Pediatric Trauma in The Netherlands: Incidence, Mechanism of Injury and In-Hospital Mortality

Background In the Netherlands, there are no specialized or certified pediatric trauma centers, especially for severely injured children. National and regional agreements on centralization of pediatric trauma care are scarce. This study aims to describe the incidence, injury mechanism and in-hospital mortality of pediatric trauma in the Netherlands, as a prelude to the further organization of pediatric trauma care. Methods A retrospective cohort analysis of data from the Dutch National Trauma Registry in 2009–2018, concerning all children (0–16 years) hospitalized due to injury in the Netherlands. Results The annual number of admitted injured children increased from 8666 in 2009 to 13,367 in 2018. Domestic accidents were the most common cause of non-fatal injury (67.9%), especially in children aged 0–5 years (89.2%). Severe injury (injury severity score ≥  16) accounted for 2.5% and 74% of these patients were treated in level-1 trauma centers. In both deceased and surviving patients with severe injuries, head injuries were the most common (72.1% and 64.3%, respectively). In-hospital mortality after severe injury was 8.2%. Road-traffic accidents (RTAs) were the leading cause of death (46.5%). Conclusions Domestic accidents are the most common cause of injury, especially in younger children, whereas RTAs are the lead cause of fatal injury. Severe pediatric trauma in the Netherlands is predominantly managed in level-1 trauma centers, where a multidisciplinary team of experts is available. Improving the numbers of severely injured patients primarily brought to level-1 trauma centers may help to further reduce mortality.


Introduction
Since 2000, an average number of 80 children younger than 15 years old have died due to unintentional injuries each year in the Netherlands [1]. Regional studies showed that about 80% of the children that survive severe injuries in the Netherlands recover well, whereas 20% remain disabled to some extent after long-term follow-up [2,3].
In 1999/2000, the Dutch trauma care system was regionalized with dedicated trauma centers appointed per region [4]. As a result, the injury-related in-hospital mortality rate in children aged 13-18 years has decreased over the period 1996-2006. However, no change was observed for children younger than 12 years. Injury prevention strategies, such as road safety regulations, have also led to a decrease in mortality related to some types of childhood injuries. In 2018, 19 children had died in a road-traffic accident in the Netherlands; on average 46 less fatalities per year than in the period 1996-2000 [5], due to improved road and traffic safety over the last decade.
Trauma care in children is fundamentally different from that in adults, because of their anatomical and physiological dissimilarities [6]. In the Netherlands, there are no specialized pediatric trauma centers and severely injured children are treated in any of the level-1 trauma centers. This may affect the outcome of care, especially for severely injured children, since it has been proven difficult to provide optimal pediatric trauma care without adequate experience [7].
The objective of this study is to describe the current epidemiology of pediatric trauma in the Netherlands, since an up-to-date overview is lacking. By analyzing data from the Dutch National Trauma Registry (DNTR), we aimed to determine the incidence of pediatric injury, and to identify trends in cause of injury, injury patterns and outcome of pediatric patients who were admitted to a hospital after trauma. Insight in these data will help to further optimize pediatric patient trauma care as well as trauma prevention on a national level in the Netherlands.

Study design and population
This retrospective cohort study analyzed data obtained from the DNTR, which was established in 2007 and includes information from all Dutch hospitals of patients admitted due to injury [8]. All pediatric patients aged 0 to 16 years who were admitted to one of the Dutch hospitals due to injury between January 1st, 2009, and December 31st, 2018, were included.

Data collection
The primary outcome of this study was in-hospital mortality. Secondary outcomes were length of hospital stay (LOHS) and admittance to the intensive care unit (ICU). Other data included age, sex, type of hospital (trauma (level-1) vs. non-trauma center), injury characteristics (injury mechanism, injury type, abbreviated injury scale (AIS) score, injury severity score (ISS)) and vital signs on admission to the Emergency Department (ED) (Glasgow Coma Scale (GCS), respiratory rate (RR), systolic blood pressure (SBP)). In the DNTR, injuries are classified according to the AIS98 (up to 2015) and AIS08 (since 2015).

Definitions
Patients with an ISS C 16 and ISS C 25 were considered as severely and critically injured, respectively [9]. A GCS of 13-15 corresponds with mild head injury, 9-12 with moderate head injury and 3-8 with severe head injury. Normal values for systolic blood pressure and respiratory rate vary per age group [10]. For each age category, values below normal were considered low and values above normal were considered high.

Statistical analysis
To allow comparisons between age groups, children were divided in three age categories (0-5, 6-11, 12-16 years old). Categorical data were presented as number and percentage and compared using the Chi-square test. Injury severity scores were presented as median (interquartile range [IQR]) and compared using the Mann-Whitney test or Kruskal-Wallis test for independent samples. Additional analyses compared deceased and survived children and children treated in trauma vs. non-trauma centers. Missing data in this descriptive study were not accounted for in the analysis. The statistical analyses were performed in SPSS (IBM Corp. Released 2020. IBM SPSS Statistics for Windows, Version 27.0. Armonk, NY: IBM Corp).
No formal approval from the Institutional Review Boards was necessary, since only summarized data from the DNTR were obtained.

