Counting the costs of trauma: the need for a new paediatric injury severity score

While the COVID-19 pandemic continues to dominate news headlines, little attention has been given to the announcement that the USA and UK will withdraw combat forces deployed to the so-called war on terror in Afghanistan. Protracted conflicts in the Middle East (eg, Syria, Yemen, and Afghanistan) and Africa (eg, Libya and the Sahel) continue unabated and under-reported, with well over 100 000 conflict-related deaths in 2020.¹ The toll on populations, especially children, is a shameful signal lost in the noise of other world events. The number of children living in conflict zones has doubled since 1991, such that by 2017, nearly a fifth of the world's children—420 million—lived in conflict-affected areas.² The UN report on children and armed conflict shows the annual number of children killed and injured in conflict has exceeded 10 000 since 2014, with a dramatic rise over the past 15 years.³

Children have been reported to make up 18% of patients seen in humanitarian hospitals in conflict zones and constitute 30% of seriously injured patients requiring surgical intervention.⁴ Despite the steadily growing need for paediatric expertise in medical care provided in conflict zones, several reports from military and humanitarian organisations describe a deficit in paediatric training, experience, and equipment on the ground.⁴

The first logical step to improve the quality of trauma care for children injured in conflict is to analyse the data from past conflicts and identify areas for improvement. However, even in 2021, this is not possible. Many conflicts occur in low-resource settings where already-fragile health systems are confronted with many acutely ill or injured patients. Under these conditions, close attention to data collection for retrospective analysis is rarely a high priority. One systematic review⁵ reported that fewer than 25% of publications describing conflict-related injuries among civilians reported minimum trauma data registry elements such as initial Glasgow coma scale score, the need for intensive care unit admission, or operative intervention.⁵ Trauma care in conflict is provided by a combination of different military, local, and non-governmental organisation actors, who do not share these data or make data available to outside researchers. Data collection is therefore non-standardised, severely limiting the utility of available data, which are inadequate to identify opportunities for system improvement.

Injury patterns can vary substantially across different conflicts. Several systems for standardising injury severity exist and are designed such that outcomes can be compared across different settings and patient populations. Most scoring systems take into account the distribution of injuries to specific anatomical regions, physiological derangements, or a combination of the two. Those incorporating physiological variables are often formulated as triage tools, with markers of haemodynamic instability to identify those in need of urgent resuscitation, alongside a simple quantification of the patient's physical injury pattern. Such composite systems are not widely used and more often anatomical scores are used in the literature on conflict-related injury.⁵ Additional challenges for accurate assessment of paediatric trauma in conflict are that many of the scores were based on anatomical injury descriptions in adult patients in civilian settings. For example, the injury severity score (ISS), the predominant injury scoring system used in large national trauma registries such as the UK's Trauma Audit and Research Network and the basis of several trauma scoring systems, was developed in 1974 and was based on analysis of 2128 patients injured in traffic collisions.⁶

The injuries inflicted in conflict are very different to those encountered in the civilian population upon which the ISS was based. Moreover, there are mounting data suggesting that both mechanism and location of injury might affect paediatric survival, such that children with identical ISS scores can have different expected mortality, since this data is not captured by the ISS. One study of 26 000 children showed that those with stab and gunshot wounds had a significantly increased risk of death compared with those injured in road traffic accidents, even after correcting for injury severity with the ISS.⁷ Data from a large paediatric trauma registry showed, for children with identical ISS scores, increasing severity of head injuries was strongly associated with increasing mortality but increasing injury severity in extremities made no difference.⁸ Blast injuries in children from improvised explosive devices, activated unexploded ordinance (eg, landmines or unexploded bombs), mortar fire, or suicide bombing can produce injuries of the type and severity rarely encountered in civilian trauma. Of the 17 401 child deaths from the Syrian conflict reported to the Violation Documentation Centre by 2016, 83% were a result of blast injury.⁹ In the original cohort from which the ISS was formulated, almost half of the patients had very low scores, with their most severe injury (score of 3) calculated to have an expected survival rate of over 97%, representing injuries such as uncomplicated open fractures, liver contusions, or pneumothoraces.⁶ Given this cohort contained low numbers of both children and severely injured patients, the data are likely to have limited applicability to severely injured children, such as those with complex blast injuries. Unsurprisingly, trauma scores such as the ISS do not perform well for paediatric trauma in conflict settings. The major challenge is that children with identical ISS scores might have very different expected mortality, which might be amplified in conflict injury and this shortcoming undermines the utility of the score for comparative analysis.

Several measures could be implemented to improve paediatric trauma care in conflict settings. The first might be greater coordination between different organisations providing health care in conflict zones, such that anonymised patient records could be pooled and analysed to improve patient care. This coordination could augment role-specific paediatric training packages that are standardised for health-care workers deployed to conflict zones. Second, to increase the quality of future data collection, a standardised minimum dataset to document essential information regarding the nature of patient injuries, care rendered, and outcomes could be introduced across health facilities.¹⁰ Finally, to facilitate more accurate mortality comparisons between areas and organisations, a paediatric-specific injury severity scoring system applicable to conflict and other austere settings could be developed that addresses some of the shortcomings of the ISS.

© 2021 akramalrasny/Shutterstock

We declare no competing interests. WS and HW are joint first authors.