March-April 2020 and March-April 2019. Ambulances’ waiting time at the hospitals before discharging patients and the patients’ severity at emergency department admission were also analyzed.

Results

EMS managed 37,340 events in March-April 2020, +51.5% versus 2019. “Breathing” or “Infective” events reported to the dispatch center increased more than ten-fold (OR 25.1, p < 0.0001) in March 2020 and two-fold in April 2020 compared to 2019 (OR 3, p < 0.0001). Deaths increased +246% (OR 1.7, p < 0.0001), and patients not transported to hospital +481% (OR 2.9, p < 0.0001) in March 2020 compared to 2019. In some hospitals, ambulances waited more than one hour before discharging the patients, and the emergency departments doubled the admission of critically ill patients. Transported patients for “Breathing” or “Infective” events were primarily males (OR 1.5, p < 0.0001). The patients had lower SpO₂ in 2020 than in 2019 and they were younger.

Conclusions

The Lombardy region experienced an unexpected outbreak in an extremely short timeframe and in a limited area. The EMS coped with this pandemic, covering an extremely higher number of requests, with a ten-fold increase in the number of events managed.

Key words:

• Emergency Medical System (EMS)
• COVID-19
• Lombardy
• pandemic
• respiratory failure

Objective

February 21^st, 2020. The Lombardy COVID-19 era began in Italy. From that day on, the Lombardy region was among the first areas to experience an outbreak of COVID-19 outside of China, and Italy appeared to be one of the countries most affected by the SARS-CoV2 infection. The Italian public authorities ordered a lockdown, initially of some areas in Lombardy, and subsequently of the whole country. The Lombardy Emergency Medical System (EMS) and several hospitals of the region were the first structures involved in the management of this health emergency (1).

In the winter months, it is customary to witness an increase in respiratory diseases, usually due to the well-known flu epidemics (Influnet report, ISS) (2,3). However, compared to previous years, during this pandemic, public authorities privileged hospital access through emergency vehicles (1): the Italian Ministry of Health and the National Health Service recommended to avoid reaching the hospitals directly, but to dial the emergency number 112 (the European Emergency Number, EENA) to evaluate the patients at home and to decide for any transport to hospital. In Lombardy, the 112 EENA is managed by the “Regional Agency for Emergencies” (Azienda Regionale Emergenza Urgenza, AREU). This led to an overload for the Lombardy EMS, which found itself in the front line in the management of the pandemic, managing a plethora of patients whose characteristics were not yet fully known and were defined day by day (1,4–6). In addition to the increased requests, the EMS was overwhelmed by the initial contagion of numerous health workers, who had to abstain from work; this emergent situation required a timely organization, moving part of the staff across the Lombardy provinces, in order to guarantee an adequate advanced vehicles service to meet the increasing needs. In the meantime, the number of deaths began to increase (4,7,8); hospitals were unable to accommodate all the patients who required hospitalization. New hospital beds were created in a noticeably short time, even outside the hospital (6). Field hospitals were structured in a few days, first in Brescia and in Bergamo. The worsening of the patients’ clinical conditions occurred also at home; when COVID-19 broke out, its severity was not completely known, and some dramatic events precipitated the clinical picture despite the apparent well-being of the patients, increasing the risk of death also at home (4,7,8). EMS was therefore faced with requests for patients’ home evaluation, transport from home to hospitals, transfers between hospitals, transfers outside the region by land and by air, increasing the disproportion between the need and the available resources (4,9,10)

In Lombardy, the COVID-19 outbreak created serious problems to the health system, due to the speed of propagation of the infection in an extremely short time and in a limited area. The greatest spread of the virus concentrated in very few provinces, resulting in one of the highest mortality rates from COVID-19 worldwide (from Johns Hopkins website); we cannot exclude that this case mortality rate may be due to the age of the people who were getting infected (11). In this scenario, understanding how to deal with the increasing request of first aid was the top priority. In this perspective we aimed to analyze the data recorded by AREU through the 112 EENA, comparing the activity performed by the EMS in March and April 2020 to the activity carried out, in the same months of the previous year, in two of the hardest hit areas of the region: the provinces of Brescia and Bergamo. We analyzed the events managed by the dispatch center, the number of deaths recorded by the Lombardy EMS, the mean waiting time of the ambulances at the hospitals, the typology of the patients managed, in order to investigate what occurred in the Lombardy region when the COVID-19 broke out.

