Presentations of children to emergency departments across Europe and the COVID-19 pandemic: A multinational observational study.

BACKGROUND: During the initial phase of the Coronavirus Disease 2019 (COVID-19) pandemic, reduced numbers of acutely ill or injured children presented to emergency departments (EDs). Concerns were raised about the potential for delayed and more severe presentations and an increase in diagnoses such as diabetic ketoacidosis and mental health issues. This multinational observational study aimed to study the number of children presenting to EDs across Europe during the early COVID-19 pandemic and factors influencing this and to investigate changes in severity of illness and diagnoses. METHODS AND
FINDINGS: Routine health data were extracted retrospectively from electronic patient records of children aged 18 years and under, presenting to 38 EDs in 16 European countries for the period January 2018 to May 2020, using predefined and standardized data domains. Observed and predicted numbers of ED attendances were calculated for the period February 2020 to May 2020. Poisson models and incidence rate ratios (IRRs), using predicted counts for each site as offset to adjust for case-mix differences, were used to compare age groups, diagnoses, and outcomes. Reductions in pediatric ED attendances, hospital admissions, and high triage urgencies were seen in all participating sites. ED attendances were relatively higher in countries with lower SARS-CoV-2 prevalence (IRR 2.26, 95% CI 1.90 to 2.70, p < 0.001) and in children aged <12 months (12 to <24 months IRR 0.86, 95% CI 0.84 to 0.89; 2 to <5 years IRR 0.80, 95% CI 0.78 to 0.82; 5 to <12 years IRR 0.68, 95% CI 0.67 to 0.70; 12 to 18 years IRR 0.72, 95% CI 0.70 to 0.74; versus age <12 months as reference group, p < 0.001). The lowering of pediatric intensive care admissions was not as great as that of general admissions (IRR 1.30, 95% CI 1.16 to 1.45, p < 0.001). Lower triage urgencies were reduced more than higher triage urgencies (urgent triage IRR 1.10, 95% CI 1.08 to 1.12; emergent and very urgent triage IRR 1.53, 95% CI 1.49 to 1.57; versus nonurgent triage category, p < 0.001). Reductions were highest and sustained throughout the study period for children with communicable infectious diseases. The main limitation was the retrospective nature of the study, using routine clinical data from a wide range of European hospitals and health systems.
CONCLUSIONS: Reductions in ED attendances were seen across Europe during the first COVID-19 lockdown period. More severely ill children continued to attend hospital more frequently compared to those with minor injuries and illnesses, although absolute numbers fell. TRIAL REGISTRATION: ISRCTN91495258 https://www.isrctn.com/ISRCTN91495258.

Introduction

Healthcare systems across Europe continue to be greatly affected by the Coronavirus Disease 2019 (COVID-19) pandemic. Early in the COVID-19 pandemic, urgent and emergency facilities prepared for a potential influx of acutely unwell children and young people [1]. However, evidence emerged that children were less likely to develop symptoms of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection, when compared with adults [2-6]. Moreover, reduced numbers of unwell or injured children visiting urgent and emergency care services were reported, and these seemed to be greatest for children with infectious communicable diseases [7-11]. Typically, these studies did not compare patterns between countries or in relation to different public health strategies.

At the same time, concerns were raised about potential delays in, and higher acuity of, presentations to appropriate healthcare services, as a result of difficulties accessing these services, changes in healthcare provision preferencing virtual consultations, fear of exposure to SARS-CoV-2 in healthcare facilities, and blanket “Stay at Home” statements [12-14]. In the United Kingdom, this resulted in a statement from the Royal Society of Pediatrics and Child Health to reassure parents and caregivers, urging them to seek appropriate urgent and emergency medical attention when worried about the acute illness or injury of their child [15]. Additionally, mostly anecdotal evidence reported increased numbers of specific childhood diagnoses, such as diabetic ketoacidosis [16] and intussusception [17]. These hypothesized a possible link with acute or prior SARS-CoV-2 infection, yet evidence from large-scale cohorts is lacking. Concerns were also raised for the mental health of children resulting from school closures and stay at home orders [18,19].

In this study, we aimed to compare the number of children presenting to emergency departments (EDs) across Europe during the first phase of the COVID-19 pandemic with the 2 previous years; investigating any change in severity of illness and describing the associations with specific diagnoses potentially related to SARS-CoV-2.

Methods

Study design, setting, and participants

This retrospective, observational study included 38 sites from 16 European countries as part of the “Epidemiology, severity and outcomes of children presenting to emergency departments across Europe during the SARS-CoV-2 pandemic” (EPISODES) study (trial registration number: ISRCTN91495258) (S1 and S2 Files). Sites were selected from the Research in European Pediatric Medicine (REPEM) and the Pediatric Emergency Research in the United Kingdom and Ireland (PERUKI) networks following the earlier work of Bressan and colleagues [1]. Routine clinical data from all children presenting to the ED were extracted from electronic health records for the period January 1, 2018 to May 17, 2020. The upper age limit varied between sites at between 16 and 18 years old. This study is reported as per the REporting of studies Conducted using Observational Routinely collected health Data (RECORD) statement (S1 Checklist) and the study protocol is available in the Supporting information (S1 File).

Aggregated, standardized data were uploaded using the REDCap online platform. For the period January 1, 2018 and February 1, 2020, data were collected on a monthly basis. For the period February 2, 2020 to May 17, 2020, on a weekly basis. This amounted to a total of 40 time windows (S1 Table). The clinical report form included 10 different data domains: (1) moment of presentation; (2) patient characteristics; (3) mode of arrival and referral pathway; (4) triage urgency; (5) type of presenting problem and vital signs; (6) diagnostics performed in the ED; (7) treatment in the ED; (8) diagnosis; (9) hospital admission; and (10) duration of ED and hospital stay (S3 File); data availability varied between sites (S1 Fig).

Triage urgency levels, used to determine the urgency of care in the ED, were categorized in 3 predefined categories, defined as emergent-very urgent (or RED-ORANGE, or level 1 to 2), urgent (or YELLOW, or level 3), and standard-nonurgent (or GREEN-BLUE, or level 4 to 5) to allow uniform coding between sites. For diagnosis coding, ICD-10 codes were issued for guidance (S2 Table), but an internally and temporally consistent coding approach was encouraged for each of the individual sites, acknowledging different coding systems and strategies in the ED. This was checked by plotting the diagnoses coding in time as percentage of total number of attendances for each site. To achieve reliable and accurate transformation of local (non-ICD-10) coding systems into the predefined diagnosis categories, training sessions were held and support offered to study sites by the lead investigators. Diagnoses were selected to reflect the broad spectrum of presenting problems to EDs, and their perceived change in incidences during the initial phase of the COVID-19 pandemic, following a consensus methodology among the study steering group. Final selection of diagnoses for analyses, after completion of data quality control process, included (1) common communicable diseases (e.g., tonsillitis, otitis media, lower respiratory tract infection (LRTI), gastrointestinal infection); (2) common minor injuries (e.g., minor head injury, radius fracture); (3) mental health issue; (4) diabetic ketoacidosis; and (5) surgical presentations (e.g., appendicitis, volvulus-intussusception-malrotation, testicular torsion). Severity was defined based on level 1 to 2 urgency classification at triage, any hospital admissions, pediatric intensive care unit (PICU) admission, or death in ED.

