Objective: Trauma accounts for a large proportion of childhood deaths. No data exist about injury patterns within paediatric trauma in the UK. Identification of specific high-risk injury patterns may lead to improved care and outcome.
Methods: Data from 24 218 paediatric trauma cases recorded by the Trauma Audit and Research Network (TARN) from 1990 to 2005 were analysed. Main injury, injury patterns and outcome were analysed. Mortality at 93 days’ post-injury was the major outcome measure.
Results: Limb injuries occurred in 65.0% of patients. In infants 81.4% of head injuries were isolated, compared with 46.5% in 11–15-year-old children. Thoracic injuries were associated with other injuries in 68.4%. The overall mortality rate was 3.7% (n = 893). Mortality decreased from 4.2% to 3.1%; this was most evident in non-isolated head injuries. It was low in isolated injuries: 1.5% (n = 293). In children aged 1–15 years the highest mortalities occurred in multiple injuries including head/thoracic (47.7%) and head/abdominal injuries (49.9%). Having a Glasgow Coma Scale of <15 on presentation to hospital was associated with a mortality of 16%.
Conclusions: Differences in injury patterns and mortality exist between different age groups and high-risk injury patterns can be identified. With increasing age, a decline in the proportion of children with head injury and an increase in the proportion with limb injury were observed. This information is useful for directing ongoing care of severely injured children. Future analyses of the TARN database may help to evaluate the management of high-risk children and to identify the most effective care.
Statistics from Altmetric.com
Trauma still accounts for a large proportion of childhood deaths in the UK, ranging from 16% in 1–4 year olds to 40% in teenagers 15–19 years old.1 It is the leading cause of death between the ages of 1 and 44 years.2 Information is available about the causes of injury, but no UK data exist about injury patterns in paediatric trauma. It has been shown that improved hospital care results in lower mortality3 and that care is best delivered at a paediatric centre.4 Therefore the identification of high-risk injury patterns may lead to improved care and ultimately further improvements in outcome in children admitted to hospital with major trauma. This will be of value for the proposed trauma centres in London.5 We analysed paediatric trauma cases recorded by the national database of the Trauma Audit and Research Network (TARN) with the aim of identifying the most common injuries, injury patterns, temporal trends and related mortality in different age groups.
MATERIAL AND METHODS
TARN was established in 1989 and collects anonymised data from injured patients presenting to 109 (60%) of trauma-receiving hospitals in England and Wales. At present TARN membership is voluntary and the 60% of hospitals are an appropriate mix of teaching, general, urban, suburban and rural hospitals. Injured children are eligible for inclusion, if their trauma admission is greater than 72 h or results in high dependency unit/intensive care unit care, interhospital transfer or death at any time up to 93 days’ post-initial hospital presentation. Children who present with uncomplicated closed limb injuries, closed facial injuries, simple skin injuries and superficial or partial burns less than 10% body surface area are excluded. Analysis on anonymised TARN data for research is permitted through the Patient Information Advisory Group.
For this analysis only children with moderate and severe injuries as defined by the Abbreviated Injury Scale (AIS) ⩾2 were included. The AIS coding system uses an internationally acknowledged dictionary to assign a severity score for the injury sustained; the scores range from 1 (minor) to 6 (fatal).6 Injuries were categorised into five main body areas: head, thorax, abdomen, limbs and a group encompassing face, neck, spine and burns (FNSB) depending on each child’s most severe injury.
We studied these injury groupings with respect to four age groups (<1 year, 1–5 years, 6–10 years, 11–15 years) and four time periods (1990–1993, 1994–1997, 1998–2001, 2002–2005).
We analysed whether the major injury was an isolated injury or combined with significant injury in another body area and whether more than two body areas were significantly injured. In order to correlate mortality to injury patterns we excluded children who presented to a TARN hospital but were then transferred to a hospital that did not participate in TARN and for whom we therefore did not have outcome data.
There were 28 252 paediatric cases on the TARN database between 1990 and 2005. The median age (interquartile range; IQR) of children presenting with serious injury in this series was 9.5 (4.9–13) years. For these children the most common mechanisms of injury were Road Traffic Collisions (RTCs) (40.9%), followed by falls (36.9%), assault or non-accidental injury (1.9%) and sport (0.2%). The median Injury Severity Score (IQR) of these children was 9 (9–10). On arrival at the Emergency Department most children were fully conscious (median Glasgow Coma Scale (GCS) 15(15–15)). GCS on initial presentation of 15 was associated with a mortality of 0.2%, but a GCS of <15 was associated with a mortality of 16% (p<0.001). As 4034 (14.3%) children were transferred to non-TARN hospitals they were excluded from further analysis; their median age, GCS and Injury Severity Score were not significantly different to that of the main group.