Total admitted pediatric trauma population
The annual number of admitted injured children increased from 8666 in 2009 to 13,367 in 2018 (Fig. 1). During the study period, a total of 129,840 children were hospitalized due to injury, of whom 23.1% were treated in a level-1 trauma center (Table 1). The largest group was the 0-5year-old group (44.9%). Over time, the proportion of hospitalized children aged 0-5 years increased and decreased for children aged 12-16 years (Fig. 2). Boys represented 58.8% of the population and formed the majority in all age categories (Table 1). Sex distribution remained unaltered over time.
Penetrating trauma was rare in each age group, but highest in the 12-16-year-old group (3.7%) ( Table 1). Domestic accidents were the most common injury mechanism in children aged 0-11 years. The proportion of hospitalized children due to sport-related accidents and RTAs increased with age ( Table 1). The distribution of other types of injury mechanisms was stable over time ( Fig. 3).
Children aged 0-5 years were the least severely injured ( Table 1). In 2015-2018, ISS was lower in all age categories compared to 2009-2014 due to the switch in the DNTR to AIS version 2008 in 2015 (Table 1).

Severely injured (ISS ‡ 16) pediatric trauma population
Characteristics per age group and injury mechanism.
Severe injury accounted for 2.5% of all pediatric injuryrelated hospital admissions and children aged 12-16 years formed the largest age group in this population (44%) ( Table 2). The majority in each age group were boys (62.8% in the total group). The age distribution of the severely injured children varied somewhat over time (Fig. 2).
RTAs were the leading injury mechanism in severely injured children aged 12-16 years (59.8%), while domestic accidents (41.3%) were most common in children 0-5 years (74.5%) and children aged 6-11 years (44.8%) ( Table 2). Over time, the proportion of children with severe injury due to RTAs increased and there was a decrease in the proportion of children with severe injury due to domestic accidents (Fig. 3).
Children  Table 2). More than one-third of the severely injured population was critically injured (ISS C 25). This proportion differed significantly by age group (Table 2) and injury mechanism (Table 3).
Head injuries were the most common severe injuries (AIS C 3) (65%, Table 2). Severe injuries to the head and external injuries were the most frequent in children 0-5 years, whereas severe injuries to the face, thorax, abdomen, spine and extremities were more common in older children (Table 2). Severe head injury was present in the majority of children with severe injury after RTAs, domestic accidents and violent accidents (Table 3). About 20% of the severely injured children had severe injuries to the thorax and abdomen, mostly after RTAs, sport accidents (abdomen), self-harm (thorax), and violent accidents ( Table 3).
Vital signs of severely injured children differed between the age groups. Age-specific low respiratory rates and high systolic blood pressure were more common in the youngest age group (Table 2). Children 12-16 years had the longest LOHS (median 6 days) and were admitted most often to the ICU (63%), whereas children 0-5 years had the shortest LOHS (median 4 days) and were less frequently admitted to the ICU (58%, Table 2).

Trauma center admission
Seventy-four percent of the severely injured children were treated in a level-1 trauma center (Table 2). Severely injured children who were injured in an RTA, had a lower level of consciousness (GCS \ 13), high RR, low SBP, higher ISS and severe injuries to the head, thorax, spine and lower extremities were more likely to be treated in a level-1 trauma center (Table 4). Consequently, severely injured children who were treated in a trauma center were longer hospitalized, were more frequently admitted to the ICU and had a higher mortality rate than those treated in a non-trauma center. Severely injured children who suffered a domestic or sport-related accident or had severe abdominal injury were more likely to be treated in a nontrauma center.

In-hospital mortality
In the study period 2009-2018, 262 (8.2%) of the children with severe injuries died in the hospital (Table 2), mostly after an RTA or domestic accident ( Table 5). Most of them were critically injured (ISS C 25: 89.2% in the years 2009-2014; 93.5% in the years 2015-2018) and were in a coma (93.9% with GCS of 3-8). Nearly half of them were hypotensive on hospital admission (49.3%) ( Table 5). Nonsurvivors more frequently had severe injuries to the head, neck, thorax and spine, as well as more severe external injuries (Table 5).