Data analyses

The completeness, quality, and internal consistency of data were checked by plotting the absolute numbers, as well as percentage of total attendances, for each variable of interest in time for the whole study period 2018 to 2020. In order to quantify changes in attendances, we compared observed attendances with predicted numbers of attendances. Predicted numbers of attendances were estimated using monthly data for the 25 months prior to February 3, 2020. As the data had both a trend and seasonal component, we used Holt–Winters exponential smoothing to make short-term monthly forecasts for February, March, April, and May 2020. We adjusted these to weekly estimates of predicted numbers. We plotted predicted ED attendances against the introduction of national infection prevention measures [20]. We also calculated 28-day mean numbers for selected diagnoses, PICU and hospital admission, and death in ED for each month from January through April for the years 2018 to 2020.

We used a Poisson model, adjusted for time since February 3, 2020, to determine if there were differences between age groups, diagnoses, and disposition for patients. For each model, the outcome was the count of attendances per week from the week beginning February 3, 2020 to the week beginning May 4, 2020, with an offset of the predicted number of attendances in each week. An incidence rate ratio (IRR) >1 indicates higher numbers compared with the reference group, whereas an IRR <1 reflects a higher reduction in numbers. For age groups, the analysis was adjusted for site; for diagnoses and disposition, numbers were too small to make forecasts at site level and we therefore aggregated these across the whole sample. For diagnoses, we completed 2 models, one with 8 separate diagnoses and one where these were divided into 3 groups: surgical presentation (i.e., appendicitis), communicable diseases (i.e., tonsillitis, otitis media, LRTI, and gastroenteritis) and “other” (i.e., mental health issue, radius fracture, and minor head trauma). For 3 diagnosis groups, the number of attendances was too low to make sensible forecasts, namely diabetic ketoacidosis, testicular torsion, and the combined group of intussusception, volvulus, and malrotation. In addition, we determined if there were associations between the change in hospital attendances and the prevalence of SARS-CoV-2 in the country, as per the European Centre for Disease Prevention and Control (ECDC), and the number of COVID-19 measures that were introduced in each hospital in response to the pandemic as previously detailed by Rose and colleagues [21]. Rose and colleagues performed a survey study describing changes in local and regional healthcare pathways, including the diverting of patient groups to or away from the ED, and service provision. The survey covered a total of 37 possible points of change in provision of care for sites without a short stay unit (20 pertaining to service provision and 17 to patient pathways) and 38 possible points of change for those that did (21 service provision and 17 patient pathways).

High-prevalence countries were defined as a cumulative 14-day rate of >80 new cases per 100,000 of the population. For countries with multiple sites, we used an ANOVA to determine if there was evidence that within country differences were greater than between country differences, for total attendances in March and April, adjusted for predicted numbers to account for differences in site sizes. One site (MAL001) was unable to provide information on diagnosis so it was excluded from this section of analysis; 3 sites (SLO001, POR005, and TUR001) did not provide triage data. Two sites were excluded (NL002 and HUN002) from the forecasting analyses and Poisson models as they had missing data in the period before the pandemic (2018). One site (IRE003) was excluded from the Poisson models because it closed to pediatric attendances in response to the pandemic. One site (TUR003) accounted for 18% of all attendances, and we carried out sensitivity analysis to confirm the changes to our findings when including this site. Analyses were performed using R v4.0.0.

Ethics

Following initial approval by the UK Health Research Authority, all participating sites obtained approval from their national/local institutional review boards (S3 Table). The need for individual patient informed consent was waived. Data sharing agreements were in place.

Results

Description of sites, infection prevention measures, and SARS-CoV-2 prevalence

Sites included in the study varied in terms of size and service provision (S4 Table and S2 Fig). The annual number of ED attendances ranged from 4,961 (NL001, 2019) to 295,787 (TUR003, 2019) (S3 Fig). All but 3 sites were tertiary academic hospitals with specialized pediatric EDs; the remaining 3 sites were general teaching hospitals, two of which had dedicated pediatric sections and staff, and one of which had a mixed ED. Sites in Austria, Slovenia, and the Netherlands mainly saw medical presentations, whereas the other sites saw both medical and surgical/trauma presentations. Timing and degree of infection prevention measures were similar across European countries (Fig 1 and S5 Table). Notably, Iceland and Sweden did not close day care, nurseries, or primary education; Germany and the UK kept higher education open; Sweden did not close any public spaces; Hungary and Sweden did not advocate use of face masks; Malta, Iceland, and Sweden did not introduce stay-at-home recommendations; and Germany, Hungary, and Iceland did not formally close workspaces. Highest national prevalence of SARS-CoV-2 varied between countries (Fig 1 and S6 Table).

Fig 1

Timelines of first phase of COVID-19 pandemic in participating countries.

Timelines of the introduction of national infection prevention measures (“Response measures”), as well as dates for the first and first 100 cases of SARS-CoV-2 for each of the countries participating in the EPISODES study. The black circle depicts the date of the highest 14-day cumulative rate of new SARS-CoV-2 cases per 100,000, with the size reflecting the actual case rate. COVID-19, Coronavirus Disease 2019; EPISODES, Epidemiology, severity and outcomes of children presenting to emergency departments across Europe during the SARS-CoV-2 pandemic; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2.

Changes in total attendances

All 38 sites had significant reductions in attendances in spring 2020 (Table 1 and Figs 2 and S4). The largest reduction was seen in AUS001 with observed numbers at 5% (95% CI 5% to 6%) of predicted in the week starting March 30, 2020; the smallest peak reduction in ED attendances was at 56% (95% CI 52% to 60%) of predicted in SWE001 during the same week. IRE003 closed for pediatric visits from March 30, 2020 onwards, with most patients diverted to IRE001. Poisson models, adjusted for time since intervention and predicted numbers of attendances, showed that there were significant differences between sites. Observed attendances, with respect to predicted, were relatively higher in sites in France, Sweden, Ireland, Iceland, Latvia, and the Netherlands, where observed attendance rates were greater than 50% of predicted. However, there was considerable overlap between all sites when 95% confidence intervals were considered. Results of the Poisson models suggest that attendances in Spring 2020 were higher in EDs in countries with lower SARS-CoV-2 prevalence (IRR 2.26, 95% CI 1.90 to 2.70, P < 0.001) (Table 2). We found a relationship between the number of introduced organizational COVID-19 measures and ED attendances and more organizational COVID-19 measures were associated with lower numbers of ED attendances when adjusted for predicted ED attendances (IRR 0.13, 95% CI 0.11 to 0.16, when sites with 4 or more measures were compared to sites with no measures, P < 0.001). Similarly, larger reductions in ED attendances were seen in mixed adult and pediatric academic hospitals (versus standalone children’s hospital, IRR 3.49, 95% CI 2.89 to 4.24, P < 0.001; general nonuniversity hospital, IRR 2.73, 95% CI 2.28 to 3.30, P < 0.001) and urban hospitals (versus mixed urban and rural hospitals, IRR 5.33, 95% CI 4.44 to 6.46, P < 0.001). ED attendances across all age groups significantly reduced (S5 and S6 Figs). Attendances in children aged above 12 months were reduced more than children below 12 months (12 to <24 months IRR 0.86, 95% CI 0.84 to 0.89; 2 to <5 years IRR 0.80, 95% CI 0.78 to 0.82; 5 to <12 years IRR 0.68, 95% CI 0.67 to 0.70; 12 to 18 years IRR 0.72, 95% CI 0.70 to 0.74; versus age <12 months as reference group, all P < 0.001) (Table 2). There was insufficient evidence to conclude that this pattern continued with increasing age for children aged 12 months and older. Patterns between sites within the same country appeared similar (S7 Fig) with strong evidence that between country differences were greater than within country differences (F value: 6.453; p: 0.002).