Twenty four thousand two hundred and eighteen children fell into the five main injury groups with AIS ⩾2 injuries. Limb injuries were most common. Injuries occurred in isolation in 84.3% (n = 20 426) of cases. Mortality was low overall: 3.7% (n = 893, 95% CI 3.4 to 3.9%) and 1.5% (n = 293, 95% CI 1.3 to 1.6%) in isolated injuries. The percentage of each injury and whether it was isolated is shown in table 1.
In the group of children with injuries in two main body areas, the results were as follows: non-isolated head injuries (n = 2603) were most commonly associated with limb injuries and occurred in 58.7% of patients (n = 1528, 95% CI 56.8 to 60.6%). Similarly limb injuries were most commonly associated with head injuries. Thoracic injuries in children 1–15 years old (n = 568) were most commonly associated with limb injuries, occurring in 48.8% (n = 277, 95% CI 44.7 to 52.9%).
Mortality by body region injured
The highest mortality rate for isolated injuries occurred in children with head injuries; this was unchanged over different time periods and in different age groups. In children with multiple injuries, defined as AIS ⩾2 in more than one body region, mortality was also highest when head injuries occurred. Within this group the highest mortalities occurred in multiple injuries including thoracic injuries (n = 538) at 49.2% (n = 267, 95% CI 45.0 to 53.4%) and in combined head and FNSB injuries (n = 881) at 23.2% (n = 188, 95% CI 20.4 to 26.0%). A high mortality rate occurred in thoracic injuries when associated with multiple other injuries (n = 411): mortality 8.8% (n = 36, 95% CI 6.1 to 11.5%). There were a small number of children with a combination of thoracic and abdominal injuries (n = 76); mortality in this group was 14.5% (n = 11, 95% CI 6.6 to 22.4%); these injuries were most likely to be seen in association with other injuries (n = 96) and had minimal effect on mortality. Combined head/abdominal injuries (n = 375) were also most likely to be associated with other injuries (n = 256). Associated injuries increased the mortality from 42.9% to 53.1%. Figure 1 shows the mortality associated with isolated and combined injuries for the major injury groups. Head injuries were associated with cervical spine injuries in 2.2% of children.
Analysis by age
The number of injuries in different age groups is illustrated in table 2. The total number of injuries for each age group represents the number of children with injuries and is therefore lower than the total number of injuries in each age group as some children had more than one injury.
In all age groups the most common injury was a limb injury. In infants, head injuries occurred in 42.9%, almost double the incidence in older children, but similar to the percentage of limb injuries. The only significant change in injury pattern with age was related to head injuries. In infants 81.4% of head injuries were isolated, again almost double the rate seen in the 11–15-year-old age group (p<0.001). Abdominal injuries were extremely rare in infants (n = 7). With increasing age, limb injuries became more prevalent and head injuries less prevalent.
Mortality was also analysed for the five main injury groups by age and is shown in table 3. Mortality associated with non-isolated head injuries decreased with increasing age but not significantly. The age group with the best outcome appeared to be the 6–10 years age group, with infants having the poorest outcome. There was a marked decrease in mortality from abdominal injuries outside of infancy, but this was not statistically significant as the numbers of infants with abdominal injuries were so low. The presence of further injuries associated with combined thoracic and head injuries increased mortality (p = 0.012) up to the age of 10 years from 34.5% (n = 29/84, 95% CI 24.3 to 44.7%) where further injuries were absent, to 50.4% (n = 117/232, 95% CI 44.0 to 56.8%). In children with combined head and thoracic injuries only, mortality showed a progressive increase with increasing age. In this group, infants (n = 19) had a mortality of 26.3% (n = 5, 95% CI 6.5 to 46.1%) and children 11–15 years old (n = 61) had a mortality of 49.2% (n = 30, 95% CI 36.7 to 61.7%) (p = 0.079).
Outside of infancy increasing age was associated with a decrease in mortality with combined head, abdominal and other injuries (20% 0–1, 64.4% 1–5, 60.3% 6–10, 46.4% 11–15). Abdominal injuries conveyed a higher mortality in the under 5 years age group: 9.4% (n = 14), compared with older children: 3.3% (n = 27); p = 0.02.
Analysis by year
Details of the numbers of children and mortality in the different time periods are shown in table 4. Overall mortality in the five main injury groups fell from 4.2% (n = 139, 95% CI 3.5 to 4.9%) in 1990–1993 to 3.1% (n = 179, 95% CI 2.7 to 3.6) in 2002–2005.
A statistically significant reduction in mortality throughout the study period was found in head associated with other injuries (chi square for trend p = 0.006). There was a significant decrease in the proportion of head injuries sustained balanced by a significant rise in limb, thoracic and abdominal injuries (chi squared for trend p<0.001). For non-isolated head injuries the percentage associated with spinal injuries increased over the study period, from 7.5% to 15.3% (chi square for trend p<0.001).