Methods

This is a retrospective observational cohort study of two provinces of the Lombardy region, Brescia and Bergamo, distributed over an area of 7,540.52 km² (2.50% of the overall Italian area), with a population of 2,380,544 people (3.94% of the Italian population as of 1^st January 2020). Brescia and Bergamo were the provinces most affected by the outbreak of COVID-19. Until the 30^th of April, these two provinces had 5,216 deaths (18.7%) out of the 27,967 deaths occurred in the whole of Italy.

The study was conducted in accordance with the principles of the Helsinki declaration and was approved by the AREU Data Protection Officer on 30^th April 2020.

The EMS Italian System

Italy has 20 regions and the EMS is managed at a regional level. Each region organizes its healthcare system by means of 112 EENA (Public Safety Answering Points, PSAPs), following the directives of the Ministry of Health. 112 PSAPs receive all the emergency calls, geolocating them and forwarding them to the most appropriate second level PSAPs for the rescue: Police or Carabinieri, Fire Brigades, EMS. With regard to the EMS dispatch center, if the call refers to a patient, an “event” is generated, within which several calls can converge.

The Lombardy EMS is currently managed by AREU, created in 2007. AREU has since managed all regional EMS PSAPs, rescue coordination, blood transfusion and organ transportation activities, and as of 2009, AREU is responsible for the first 112 PSAP in Italy (Varese, Lombardy). The 112 PSAPs migration process is still ongoing in Italy. Nowadays AREU has authority on two regional services, and manages both the 112 PSAPs as the first level citizen contact point, and the EMS PSAPs as the second level dispatching centers for medical rescue (12).

In Brescia and Bergamo provinces, AREU manages the ambulances and the advanced vehicles, the number of which varies each year based on the local needs. Ambulances are usually operated by rescue volunteers or employees (EMTs), while advanced vehicles have a variable crew of two or three members. In the former case there are a rescue driver and a nurse, in the latter there is also a doctor.

Data Registry

Data were provided by the AREU Lombardy Headquarters Register. We analyzed the “events” of the two months of the pandemic peak in Italy, March and April 2020, and compared them to those of the previous year (March and April, 2019). The “event” represented the management of one patient. It means that the total number of events was lower compared to the call flow managed by the PSAPs, indeed for one single event more calls may have been necessary (i.e. the first rescue incoming call performed by bystanders, the call to send the ambulance, the call to describe the patient status, and so on). Each event implied sending a rescue vehicle, basic or advanced, based on the telephone dispatch. The event was classified in categories, one of which referred to “medical events”. This category was further divided into other groups, including “Breathing” or “Infective” labels. Within the “Breathing” or “Infective” events, we analyzed some of the patients’ characteristics related to the spread of COVID-19 (age, sex, peripheral oxygen saturation - SpO₂) and compared them to the previous year. We also registered the data related to the patients transported to hospitals, patients not transported to hospitals and deaths at home.

To understand the critical aspects of this pandemic, we investigated the severity of patients admitted to the emergency departments of the hospitals based on the color code, from the red (life-threatening cases) to the white code (patients with non-urgent and deferrable diseases). We also analyzed the mean waiting time of ambulances, before admitting the patients to the emergency departments of the hospitals.

Statistical Analysis

The categorical variables are presented as number and percentage, and the continuous variables are presented as median and interquartile range (IQR). The categorical variables were analyzed by means of χ² test, and the relative odds ratios (OR) and 95% interval confidences (IC95%) were provided. Continuous variables were treated by means of the Mann-Whitney test, or by means of Kurskal-Wallis test and the Dunn’s multiple comparison post-hoc analysis, when appropriate.

The effects of the year, age and sex on the SpO₂, considered as dependent variable, were analyzed by means of multiple regression.

Differences were considered significant when p < 0.05, otherwise they were considered non-significant (NS). The Prism 8.0.1 statistical software (GraphPad Software LLC, San Diego, CA, USA) was used to this aim.