Table 1

Summary data on the lowest observed number of ED attendances during COVID-19 for each participating center.

Site	Week of first public health measure	Week of highest reduction	Observed number of ED attendances	Total % of predicted numbers (95% CI)	Overall reduction$	Date of highest 14-day cumulative rate of new cases per 100,000	Highest cumulative 14-day rate of new cases per 100,000	Total changes to health system	Total changes as % of possible changes
Austria	09-03-2020				75%	02-04-2020	102.33
AUS001		13-04-2020	16	5% (5% - 6%)*	85%			5	13.5%
AUS003		30-03-2020	119	17% (15% - 18%)	72%			1	2.7%
AUS004		30-03-2020	114	21% (19% - 23%)	71%			0	0.0%
France	24-02-2020				18%	11-04-2020	86.12
FR001		30-03-2020	470	41% (35% - 48%)	28%			2	5.3%
FR002		30-03-2020	154	52% (40% - 77%)	-6%			1	2.6%
FR003		30-03-2020	285	42% (36% - 49%)	28%			0	0.0%
FR004		13-04-2020	115	34% (27% - 48%)	23%			0	0.0%
Germany	02-03-2020					09-04-2020	86.36
GER001		30-03-2020	165	36% (33% - 39%)	53%			0	0.0%
Hungary	09-03-2020					23-04-2020	13.34
HUN001		23-03-2020	279	35% (33% - 37%)	52%			4	10.5%
Iceland	16-03-2020					03-04-2020	277.04
ICE001		06-04-2020	132	49% (44% - 55%)	40%			3	7.9%
Ireland	09-03-2020				56%	23-04-2020	213.02
IRE001		23-03-2020	404	50% (46% - 55%)	27%			3	7.9%
IRE002		30-03-2020	333	32% (28% - 36%)	55%			4	10.5%
IRE003		30-03-2020	0	0%+	86%			13	34.2%
Italy	02-03-2020				70%	02-04-2020	124.03
IT001		23-03-2020	105	20% (20% - 21%)	67%			1	2.6%
IT002		23-03-2020	36	12% (11% - 13%)	73%			8	21.1%
Latvia	16-03-2020					06-04-2020	20.52
LAT001		30-03-2020	491	38% (37% - 39%)	53%			0	0.0%
Lithuania	09-03-2020					04-04-2020	25.12
LIT001		30-03-2020	164	26% (24% - 27%)	62%			9	23.7%
Malta	09-03-2020					12-04-2020	46.20
MAL001		30-03-2020	68	13% (12% - 15%)	80%			1	2.7%
The Netherlands	09-03-2020					19-04-2020	86.57
NL001		16-03-2020	38	36% (33% - 38%)	44%			1	2.7%
Portugal	09-03-2020				71%	11-04-2020	109.02
POR001		13-04-2020	305	25% (23% - 28%)	70%			1	2.6%
POR003		30-03-2020	365	22% (20% - 24%)	70%			2	5.3%
POR004		30-03-2020	132	13% (12% - 15%)	74%			6	15.8%
POR005		20-04-2020	117	21% (18% - 23%)	69%			1	2.6%
Slovenia	02-03-2020					05-04-2020	28.55
SLO001		30-03-2020	12	8% (7% - 8%)	74%			3	7.9%
Spain	09-03-2020				70%	05-04-2020	217.56
SP001		23-03-2020	256	26% (24% - 28%)	62%			1	2.6%
SP002		30-03-2020	51	11% (10% - 11%)	77%			0	0.0%
Sweden	09-03-2020				36%	01-05-2020	83.60
SWE001		30-03-2020	661	56% (52% - 60%)^	31%			0	0.0%
SWE002		06-04-2020	253	48% (44% - 52%)	40%			0	0.0%
Turkey	09-03-2020				68%	22-04-2020	74.97
TUR001		27-04-2020	98	33% (29% - 38%)	59%			3	7.9%
TUR002		06-04-2020	233	15% (14% - 17%)	72%			6	15.8%
TUR003		13-04-2020	882	14% (13% - 15%)	75%			5	13.5%
United Kingdom	16-03-2020				63%	01-05-2020	99.25
UK001		30-03-2020	387	30% (28% - 32%)	63%			3	7.9%
UK002		30-03-2020	72	18% (17% - 21%)	71%			3	7.9%
UK004		23-03-2020	487	36% (34% - 39%)	55%			2	5.3%
UK005		13-04-2020	76	14% (12% - 17%)	71%			1	2.6%
UK006		30-03-2020	401	33% (31% - 35%)	54%			2	5.3%

The starting date of the week where the observed numbers of ED attendances had the largest difference from predicted numbers expressed in % with 95% confidence intervals. With

* AUS001 having the largest reduction from predicted numbers and ^ SWE001 the least change from predicted numbers. In addition, highest national SARS-CoV-2 infection rates are given for each of the study sites, as reported by the ECDC [20],with a threshold of 80 cases per 100,000 to indicate low (green) and high (orange) prevalence. Total changes to health system as per Rose and colleagues [21].

$Overall reduction calculated from introduction of first national public health measure until end of study period. If multiple sites per country, overall reduction for the country was estimated using the average of the overall reductions of each individual site.

+IRE003 was closed for pediatric attendances from March 30, 2020 onwards.

HUN002, NL002 excluded from table owing to missing data in 2018.

COVID-19, Coronavirus Disease 2019; ECDC, European Centre for Disease Prevention and Control; ED, emergency department; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2.

Fig 2

Observed versus predicted ED attendances (%).

The observed versus predicted number of children presenting to EDs in countries across Europe in the weeks following February 2, 2020 until May 11, 2020, for all sites combined. The color and the size of the dots reflect the actual number of ED attendances for each site and for each time window. The line connects the mean of the observed vs. predicted point estimates for each of the individual sites for each time window.

Table 2

Poisson regression models for ED attendances.

	IRR (95% Confidence Interval)	p-value
Number of COVID-19 measures in hospital $
No measures–reference group
1 measure	0.426 (0.391 - 0.463)	<0.001
2 measures	0.760 (0.717 - 0.806)	<0.001
3 measures	0.454 (0.413 - 0.499)	<0.001
4+ measures	0.130 (0.108 - 0.155)	<0.001
SARS-CoV-2 prevalence ^
Low prevalence	2.256 (1.904 - 2.700)	<0.001
Type of hospital
Mixed tertiary hospital–reference group
Standalone tertiary children’s hospital	3.490 (2.894 - 4.241)	<0.001
General nonuniversity teaching hospital	2.733 (2.280 - 3.303)	<0.001
Urban–reference group
Urban and rural mixed	5.333 (4.439 - 6.460)	<0.001
Age group
0-<12 months–reference group
12-<24 months	0.862 (0.838 - 0.886)	<0.001
2-<5 years	0.799 (0.780 - 0.819)	<0.001
5-<12 years	0.682 (0.666 - 0.698)	<0.001
12–18 years	0.719 (0.698 - 0.739)	<0.001
Triage urgency classification
Nonurgent and standard triage categories–reference group
Urgent	1.096 (1.076 - 1.116)	<0.001
Emergent and very urgent	1.530 (1.488 - 1.573)	<0.001
Diagnosis I
Appendicitis–reference group
Gastro-intestinal infections	0.279 (0.253 - 0.308)	<0.001
Minor head injury	0.783 (0.709 - 0.866)	<0.001
LRTI	0.357 (0.323 - 0.396)	<0.001
Mental health issues	0.688 (0.609 - 0.777)	<0.001
Otitis media	0.231 (0.206 - 0.260)	<0.001
Radius fracture	0.732 (0.654 - 0.819)	<0.001
Tonsillitis	0.189 (0.172 - 0.208)	<0.001
Diagnosis II
Surgical presentation–appendicitis–reference group
Communicable diseases	0.238 (0.253 - 0.308)	<0.001
Other	0.754 (0.709 - 0.866)	<0.001
Outcome
Admission–reference group
Death	1.749 (0.876 - 3.069)	0.077
PICU Admission	1.295 (1.157 - 1.445)	<0.001

To derive IRRs, the predicted counts for each individual site were used as an offset in the Poisson model to account for case-mix differences between sites.