This study reports the first results about the most common injuries and injury patterns in severe paediatric trauma in the UK and is the largest such international study. It shows RTCs and falls to be the main cause of severe trauma in children with head injuries being the main cause of death even though most children presented to hospital with a GCS of 15. There were differences in injury pattern and mortality between different age groups. Injury patterns with a high risk can be identified. Having a GCS of 15 on arrival to hospital is a good prognostic indicator. Mortality is the main outcome measure in this paper; morbidity is an important factor in trauma and is poorly researched.7–9
The results of this study showed that the majority of children who sustain injuries in two body areas other than limbs have associated further injuries. The highest number of deaths occurred in children with head injuries; the highest mortality was in children with combined head/thorax or head/abdominal injuries increasing further when other injuries were also present. The majority of thoracic injuries were associated with other injuries. To date, there have been only a few studies published on the most common injuries in children.10–17 Two studies from American trauma centres compare the most common injuries in adults and children in pedestrian road traffic accidents.10 11 The most common injuries in 29 and 73 children (<16 years), respectively, were to the extremities (60% and 44%, respectively), head (57% and 32%, respectively), abdomen (12% and 10%, respectively) and thorax (8% and 3%, respectively). In this study fewer head (26%) and abdominal injuries (3.8%) were found but the other results were similar. Other studies reviewed only very selected groups of children (those with polytrauma12 or those who required intensive care treatment13 or died14). Their findings are not directly comparable to the results of our study. More recent studies have not selectively reviewed severe trauma, but analysed all paediatric trauma treated in emergency departments (infants <1 year of age15 and children 1–10 years16 and 0–18 years17). The third of these looked at discharge diagnoses and in 54% the children had an isolated injury with 17% being head injuries and 78% limb injuries. Our study showed a higher number of isolated injuries, but a similar percentage of head and limb injuries. Only two studies report on injury pattern in children18 19 Both reviewed fatal injuries in children who died after RTCs. It is difficult to compare injury patterns of children sustaining fatal RTCs with our study. Nevertheless, the reported injury combinations in fatal RTCs show similarities to the injury patterns in our study, with the highest mortality being for combined head and thoracic injuries. Two further studies have also highlighted an increased mortality when a child’s multiple injuries include thoracic injuries.12 20
Injury pattern by age
The analysis of injury patterns in different age groups requires large numbers of patients. Data used for this analysis were collected on a national database with approximately 60% of UK hospitals taking part in data collection. Some children were transferred to a non-TARN centre, no outcome data were available and they were excluded from the analysis. Even with data collected over 16 years the numbers of children, particularly with abdominal (n = 964) and thoracic injuries (n = 601), were relatively small. It is therefore difficult to give reliable results in these injury groups for different age groups. These numbers are higher than total numbers of patients included in most other studies. Eighty per cent of head injuries in infants were isolated and carried a higher mortality, with the number of associated injuries increasing in every age group. It is difficult to account for the improved prognosis in the 6–10-year age group.
Injury pattern by year
Overall mortality from childhood trauma is low and declined over the study period. This corresponds with a previous analysis, which showed reduced mortality after paediatric trauma with improved hospital care.3 The anatomical pattern of injuries in this denominator population of severe paediatric trauma (admission >3 days) showed a decline in the proportion of head injuries and an increase in the proportion of limb injuries over the timeframe studied. The rise in spinal injuries associated with head injuries is a concern. There was variability in actual numbers of injuries occurring in the different time periods; this is likely to be due to different numbers of centres submitting data and a variability in reporting.
This study is the first study of trauma patterns in children and the largest such international study.
Mortality from trauma is low and is decreasing.
Any decrease in GCS in the context of trauma causes a significant increase in mortality.
Infants have proportionately more head injuries than other age groups and increased mortality from isolated head injuries.
Children of different ages have differing injury patterns and differing mortalities.
Limb injuries have proportionately increased over the last 16 years and also with increasing age.
Thoracic injuries are increasing and are associated with a high mortality.
Highest mortality is seen in head injuries associated with either chest or abdominal injuries, either in isolation or with other injuries.
The number of spinal injuries related to head injuries is increasing.
We must continue to improve road, home and playground safety to reduce injury occurrence. Once injury has occurred there is much evidence on the optimal management of head injuries21 22 and we must strive to find optimal treatment for thoracic, abdominal and FNSB injuries. The relatively low numbers of these injuries makes this more difficult, but no less important. We have identified high-risk injury patterns; there is a need for further studies to evaluate the care of children with such injury patterns in order to identify areas for improvement.
We thank Professor DW Yates for advice on the study design and manuscript.
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.