Results

The total events processed by the EMS dispatch center in March and April 2020 were 37,340, +51.5% compared to 2019 (24,653 events). There was an over ten-fold increase in the “Breathing” or “Infective” events in March 2020 compared to 2019 (OR 25.1 [23.6–53.3], p < 0.0001, Figure 1 Panel A), and a two-fold increase in April 2020 compared to 2019 (OR 3.5 [3.3–3.7], p < 0.0001, Figure 1 Panel A). This trend was maintained in both provinces (see Supplementary Material Figure S1, Panels A and C).

Figure 1. Number of events managed by the EMS in March and April 2019 and 2020. Panel A reports the number of events managed by the dispatch center, the number of “Breathing” or “Infective” events and the number of patients transported to hospital. Panel B reports the number of patients not transported to hospital and the number of patients died. *: p < 0.0001; #: p = 0.054.

Concerning the number patients transported to hospital (Figure 1, Panel A, white bars) in March 2020, there was a + 56.7% increase compared to the previous year, although the ratio over the total events was lower during the COVID-19 pandemic (OR 0.29 [0.17–0.31], p < 0.0001, Figure 1, Panel A). In April 2020 the number of patients not transported to hospital was lower compared to the previous year, with a lower ratio for the former compared to the latter (OR 0.43 [0.40–0.46], p < 0.0001).

During the pandemic, the number of not hospitalized patients increased in both March (+481%) and April (+114%) (OR 2.9 [2.7–3.0] and OR 2.4 [2.3–2.6], p < 0.0001, respectively; Figure 1, Panel B, gray bars). Deaths increased by +246% in March 2020 compared to March 2019 (OR 1.7 [1.5–2.0], p < 0.0001), and +20% in April 2020 versus April 2019 (OR 1.2 [0.999–1.49], p = 0.054, Figure 1, Panel B, black bars). The previous data, divided by province, are reported in the Supplementary Material, Figure S1, Panels B and D. Similar differences between the years 2020 and 2019 were found for the two provinces, apart from the number of deaths, which was significantly higher in the province of Brescia in April 2020 than in April 2019 (+25%, p = 0.027), but not in Bergamo.

Once transported to the hospitals, the patients waited in the ambulance a longer time in 2020 than in 2019 before being admitted into the emergency department (Table 1). The table reports the median time (IQR) that ambulances waited before discharging the patients, divided in the two provinces, and subsequently with the data pooled together. The highest waiting time was recorded in March 2020. No differences were found between the two years (March versus April). The waiting time for each single hospital in the provinces, divided by month and year, is reported in the Supplementary Material, Table S1, where it is possible to note that for some hospitals the waiting time exceeded one hour.

Table 1. Waiting time for ambulance in hospital

	2019		2020
	March	April	March	April
Brescia	10 (2)	10 (1)	17 (3)*^#	15 (2)*^#
Bergamo	15 (2)	15 (2)	32 (28)*^#	17 (3)
Bergamo and Brescia	12 (6)	11 (4)	22 (16)*^#	16 (4)*^#

Ambulance median waiting time (IQR) before patient admission into the emergency department. Data refer to both areas (first two rows), and to the data pooled together (last row). Data are presented as median (IQR). Statistical analysis was performed by means of Kruskal Wallis test for non-parametric data. *: p < 0.05 compared to March 2019; ^#: p < 0.05 compared to April 2019.

Once admitted into the emergency department, patients were sorted by disease severity. In 2020 the patients admitted to the hospitals of the two provinces increased from 20,128 to 26,029 (+29%). Red and yellow code patients doubled in 2020 compared to 2019 (Table 2); contrariwise the number of “green codes” was significantly lower. Data divided by province are reported in the Supplementary Material, Table S2.

Table 2. Severity at admission into emergency department

	March and April 2019	March and April 2020	OR [IC95%]	p
RED CODE	359	624	1.4 [1.2–1.5]	<0.0001
YELLOW CODE	4,245	9,297	2.1 [2.0–2.2]	<0.0001
GREEN CODE	16,657	14,751	0.27 [0.26–0.28]	<0.0001
WHITE CODE	58	166	2.2 [1.7–4.9]	<0.0001
Total	20,128	26,029

Number of events in the hospitals of the two considered areas grouped by color code, data for Brescia and Bergamo pooled together. See Supplementary Material, Tables S1 and S2, for the data divided by single hospital and single province. The table reported the odds ratio; χ² test was performed to set differences.