$The number of changes made in each hospital in response to the pandemic as previously detailed by Rose and colleagues [21].

^Low-prevalence countries were defined as a cumulative 14-day rate of < = 80 new cases per 100,000 of the population as per the ECDC [20].

COVID-19, Coronavirus Disease 2019; ECDC, European Centre for Disease Prevention and Control; ED, emergency department; IRR, incidence rate ratio; LRTI, lower respiratory tract infection; PICU, pediatric intensive care unit; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2.

Triage urgency

Overall, there was a higher reduction (observed compared to predicted) in children with lower triage urgency when compared to children with higher triage classification (urgent triage, IRR 1.10, 95% CI 1.08 to 1.12, P < 0.001; emergent and very urgent triage IRR 1.53, 95% CI 1.49 to 1.57, P < 0.001; versus nonurgent triage category), even though clear reductions were seen for all triage categories (S8 Fig).

Hospital and PICU admissions

Hospital and PICU admissions were fewer than predicted (Figs 3, 4 and S9). We did not observe an increase in the number of deaths in ED (IRR 1.75, 95% CI 0.88 to 3.07, p = 0.08). The change in PICU admissions (IRR 1.30, 95% CI 1.16 to 1.45, versus general admissions) was not as great as the change in general admissions.

Fig 3

Observed versus predicted hospital admissions for patients attending the ED (%).

The observed versus predicted number of children admitted to hospital from the ED in countries across Europe in the weeks following February 2, 2020 until May 11, 2020, for all sites combined. The color and the size of the dots reflect the actual number of ED attendances for each site and for each time window. The line connects the mean of the observed vs. predicted point estimates for each of the individual sites for each time window.

Fig 4

Hospital admissions, intensive care admissions, and deaths in the ED for the period January–April over a 3-year period.

Percentages of total ED attendances (left) and absolute numbers (right) of children admitted to hospital (top), PICUs (middle), or died in the ED (bottom); comparing the 28-day mean numbers for the months of January–April for 2018 vs. 2019 vs. 2020. ED, emergency department; PICU, pediatric intensive care unit.

Diagnoses

The 28-day mean numbers for common communicable diseases decreased in absolute and relative frequencies (Table 3 and Fig 5A), in particular for tonsillitis, otitis media, gastrointestinal infections, and LRTIs. Decreases were also seen in common childhood injuries such as minor head injuries and radius fractures (Fig 5B). No increase in absolute numbers were seen for several uncommon diagnoses suggested to be linked with SARS-CoV-2 infection, such as diabetic ketoacidosis (Fig 5C), intussusception, and testicular torsion (Fig 5B), even when stratified for high-SARS-CoV-2 prevalence countries (S10 Fig). Mental health attendances declined during the first phase of the COVID-19 pandemic in absolute terms, but this corresponded with an increase in relative frequency (Fig 5C). Fig 6, reflecting the observed versus predicted numbers for the 8 selected diagnoses, shows that the change of children and young people with appendicitis was less than for the other diagnoses groups. Mental health issues, radius fractures, and minor head injuries were all affected, but there was evidence that attendances increased from the end of March. In contrast, attendances for LRTI, otitis media, gastrointestinal infections, and tonsillitis remained low. Poisson models showed no significant difference between mental health, minor head trauma, and radius fracture. There was evidence of significant difference between infections and trauma and mental health, with bigger reductions in infections. When communicable diseases were combined, there was a clear difference between surgical presentation (appendicitis), communicable diseases, and “other” (mental health, radius fracture, and head trauma) (Table 2).

Table 3

Totals of selected clinical diagnoses in the ED for the period January–April over a 3-year period.

	2018				2019				2020
	January	February	March	April	January	February	March	April	January	February	March	April
Tonsillitis	17,627(12.9%)	16,781(12.9%)	15,249(12.4%)	14,722(12.0%)	24,939(16.8%)	16,231(12.1%)	15,898(12.3%)	18,751(14.6%)	26,118(18.4%)	17,668(14.0%)	10,639(13.5%)	2,382(6.3%)
Otitis media	4,917(3.6%)	4,814(3.7%)	4,063(3.3%)	3,639(3.0%)	4,843(3.3%)	4,370(3.2%)	3,900(3.0%)	3,719(2.9%)	4,115(2.9%)	4,452(3.5%)	2,156(2.7%)	407(1.1%)
LRTIs	9,129(6.7%)	8,981(6.9%)	7,197(5.8%)	5,407(4.4%)	10,684(7.2%)	10,294(7.7%)	6,519(5.1%)	5,415(4.2%)	9,374(6.6%)	9,028(7.2%)	5,667(7.2%)	1,059(2.8%)
Gastrointestinal infections	7,809(5.7%)	8,488(6.5%)	8,946(7.3%)	9,103(7.4%)	10,313(6.9%)	9,949(7.4%)	11,345(8.8%)	11,066(8.6%)	9,158(6.4%)	8,726(6.9%)	4,719(6.0%)	1,905(5.1%)
Appendicitis	399(0.3%)	412(0.3%)	434(0.4%)	424(0.3%)	455(0.3%)	437(0.3%)	505(0.4%)	443(0.3%)	431(0.3%)	435(0.3%)	332(0.4%)	307(0.8%)
Testicular torsion	119(0.1%)	118(0.1%)	124(0.1%)	117(0.1%)	143(0.1%)	111(0.1%)	147(0.1%)	136(0.1%)	125(0.1%)	125(0.1%)	99(0.1%)	103(0.3%)
Intussusception, volvulus, and malrotation	33(0.0%)	37(0.0%)	56(0.0%)	58(0.0%)	40(0.0%)	58(0.0%)	63(0.0%)	60(0.0%)	55(0.0%)	44(0.0%)	53(0.1%)	30(0.1%)
Mental health conditions	705(0.5%)	652(0.5%)	718(0.6%)	736(0.6%)	811(0.5%)	821(0.6%)	853(0.7%)	757(0.6%)	791(0.6%)	802(0.6%)	603(0.8%)	388(1.0%)
Diabetic ketoacidosis	61(0.0%)	60(0.0%)	56(0.0%)	64(0.1%)	68(0.0%)	54(0.0%)	61(0.0%)	62(0.0%)	62(0.0%)	64(0.1%)	55(0.1%)	50(0.1%)
Radius fracture	783(0.6%)	891(0.7%)	853(0.7%)	1,311(1.1%)	886(0.6%)	1,036(0.8%)	1,153(0.9%)	1,362(1.1%)	809(0.6%)	935(0.7%)	730(0.9%)	592(1.6%)
Minor head injury	2,709(2.0%)	2,976(2.3%)	2,985(2.4%)	3,392(2.8%)	2,682(1.8%)	2,730(2.0%)	3,043(2.4%)	3,169(2.5%)	2,372(1.7%)	2,415(1.9%)	1,728(2.2%)	1,472(3.9%)

Absolute numbers (and % of total children seen in the ED) of children with diagnosis of (a) tonsillitis; (b) otitis media; (c) LRTIs; (d) gastrointestinal infections; (e) appendicitis; (f) testicular torsion; (g) intussusception, volvulus, and malrotation; (h) mental health issues; (i) diabetic ketoacidosis; (j) radius fracture; and (k) minor head injury; comparing the 28-day mean numbers for the months of January–April for 2018 vs. 2019 vs. 2020.