Discussion
The annual number of registered pediatric injury-related hospitalizations in the Netherlands increased by nearly 60% from 2009 to 2014 and remained stable thereafter. This trend may predominantly be explained as an artifact, because the hospital participation rate in the DNTR increased from 64% in 2007 to 100% in 2015 [11]. Since 2015, the registration of patients admitted due to injury in the DNTR is near complete. Children aged 0-5 years accounted for an increasing part of all hospital admissions due to injury, which may be caused by the fact that young children are more likely to be admitted as a precaution than older ones [12]. Domestic accidents were the most common injury mechanism in children. Prevention of such injuries remains a global challenge [13]. Parent-focused strategies and product and environmental modifications have been proven effective in preventing childhood injuries [14,15]. Modifications in the child's environment such as the installment  of stair gates, window guards and smoke alarms, childproofing caps on medication packaging, securing cabinets on walls and fencing a pool are some important first steps in home-injury prevention. Also, home visits by professionals can help improve the home environment and provide parents with information on safety equipment [16]. In the Netherlands, the Consumer and Safety Institute Vei-ligheidNL provides information to both parents and professionals regarding children's safety and injury prevention [17]. In cooperation with the Ministry of Public Health, they organize trainings, campaigns and behavioral interventions with national, regional and local authorities. Active participation of local authorities to enhance the role of this organization may further help spreading awareness on childhood injury and prevention. Implementation of preventative measures is challenging, however. In addition to information campaigns, legal requirement and enforcement of prevention measures that are proven effective are necessary where possible. In children aged 12-16 years, RTAs were the main cause of injury (60% of the cases). Children in this age group commonly bicycle unsupervised, making them more vulnerable in daily road traffic. Besides bicycling, use of light mopeds (two-wheeled vehicle) is quite common among Dutch 16-year-old children. Both modes of transportation do not require helmet use and make children at this age prone to head injury in RTAs. As observed in other studies, most RTA-related deaths in Dutch children involved children either as cyclists or pedestrians [18], while the majority of light-moped riders involved in fatal crashes between 2010 and 2015 were 16-17 years old [19].
Improvements in both road infrastructure and vehicle safety such as reducing the speed limit in areas where and at times when children and motorized traffic meet and introducing autonomous emergency braking systems with cyclist/pedestrian recognition will help to increase children's safety in traffic [20]. For years, the obligatory use of bicycle helmets for Dutch children has been a matter of discussion. Bicycle helmets are still not obligatory because of the fear that this will strongly decrease bicycling, while this is good exercise for children. In our study, severe head injury was present in 65% of the total group of severely injured children and in 72% of the severely injured children that died in the hospital. Weijermars et al. [21] calculated that if all children under 12 years were to wear a bicycle helmet in the Netherlands, this could annually prevent up to 5 casualties and approximately 200 cases with survivable severe road-traffic injuries.
Even though the majority of the severely injured children were treated in a trauma center, 26% were treated in a non-trauma center, suggesting that further centralization of pediatric trauma in the Netherlands is necessary. Centralization of Dutch trauma care started in 1999 and since then treatment of severely injured patients including children has shifted to level-1 trauma centers, resulting in a regional reduction of 50% in crude mortality of trauma patients [22]. The goal of further centralization of care is not only to further reduce mortality but also to optimize outcomes of severely injured children who survive by offering timely and optimal diagnostics and treatment. The American College of Surgeons Committee on Trauma (ACS-COT) recommends a maximum undertriage rate of 5% in polytrauma patients [23], while in the Netherlands, this is 10% [22]. In pediatric trauma, undertriage rates often exceed 20%, which is mostly caused by differences in triage protocols [24,25]. Currently, the Dutch National Protocol of Ambulance Services and Field Triage Decision Scheme triage protocols are used in pediatric field triage in the Netherlands. Recent studies revealed that both protocols cannot meet current sensitivity targets but also cannot accurately distinguish between low-risk and high-risk trauma patients [26,27]. Accordingly, the development of a sensitive and child-specific triage protocol is needed in order to prevent undertriage in severely injured children.
There are some limitations to take into consideration when interpreting the results of this study. The first limitation is the missing data for some of the cases. Variables with the highest percentages of missing values were mechanism of injury (39% in the total group and 46% in the severely injured group) and physiological variables (GCS 15%, RR 30%, SBP 14% in the severely injured patients), while the remaining variables were complete for 90-100% of the cases (Tables 1-2). Missing values for      [37,38]. Registration of initial observations of the injured patients must be further improved in order to increase data quality and reliability of the results. A second limitation is the update of the injury severity scoring tool, the AIS, from the 1998 to the 2005 version, which took place in 2015. According to Hsu et al. [39], the adaptation of the AIS-2005 version resulted in a significant reduction of severity of injury scores in the measurement of the same injuries. Since the aim of this study was to observe recent trends in the epidemiology of pediatric trauma, we included the years 2015-2018, for which the AIS-2005 version applied. By splitting the study period in two (2009-2014 and 2015-2018), we were able to present injury severity scores adapted to the AIS-2005 version separated from each other, in an attempt to minimize misleading outcomes of injury severity. Nonetheless, the results concerning injury severity should be interpreted with caution.

Conclusions
Domestic accidents are a significant cause of injury, especially in younger children, and RTAs are the lead cause of fatal injury. Severe pediatric trauma in the Netherlands is predominantly managed in trauma centers, where a multidisciplinary team of experts is available. Improving the numbers of severely injured patients primarily brought to level-1 trauma centers may help further reduce mortality. Also, injury prevention programs should be directed toward increasing safety measures in traffic. Injury mechanisms include road-traffic accidents (all accidents including a vehicle), self-harm accidents (including intentional intoxication and attempted suicide) and domestic accidents (accidents other than the above mechanisms of injury such as fall, accidental intoxication, accidental drowning and burns) AIS abbreviated injury scale; AIS C 3 severe injury; ED emergency department; IQR interquartile range a Calculation based on injury coding according to AIS98 b Calculation based on injury coding according to AIS08 * Respiratory rate and systolic blood pressure values are age-specific and classified according to normal values per age group