The characteristics of the studied population are reported in Table 3. There were differences between 2019 and 2020 for the median age, the gender of the transported patients, and the SpO₂.

Table 3. Characteristics of the patients

	2019	2020	p
Patients, n	3,017	19,247
Age (y)
median [IQR]	78 [56–86]	70 [56–80]	<0.0001
Sex, n				OR [IC95%]
Male	1,347	10,392	<0.0001	1.5 [1.1–1.6]
Female	1,587	6,747	<0.0001	0.49 [0.45–0.53]
Missing data	83	2,108
SpO2 (%) median [IQR]
Ambient air	96 [92.25–98]	94 [88–97]	<0.0001
n	1836	10844
Oxygen	96 [93–98]	94 [90–97]	<0.0001
n	1047	6648
ΔSpO2	7 [5–11]	10 [7–15]	<0.0001
n	416	3243

Characteristics of the patients between 2019 and 2020. The categorical variable was tested by means of χ² test. Mann-Whitney test was applied for continuous variables. ΔSpO₂: difference between SpO₂ during oxygen administration and SpO2 in air.

The effects of age, sex and years on the SpO₂ were also tested by multiple regression and tested for multicollinearity. SpO₂ was significantly lower in 2020 than in 2019 (p < 0.001), in males compared to females (p = 0.01) and a significant inverse correlation was found for age (p < 0.001). The difference in SpO₂ (ΔSpO₂), in patients presenting both measurement in ambient air and during oxygen supplementation, was computed and compared between 2019 and 2020. ΔSpO₂ was higher for patients transported in 2020 compared to the previous year (p < 0.0001). Further analyses related to SpO₂ in males versus females are reported in the Supplementary Material, Table S3.

Discussion

The Lombardy region was among the first areas hit by an outbreak of COVID-19 outside of China and two of the most affected areas by the SARS-CoV2 infection were the provinces of Brescia and Bergamo. Brescia and Bergamo have a unique EMS dispatch center for health emergencies (SOREU Alpi, one of the Lombardy AREU Departments) (1,6,7,13), accessible by means of 112 PSAPs.

Compared with the same months of 2019, the number of events increased significantly mainly due to respiratory and infective problems: in March 2020 there was a ten-fold increase in the number of events managed by the second level PSAPs compared to 2019, while in April there was “only” a two-fold increase. As soon as the pandemic began, AREU organized a toll-free information service, set up in two 112 PSAPs within 48 hours, using personnel already employed in the regional service call centers. A special triage filter in 112 PSAPs was further created and developed, to evaluate COVID-19 cases. The first version aimed to avoid all the calls that were not related to a real emergency, in order to separate patients with real medical problems from people who needed assistance or information about COVID-19. The filter evolved several times during the pandemic, and new second level PSAPs were defined: two toll-free PSAPs for information and two support PSAPs dedicated to EMS. The application of filters reduced the response time from 10 minutes to an average waiting time of 12 seconds. This effect was both the result of the re-direction of non-emergency calls to the toll-free number PSAPs, and of the reorganization of ambulance PSAPs’ priorities and care management (12).

The total lockdown was declared on the 9^th March and its effects were obvious only weeks later (14). Likewise, the absolute number of patients transported to hospitals increased significantly, mainly in March. This incredible workload required a quick reorganization of the EMS: a new ancillary dispatch center was created (SOREU Lombardia) and a “Business Intelligence” (15) was applied to improve the EMS management response. The “Business Intelligence” refers to “techniques, technologies, systems, practices, methodologies, and applications that analyze critical business data to help an enterprise better understand its business and market and make timely business decisions” (16). “Business Intelligence” allowed reallocating the resources across the region, based on real-time data recording: the resources (ambulances and advanced vehicles) and staff were increased in the areas that registered new outbreaks. This allowed to take timely decisions (15).