ED, emergency department; LRTI, lower respiratory tract infection.

Fig 5

Selected clinical diagnoses in the ED for the period January–April over a 3-year period.

Percentages of total ED attendances (left) and absolute numbers (right) of children with diagnosis of (A) common communicable diseases (tonsillitis, otitis media, LRTI, GI infections), (B) minor injuries and surgical presentations (radius fracture, minor head injury, appendicitis, intussusception, volvulus and malrotation (combined group), testicular torsion,), and (C) mental health issues and diabetic ketoacidosis; comparing the 28-day mean numbers for the months of January–April for 2018 vs. 2019 vs. 2020. ED, emergency department; GI, gastrointestinal; LRTI, lower respiratory tract infection.

Fig 6

Observed versus predicted number of selected diagnoses (%).

The observed versus predicted numbers of 8 selected diagnoses for all sites combined, for the period following February 2, 2020 until May 4, 2020. The error bars indicate the 80% prediction intervals.

Sensitivity analyses

The sensitivity analyses for the Poisson modelling without TUR003 resulted in IRRs slightly nearer to one, meaning all associations were slightly weaker (S7 Table). The change of the coefficient for tonsillitis was notable, increasing the IRR from 0.19 (95% CI 0.17 to 0.21) to 0.37 (95% CI 0.34 to 0.41). There was also a change between PICU admissions (IRR 1.13, 95% CI 1.00 to 1.28, p = 0.045) and admissions in general. Though the comparison remained statistically significant, the association was weaker.

Discussion

Reductions in the numbers of children attending EDs were consistently seen across Europe during the first phase of the COVID-19 pandemic. There was variation between countries, but within countries patterns were similar. The levels to which ED attendances decreased appeared to be related to the introduction and extend of infection prevention measures, changes made to local health systems, type of hospital, and national SARS-CoV-2 prevalence. Attendances were relatively higher in some sites with fewer or less strict national infection prevention measures (e.g., Sweden, Iceland), but this was not true for others (e.g., France). ED attendances were seen for all age groups, with smaller reductions in children aged below 1 year. The reduction in numbers was largest and sustained for communicable diseases, whereas other groups of diagnoses trended towards normal levels of ED attendances by the end of the study period after initial reduced ED attendance rates.

Our findings of reduced pediatric ED attendances are consistent with other studies from around the world [7-11,22-24]. The observed reduction in ED attendances will likely be multifactorial, including changed parental health-seeking behavior, modified and newly introduced healthcare pathways, and fewer circulating and reduced transmission of infectious pathogens. For example, children with asthma often frequent EDs, but they had fewer exacerbations needing ED visits during the first phase of the COVID-19 pandemic. Proposed reasons include reduced air pollution, reduced social mixing with exposures to viral trigger, and improved compliance with medication at home [25,26]. All this appeared to have affected presentations of children and young people with a low triage urgency and with minor injuries and illnesses most.

Earlier studies suggested that infection prevention measures may have resulted in delayed presentations to hospitals [12,13,27-29]. In our study, children with more severe conditions, as measured by triage urgency, need for hospital admission, and PICU and death, continued to attend hospital more frequently compared to those with minor injuries and illnesses, although overall absolute numbers fell. This was in line with other studies reporting similar reductions in children with high triage urgency or need for hospital admission [7,30-38].

Defining the harm of delayed presentations, as well as establishing what contributed to a possible delay, can be difficult [39]. In an attempt to distinguish the delay in seeking care from harm sustained, Roland and colleagues concluded that only a minority (6 out of 51 (11.8%)) of children with a potential delay in presentation were admitted to 1 of 7 hospitals [40]. Contradictory conclusions have been reported for the delay in presentations and for potential harm sustained for diagnoses of appendicitis [41,42] and testicular torsion [43-46], portraying a picture that organizing regional healthcare delivery is important to ensure continued access to pediatric urgent and emergency care during a pandemic. In addition, our data showed that, despite overall falling ED attendances, presentations requiring surgical interventions remained stable, reiterating that access to surgical teams and the ability to perform emergent surgical procedures are crucial.

Evidence is mounting that SARS-CoV-2 is directly involved in the pathogenesis of new onset diabetes [47,48]. Unsworth and colleagues first reported an increase of new onset type 1 diabetes in children and a possible link with SARS-CoV-2 in the UK [16]. Additional cohort studies found divergent associations between SARS-CoV-2, new onset diabetes, and decompensation of preexisting diabetes [49-52]. Our data did not identify increased incidence of diabetic ketoacidosis during the first phase of the COVID-19 pandemic. It might well be that clusters of new onset diabetes can be found in high-prevalence areas and that we failed to capture this in our study. Likewise, if SARS-CoV-2 acts as a precipitator, there might be a delay in the manifestation of new onset diabetes, and with reduced prevalence of typical viral triggers, this increase might only become apparent later in the pandemic [53]. We were not able to differentiate between new onset diabetes and decompensation of preexisting diabetes.

We found a reduction of children with mental health conditions presenting to the EDs during the first phase of the COVID-19 pandemic in Europe, similar to findings from studies elsewhere [9,54-57]. This is unlikely to reflect the considerable mental health issues encountered in the wider pediatric and adolescent populations [58] and of the experiences later in the pandemic, with, among others, reported increases in eating disorders in children and young people [59]. Joyce and colleagues observed an overall decrease in mental health issues in ED, albeit an increase in self-harm and deliberate ingestions [60]. Despite the reduction in absolute numbers, there was an increase in the proportion of attendances attributable to mental health potentially contributing to the heightened awareness for mental health issues in the first COVID-19 wave.

Prior to the current COVID-19 pandemic, limited data were available describing the association of infection prevention measures on urgent and emergency pediatric care in high-income countries. One study found a decrease in respiratory infections of 42% and decreased ED attendances of 28% following school closures for an influenza outbreak in Israel [61]. Similar patterns were seen during the SARS outbreak in 2003 [62-64] and the MERS outbreak in 2015 [65]. These studies also reported a larger reduction in ED utilization for children than for adults. In contrast, the 2009 H1N1 influenza pandemic generally led to increased ED utilization, with higher levels of acuity [66-68]. One previous study had reported reduced pediatric ED attendance rates for flu-like illness and respiratory tract infections following school closures [69]. Another study reported increased pediatric ED attendance numbers following media reports on health threats of the H1N1 virus [70]. Altogether, previous evidence of infectious disease outbreaks suggests a similar impact on pediatric urgent and emergency care following the introduction of public health and infection prevention measures. However, this is to a lesser extent than what was observed with the COVID-19 pandemic and one that is dependent on childhood susceptibility for the infectious pathogen.

Strengths and limitations

Our study presents multinational data enabling the comparison between infection prevention measures, national SARS-CoV-2 prevalence, and the association with acute illness and injuries in children between European countries. Most participating sites were tertiary institutions, with dedicated pediatric emergency medicine teams, with potential implications for the generalizability of our findings. At present, no standardized data extraction system for pediatric urgent and emergency care exists between European countries; and the EPISODES study is, to our knowledge, the first to navigate the difficulties of dealing with different data systems, data availability, and varying coding practices. Hence, also limited by the time restrictions caused by the COVID-19 pandemic, some sites were not able to provide data for all domains, and 2 sites (NL002 and HUN002) were only able to provide data for part of the study duration.

Limitations of electronic health records to describe patients’ diagnoses are well known [71]. Some of the participating study sites had unique non-ICD-10-based coding systems, and we urged all study teams to be consistent in transforming local data to fit the study clinical report form and we implemented a rigorous data quality process to ensure validity of coding in time. Although most diagnoses linked to SARS-CoV-2 in children were included in the predefined clinical report form, other diagnoses might be of interest in future studies. Of note, coding for children with Multi Inflammatory Syndrome in Children (MIS-C) proved unreliable, with no unique diagnostic codes available for this new disease in automated coding systems.

As the data were collected in aggregated form, thereby negating some of the difficulties with data protection regulations, we were not able to stratify for severity of specific diagnosis or age groups. This also introduced risks of overstratification during analyses. We observed large differences between sites for the number of annual ED attendances and the number of patients with high triage urgency and hospital admissions, reflecting both case-mix and diversity of patient management. We analyzed data mostly on a site-by-site basis, by using predicted versus observed ratios, and thus dealing with heterogeneity between sites. In addition, although very few study sites restructured local healthcare pathways diverting urgent and emergency care to alternative healthcare facilities, this does not fully rule out changes to access to healthcare or parental health-seeking behavior. Mongru and colleagues showed that, for one of the sites included in this study, the distribution of patients across primary and secondary care was similar before and during the first year of the COVID-19 pandemic, suggesting the observed reductions in patient numbers, especially for those with minor injuries and illnesses, were a true reflection of fewer children in the community in need of urgent and emergency care [72].

We used the cumulative 14-day rate of new cases per 100,000 of the population to identify high-prevalence countries, but indications for SARS-CoV-2 testing and reporting mechanisms differed between countries, and this could have led to under- or overestimation of national prevalence rates. Moreover, national prevalence numbers might wrongly reflect any regional variation, but, for example, in the UK, identical patterns in ED attendances were seen across the 5 sites, despite large variations in SARS-CoV-2 prevalence during the first phase of the COVID-19 pandemic [73]. Due to the rapid escalation and near-universal introduction of infection prevention measures in European countries during the study period, we were not able determine the role of each of the individual measures on reducing ED attendances.

Conclusions

Reductions in overall ED attendances were seen across our study sites during the first phase of the COVID-19 pandemic, with health systems across Europe affected similarly. In most sites, there was no suggestion of disproportionate numbers of more severely unwell children. In the first phase of the COVID-19 pandemic, the relative increase in cases of diabetic ketoacidosis or mental health issues might have contributed to a biased perception about increased occurrence, yet this is not supported by an increase in absolute numbers of cases in our data. Our study informs how pediatric emergency medicine can prepare for future pandemics, taking into account that different infectious diseases outbreaks can affect children differently, and illustrates the potential of electronic health records to monitor trends in urgent and emergency care for children.

Membership of the EPISODES study group.

(PDF)

Click here for additional data file.

EPISODES study protocol.

(PDF)

Click here for additional data file.

EPISODES study registration.

(PDF)

Click here for additional data file.

Clinical report form.

(PDF)

Click here for additional data file.

RECORD checklist.

The RECORD statement—checklist of items, extended from the STROBE statement, which should be reported in observational studies using routinely collected health data.

(PDF)

Click here for additional data file.

List of time windows for data entry.

(PDF)

Click here for additional data file.

ICD-10 guidance for coding of diagnosis.

(PDF)

Click here for additional data file.

List of approvals.

(PDF)

Click here for additional data file.

Overview of participating study sites.

(PDF)

Click here for additional data file.

List of national social distancing measures.

(PDF)

Click here for additional data file.

List of national SARS-CoV-2 rates.

The dates and numbers of SARS-CoV02 infections in each of the study sites’ countries participating in the EPISODES study.

(PDF)

Click here for additional data file.

Sensitivity analyses: Poisson regression models for ED attendances without TUR003.

(PDF)

Click here for additional data file.

Spiderplot for availability of data.

(PDF)

Click here for additional data file.

Map of European with all participating study sites.

All participating study sites are highlighted with their study site ID; represented countries in red.

(PDF)

Click here for additional data file.

Annual ED attendance numbers in 2018 and 2019.

The total number of pediatric ED attendances for each of the study sites for 2018 and 2019 (pre-COVID-19). Overall, the numbers of ED attendances in 2018 and 2019 were similar for each of the study sites, with considerable diversity between the study sites and TUR003 seeing more children in their ED than the other study sites.

(PDF)

Click here for additional data file.

Annual ED attendance numbers 2018 to 2020 for all sites separately.

The total number of pediatric ED attendances for each of the study sites for the entire study duration (January 2018–May 2020). The y-axis is depicted in log scale.

(PDF)

Click here for additional data file.

Observed versus predicted ED attendances (%) for different age categories.

The observed versus predicted number of children presenting to EDs in countries across Europe in the weeks following February 2, 2020 until May 11, 2020, for all sites combined, for children (a) aged 0–1 years; (b) 1–2 years; (c) 2–5 years; (d) 5–12 years; and (e) 12–18 years. The color and the size of the dots reflect the actual number of ED attendances for each site and for each time window. The line connects the mean of the observed vs. predicted point estimates for each of the individual sites for each time window.

(PDF)

Click here for additional data file.

Observed versus predicted ED attendances (%) for different age categories, for individual sites.

(PDF)

Click here for additional data file.

Observed versus predicted ED attendances (%) for each country.

The observed versus predicted number of children presenting to EDs in countries across Europe for which data from only 1 study site were available in the weeks following February 2, 2020 until May 11, 2020. A timeline is plotted (dashed line) to show the dates of the introduction of national social distancing measures.[20] One site from the Netherlands and 1 site from Hungary were excluded from these analyses as these sites could not provide data for the entire study duration.

(PDF)

Click here for additional data file.

Observed versus predicted triage categories (%).

The observed versus predicted number of children presenting to EDs in countries across Europe in the weeks following February 2, 2020 until May 11, 2020, for all sites combined, for children (a) nonurgent and standard triage classification; (b) urgent triage classification; (c) emergency and very urgent triage classification. The color and the size of the dots reflect the actual number of ED attendances for each site and for each time window. The line connects the mean of the observed vs. predicted point estimates for each of the individual sites for each time window. UK001 did not use a triage system with the emergency and very urgent triage category.

(PDF)

Click here for additional data file.

Percentage of children admitted to hospital for individual sites.

Percentages of total ED attendances (left) and absolute numbers (right) of children admitted to hospital (top) and PICUs (bottom); comparing the 28-day standardized numbers for the months of January–April for 2018 vs. 2019 vs. 2020. ED, emergency department; PICU, pediatric intensive care unit.

(PDF)

Click here for additional data file.

Selected clinical diagnoses in the ED for the period January–April over a 3-year period, for high-prevalence countries.