Although the absolute number of patients transported to hospitals increased compared to 2019, the ratio over the total events was lower in 2020. In addition, there was an incredible increase in patients not transported to hospitals and the number of patients who died at home (Figure 1, Panel A and B, respectively). The increased number of patients not transported to hospitals maybe reflects an excessive request by the people who called without a real need, but it could also mean an ineluctable “selection” of the patients due to overcrowding of the reference hospitals (17,18). Indeed, hospitals were unable to accommodate all the patients requiring healthcare. Field hospitals were set up quickly to accommodate the increasing number of patients, but unfortunately it was not enough, and difficult end-of-life decisions were made in hospitals in these exceptional, resource-limited circumstances (17,19). In Bergamo, a 142-beds field hospital (72 intensive and sub-intensive beds, and 70 ordinary beds) was set up and became operative the 6^th of April, and in Brescia 70 beds managed by the local AREU organization were added outside the Spedali Civili Hospital of Brescia within two weeks from the beginning of the pandemic. Forty-two critically ill patients were flown to Germany, and seventy-seven in the nearby Italian regions by fixed or rotating wings.

The higher number of dead patients may reflect a criticality of the system, mainly linked to the lack of hospital beds; hospitals were unable to accommodate such a large number of patients in such a short time. This high number of deaths has been described also in other “in and out hospital” reports (7,20). Especially in March 2020, the hospitals were overwhelmed and the number of patients who presented to the emergency department was much higher than the number the system could cope with (21). An extraordinary effort was made to increase the maximum number of patients who could be treated (1,13,21). Moreover, as shown in Table 1 (and Supplementary Material, Table S1), the waiting time before being admitted into the emergency departments was longer. Especially in March, in Bergamo, the median waiting time was twice as long (15 minutes in 2,01,932 minutes in 2020), with peaks reaching more than 1 hour in some peripheral hospitals (Supplementary Material, Table S1). This longer waiting time meant that the ambulances were not available for new missions until the patient had been taken in charge by the hospital. The increased waiting time was inevitably linked to two factors; first, the high number of patients who reported to the emergency departments, second the severity of the patients admitted. As reported in previous papers (4,5,13) the severity of COVID-19 patients in Lombardy appeared to be greater than in other countries; the red and yellow codes in hospitals nearly doubled in 2020 compared to 2019. As described elsewhere (4,5,13), COVID-19 seemed to affect more males than females, and the respiratory pattern was significantly affected in a greater number of patients, both during breathing ambient air and also with oxygen supplementation (Table 3 and Supplementary Material, Table S3). Obviously, we were unable to diagnose the positivity to COVID-19 on the field, but we assumed that the majority of respiratory and infective events during March and April 2020 was due to SARS-CoV2 infection; indeed, since the Italian Istituto Superiore di Sanità (Influnet report, ISS) reported that the peak of the flu curve had been reached between the last week of January and the first week of February. Table S3 reports the range of SpO₂ of patients in ambient air and during oxygen supplementation. It is noteworthy that the SpO₂ range is much wider compared to the previous year, in ambient air and in oxygen, meaning that the patients had a lower SpO₂ in 2020; due to the shape of the hemoglobin dissociation curve, the patients had a higher ΔSpO₂ increase (Table 3) after oxygen supplementation.

The EMS faced with COVID-19 showed organizational ability and flexibility, covering an extremely higher number of requests; at the same time, hospitals were overwhelmed by this pandemic, and further steps had to be taken to address the outbreak.

Limitations

This is a retrospective study; thus, it does not have the strength of other typologies of study. This retrospective study allows formulating ideas about the relationship between pandemic and the management of the EMS, although causal statements cannot be made. It is noteworthy that in this case it was impossible to organize a prospective study. Unfortunately, the type of study affects the quality of data; i.e. in Table 3 some data are missing (gender). Moreover, we analyzed only the events classified in two categories, “Breathing” or “Infective”. This is to be emphasized, since many COVID-19 patients had other important symptoms that may not have been included under these labels, and a lot of details probably were missing. However, during the pandemic, the dispatch system was particularly careful and probably these two categories managed to include the majority of patients who subsequently tested positive for the virus (data not supported by this study). We cannot exclude that some patients living in outlying parts of the region may have gone to hospitals outside of the region.

Conclusion

Lombardy experienced an unexpected COVID-19 outbreak in an extremely short time and in a limited area. The lockdown was adopted to counteract the spread of the virus (14), however its effects occurred only after some weeks; meanwhile, the Lombardy EMS, AREU, fought a hard battle and held on. The EMS was flooded by requests and had to deal with a ten-fold increase in events management, while the ambulances transported a significatly higher number of patients compared to 2019. Patients were predominantly affected by respiratory failure, and their peripheral saturation level was lower compared to the same months of the previous year.