Percentages of total ED attendances (left) and absolute numbers (right) of children with diagnosis of (a) tonsillitis; (b) otitis media; (c) LRTIs; (d) GI infections; (e) appendicitis; (f) testicular torsion; (g) intussusception; (h) mental health issues; (i) diabetic ketoacidosis; (j) radius fracture; and (k) minor head injury; comparing the 28-day standardized numbers for the months of January–April for 2018 vs. 2019 vs. 2020, shown for countries with of a cumulative 14-day rate of new SARS-CoV-2 cases per 100,000 of 80 or more. ED, emergency department; GI, gastrointestinal; LRTI, lower respiratory tract infection; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2.

(PDF)

Click here for additional data file.

22 Mar 2022

Dear Dr Nijman,

Thank you for submitting your manuscript entitled "Patterns of presentations of children to emergency departments across Europe and the impact of the COVID-19 pandemic: retrospective observational multinational study." for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Mar 24 2022 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Callam Davidson

Associate Editor

PLOS Medicine

11 May 2022

Dear Dr. Nijman,

Thank you very much for submitting your manuscript "Patterns of presentations of children to emergency departments across Europe and the impact of the COVID-19 pandemic: retrospective observational multinational study." (PMEDICINE-D-22-00899R1) for consideration at PLOS Medicine.

Your paper was evaluated by an associate editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We hope to receive your revised manuscript by Jun 01 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Callam Davidson,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

Please revise your title. "Impact of" should be used only if causality can be inferred, i.e., for an RCT. Please also remove the word ‘retrospective’. I suggest ‘Presentation of children to emergency departments across Europe and the COVID-19 pandemic: a multinational observational study’, or similar.

Abstract:

* Please combine the Methods and Findings sections into one section, “Methods and findings”.

* Please change ‘Interpretation’ to ‘Conclusions’

* Please remove funding information from the abstract and instead include this in your Financial Disclosure (submission form).

* Please remove the trial registry information and instead include this in your Methods.

Abstract Background: Provide the context of why the study is important. The final sentence should clearly state the study question.

Abstract Methods and Findings:

* Please include p values when quantifying the main results.

* Please include the important variables that are adjusted for in the analyses.

* In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

Please update your line numbering to be continuous (rather than restarting on each new page).

Citations should precede punctuation.

Please confirm that individual patient informed consent was waived by all participating IRBs.

Please remove the Role of the Funding Source section.

Please ensure that the study is reported according to the STROBE guideline, and include the completed STROBE checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)."

The STROBE guideline can be found here: http://www.equator-network.org/reporting-guidelines/strobe/

When completing the checklist, please use section and paragraph numbers, rather than page numbers.

Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

Please do not cite figures that present results in the Methods section.

The shading in Table 1 does not appear to be consistent.

Please update the title of Figure 2 to ‘Observed versus predicted emergency department attendances (%)’ or similar. Please update the titles of Figure 3 and 6 in a similar way.

Figures 2, 3, and 6: You refer to ‘predicted’ in the title and legend but ‘expected’ in the axis label, please update to be consistent.

Table 2:

* Please report P<0.001 rather than P<0.0001.

* Please specify the variables controlled for in the legend.

* Please present the unadjusted comparisons as well as the adjusted comparisons.

Abstract/Results/Discussion: Your study is observational and therefore causality cannot be inferred. Please remove language that implies causality, such as ‘impact’. Refer to associations instead.

Figure 5 is very large – please consider reducing the size of the individual panels (as in Figure 4) or moving this figure to the Supporting Information.

Figure 6:

* The x-axis is labelled incorrectly.

* Please indicate the meaning of the error bars.

Please quantify the findings presented in the Sensitivity analyses subsection, with 95% CI and p-values where appropriate, rather than describing a result as less significant.

Discussion:

* Please temper claims of primacy (such as that found at line 2, page 23) by adding ‘to our knowledge’.

* Figure S17 appears to present UK government data, please consider citing the appropriate data in your References rather than including as an additional figure in the Supporting Information.

Please remove the Data sharing statement, Conflict of interest, Funding source, and Author contributions from the end of the main text and instead report this information in the appropriate parts of the submission form. In the event of publication, this information will be presented as metadata.

For internet sources in the References, please present the date cited.

Comments from the reviewers:

Reviewer #1: This study has investigated the impact of the COVID-19 pandemic and infection prevention measures on children visiting emergency departments across Europe. The study has provided important implications for the understanding of the change of the ED attendances during the pandemic. However, to make the review more valuable, I have some comments and hope these will help to improve this manuscript.

1. Page 1, Line 9: aged <16 years? Why the authors stated that upper age limit varied between sites at between 16 and <18 years old?

2. Page 2-4: Please give the full name for the "COVID-19" and "SARS-CoV-2" when they appeared for the first time.

3. Page 3, Line 4-9: I seem to disagree with the authors. The reason why children less likely to develop symptoms of severe acute SARS-CoV-2 infection may not attributable to reduced numbers of children visiting urgent and emergency care services. The authors should review more relevant studies to clarify the point clearly.

4. Page 3, Line 20: did the possible link exist? How did the authors infer this conclusion?

5. Page 4, Line 1-2: Why did author only compare the situations during the first phase of the COVID-19 pandemic with the two previous years? The pandemic has last for more than two years. The authors should consider a longer period if possible.

6. Page 5, Line 5: what were the triage urgency levels? How did the authors define the different levels?

7. Page 6, Line 1: How did the author adjust short-term monthly forecasts to weekly estimates of predicted numbers?

8. Page 8, Line 17: "Description of infection prevention measures and SARS-CoV-2 prevalence". The authors should combine the two aspects of information together in this descriptive paragraph.

9. Page 10, Line 2-3: How did the authors define the organisational COVID-19 measures? Please clarify the point in the method section.

10. Page 19, Line 21-23; Page 20, Line 1-3: The authors should discuss more about multifactorial reasons.

11. Fig 1 should be revised. There are many overlaps among the figure legends which make it hard to get important information.

12. Could authors provide a map about the distributions about the site? The map will help to the understanding about the information in the Figure 2 and Figure 3.

Reviewer #2: "Patterns of presentations of children to emergency departments (ED) across Europe and the impact of the COVID-19 pandemic: retrospective observational multinational study" examines retrospective health data from 16 European countries from Jan 2018 to May 2020, to attempt to quantify the impact of the coronavirus pandemic on ED presentations by children aged <16 years. Data prior to the pandemic was used to evaluate that from after the pandemic, after trend and seasonal adjustments. The main findings are that a reduction in ED attendances, hospital admissions and high triage urgencies was observed after the start of the pandemic, broadly in line with prior literature.

The analysis appears largely comprehensive and sound. However, some issues might be considered:

1. The relatively short period of assessment (Feb 2020 to May 2020) may be noted, despite it currently being April 2022. Might it have been possible to describe a longer period post-pandemic start, especially since that might allow potential annual cycles/patterns to be analyzed?

2. In the Abstract, it is claimed that "...ED attendances were relatively higher in ...children aged >12 months (12-<24 months IRR 0.89, 95% CI 0.86 to 0.92; 2-<5years IRR 0.84, 95% CI 0.82 to 0.87; 5-<12 years IRR 0.74, 95% CI 0.72 to 0.76; 12-<16 years IRR 0.74, 95% CI 0.71 to 0.77; vs. age <12 months as reference group)". However, it appears that attendances were actually relatively lower (IRR<1) in older children compared to age <12 months. This might be checked.

3. In Page 4, it is stated that the study included 38 sites from 16 European countries. The representativeness of the sites might be commented upon. Analysis for geographic status (urban vs. rural) might be considered.

4. In Page 4, it is stated that data from before Feb 2020 was collected monthly, and weekly after Feb 2020 (after the start of the pandemic). Was this choice of period (monthly vs. weekly) due to the limitations of the data on the online platform for the period before Feb 2020, or was it a conscious choice?

5. In Page 4, it is stated that standardized (aggregated) data was used. The next page then describes triage urgency levels that were categorized in three categories. For these relevant data/variables, they might be listed (possibly in supplementary material), and clarified as to whether all participating EDs followed the same classifications/system. If not, the conversions to a standardized norm might be given.

6. In Page 5, it is stated that "For diagnosis coding, ICD-10 codes were issued for guidance (S5 Table), but an internally and temporally consistent coding approach was encouraged for each of the individual sites, acknowledging different coding systems and strategies in the ED". This might be further clarified. In particular, while the approach was encouraged, was it actually implemented for all sites? Given the different coding systems/strategies, how were they standardized?

7. In Page 5, it is stated that Holt-Winters exponential smoothing was used to make short-term forecasts for February to May 2020. A brief analysis of the chosen smoothing method on known data (e.g. fitting with data from Jan 2018 to Jan 2019, predicting and comparing against Feb to May 2019) might be strongly considered.

8. In Page 14, Table 2 contains analysis for low SARS-CoV-2 prevalence. Analysis for more stratification (e.g. percentile groups) might be considered.

9. In Page 15, while it is claimed that no increase in number of deaths in ED was observed, the IRR from Table 2 appears to be 1.75 (0.88-3.07). This appears to actually suggest a considerable increase, if not statistically-significant. As such, this point might be discussed further.

10. In Page 16, the selection criteria for the various diagnoses presented in Table 3 might be briefly discussed.

11. In Figure 4, deaths in the ED were described. It might be considered to also show figures for (excess) deaths for the age group of interest (<16 years), for context.

12. Related to the above, medical options for children during the studied post-pandemic start period might be briefly discussed. Were they possibly diverted to private clinics/home care/telemedicine?

13. Any patterns from country-level analyses (as from Table/Figure 1) relating response measures to general outcomes might be discussed in the main text.

14. For Table/Figure 1, the source and reliability of data relating to (cumulative) rate of new SARS-CoV-2 cases might be discussed, given the possible difficulty of testing especially earlier in the pandemic.

Reviewer #3: This is an observational study that reports the changes in presentations of children to emergency departments. The authors have collected data across multiple European countries. Documenting this important change in service demand is important for understanding for future pandemic planning, so while the results are not unexpected they are still valuable.

These comments are intended to improve the usefulness of the manuscript.

1. The paper is well written and generally accessible, though the presentation of data can be a little confusing. This may be made easier by having an initial figure one that simply reports the total number of attendances and admissions over time. It feels a jump to go straight into actual / predicted

2. The finding about low acuity cases being the group that stayed away most is important. The authors should enhance the discussion about why this has occurred to explicitly consider whether there are any behavioral factors at play. Within adults, there was definitely a fear about attending hospital as a risk of exposure e.g. https://www.cureus.com/articles/34364-the-attend-study-a-retrospective-observational-study-of-emergency-department-attendances-during-the-early-stages-of-the-covid-19-pandemic It is very plausible that parents perception of risk / benefit to attending a paediatric ED with lower acuity presentations might have changed.

Any attachments provided with reviews can be seen via the following link:

[LINK]

18 Jul 2022

Dear Dr. Nijman,

Thank you very much for re-submitting your manuscript "Presentations of children to emergency departments across Europe and the COVID-19 pandemic: a multinational observational study." (PMEDICINE-D-22-00899R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by three reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jul 25 2022 11:59PM.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Please provide the URL from which data are available in your Data Availability Statement.

Please provide the full study protocol in your Supporting Information and cite this in your Methods.

Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

Line 27 and line 304: Please quantify the association with general admissions.

Line 56-57: Please quantify the key results presented in this bullet point (with 95% CI and p-values).

Please aim to trim your author summary such that there are 2-3 bullet points per question.

Your study is observational and therefore causality cannot be inferred. Please remove language that implies causality, such as that found at lines 63-64. Refer to associations instead.

My sincere apologies but in my previous decision letter I mistakenly suggested the STROBE checklist ought to be included in your Supporting Information. This ought to have been the RECORD checklist (https://www.record-statement.org/checklist.php). Please update the statement at lines 110-111 and include this checklist in place of the STROBE (again, referring to paragraph number and section headings rather than page numbers).

S8 Figure: Please confirm that the appropriate usage rights apply to the use of this map. Please see our guidelines for map images: https://journals.plos.org/plosmedicine/s/figures#loc-maps

Figure S17: AUS001 is missing.

Line 370: Please present the results of the sensitivity analyses in full in the Supporting Information.

Comments from Reviewers:

Reviewer #1: I have read the revised manuscript "Presentations of children to emergency departments across Europe and the COVID-19 pandemic: a multinational observational study". The authors have made revisions well, point by point. However, minor revisions should be made to improve the manuscript for some minor errors and the syntax and language.

For example:

1、 Line 51: "…16 years and under…" should be "…18 years and under…"?

2、 Improper use of commas in the Abstract of "Concerns were raised about the potential for delayed, and more severe, presentations, and an increase in diagnoses such as diabetic keto-acidosis and mental health issues."

3、 Abstract: "This observational multinational study …" should be "This multinational observational study…"

4、 Line 109: "Upper age limit…" should be "The upper age limit…"

Reviewer #2: We thank the authors for addressing our previous comments. Brief mention might be made about (over-)stratification in the manuscript if considered appropriate.

Reviewer #3: The authors have done a diligent job of the revisions and answered the questions adequately

Any attachments provided with reviews can be seen via the following link:

[LINK]

25 Jul 2022

Dear Dr. Nijman,

Thank you very much for re-submitting your manuscript "Presentations of children to emergency departments across Europe and the COVID-19 pandemic: a multinational observational study." (PMEDICINE-D-22-00899R3) for review by PLOS Medicine.

The remaining issues that need to be addressed are listed at the end of this email.

In revising the manuscript for further consideration here, please ensure you address the specific points made by the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

We hope to receive your revised manuscript within 3 days. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We look forward to receiving the revised manuscript by Aug 04 2022 11:59PM.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Line 106: Please add ‘…and the study protocol is available in the Supporting Information (S1 File).

Line 58: Please add ‘The findings suggest that…’.

Line 60: Please add ‘…of infection prevention measures…’.

S21 Table: Please add the corresponding footnote for the flag (^), which appears to be missing.

Your response to Reviewer #2’s previous comment does not address (or respond to) their request ‘Brief mention might be made about (over-)stratification in the manuscript if considered appropriate.’ Please either include this or provide a response to explain why you have chosen not to include this.

S3 checklist title/legend: Please update this to ‘RECORD’ rather than ‘REPORT’ checklist.

Lines 368-370: Please update to ‘There was also a change between PICU admissions (IRR 1.13, 95% CI 1.00 – 1.28, p=0.045) and admissions in general. Though the comparison remained statistically significant, the association was weaker.’

28 Jul 2022

Dear Dr Nijman,

On behalf of my colleagues and the Academic Editor, Dr Zulfiqar Bhutta, I am pleased to inform you that we have agreed to publish your manuscript "Presentations of children to emergency departments across Europe and the COVID-19 pandemic: a multinational observational study." (PMEDICINE-D-22-00899R4) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

When making the formatting changes please also update your Data Availability Statement to include the full doi URL at which your study data can be found (i.e., https://doi.org/10.14469/hpc/10685).

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Callam Davidson

Associate Editor

PLOS Medicine