Violence-Related Stabbings and Gunshots Dominate Rare Paediatric Trauma Deaths at Swedish Level 1 Centre

When a child triggers a full trauma-team activation at Karolinska University Hospital in Stockholm, the odds of survival are overwhelmingly in their favour. Between 2016 and 2025, the hospital recorded 1,792 such activations involving patients aged 0 to 17 years, and only 28 of those children died within 30 days of injury. That corresponds to a 30-day case fatality rate of 1.6 percent among activated trauma patients, a figure that underscores just how uncommon death has become in a well-resourced paediatric trauma system. Yet the small number of deaths is precisely what makes them worth dissecting, because each one represents a failure point that the system rarely gets the clinical exposure to learn from.

A new retrospective descriptive case series published in BMC Pediatrics by Ammar Yaqoob of Karolinska Institutet and Shahzad Akram of Karolinska Institutet and the Trauma Centre Karolinska examines every one of those 28 deaths in detail. Drawing on the Karolinska Trauma Registry, the study identifies children who activated the trauma team, were entered into the registry, and subsequently died within the 30-day window. The researchers then characterised the mechanisms, injury patterns, and recorded causes of death, producing one of the most granular portraits available of how children die from trauma in a modern Scandinavian setting.

The demographic profile of the cohort is striking in itself. The 28 patients comprised 16 males and 12 females, with a median age of 15 years, placing most victims squarely in adolescence. Critical injury severity, defined as a New Injury Severity Score of 25 or higher, was documented in 25 of the 26 cases where the score was available. In practical terms, nearly every child who died had sustained injuries so severe that survival was anatomically improbable from the outset, a pattern consistent with the notion that in mature trauma systems, mortality concentrates among the most catastrophically injured rather than among patients with potentially survivable wounds.

The mechanism of injury tells a more unsettling story. Penetrating trauma accounted for the majority of deaths, 15 of 28 cases, mainly stabbings and gunshot wounds. Assault was recorded in half of all cases, and prehospital cardiac arrest occurred in 17 patients, meaning that for well over half of the cohort, the heart had already stopped before definitive care began. In a country often perceived as insulated from serious interpersonal violence, the finding that violence-related penetrating injury dominates paediatric trauma deaths at Sweden’s primary level 1 trauma centre challenges common assumptions and points to a burden of violence among children and adolescents that clinical registries are uniquely positioned to reveal.

Injury patterns followed a clear hierarchy. Isolated head injury was the most common presentation, occurring in 12 of 28 cases, or 42.9 percent, followed by thoracoabdominal injury in 4 cases, or 14.3 percent. This distribution aligns with the two dominant physiological causes of death recorded in the series: haemorrhagic shock, responsible for 14 deaths, or 50.0 percent, and traumatic brain injury, responsible for 12 deaths, or 42.9 percent. Cause of death could not be determined in the remaining 2 cases, or 7.1 percent. The near-even split between bleeding and brain injury reflects the fundamental anatomy of fatal trauma, where the torso’s vascular structures and the brain’s intolerance of mechanical force represent the two pathways that modern medicine struggles most to interrupt once damage is done.

The technical significance of these findings lies partly in what they reveal about the limits of trauma-care improvement. Haemorrhagic shock deaths, for instance, are the principal target of advances in prehospital blood product delivery, damage-control surgery, and massive transfusion protocols. That half of the deaths in this series still stemmed from exsanguination, often after prehospital cardiac arrest, suggests that the decisive events frequently occur before the patient reaches the hospital. Traumatic brain injury deaths, meanwhile, remain largely refractory to post-injury interventions, driving prevention efforts toward helmets, vehicle safety, and, in this cohort’s case, violence reduction.

The authors are careful to frame the study’s limitations with precision. The reported case fatality rate applies only to trauma-call activations captured in the registry and should not be interpreted as mortality among all injured children in the region. Children who died at the scene without triggering an activation, or who were treated at other facilities, fall outside this denominator. This distinction matters enormously for interpretation: the 1.6 percent figure describes what happens when the most severely injured children in the region do reach Karolinska’s trauma team, not the overall risk of death from paediatric injury in Stockholm.

What the series does offer is a template for system-level learning in environments where paediatric trauma deaths are rare. The authors argue that detailed case-level review can support clinical audit, service evaluation, and prevention strategies targeting interpersonal violence among children and adolescents. In a mature trauma system, the marginal gains no longer come from faster CT scanners or better intensive care alone, but from understanding the social and clinical circumstances that precede each death. When assault features in half of paediatric trauma deaths, the trauma centre becomes, in effect, a surveillance point for violence that public health systems must then act upon.

The study also highlights the value of long-term registry infrastructure. Ten years of prospective data collection at a single centre, spanning 1,792 activations, allowed the researchers to assemble a cohort large enough to characterise despite the rarity of the outcome. Ethical approval was granted by the Swedish Ethical Review Authority, and all data were anonymised, with consent requirements waived because the analysis was registry-based. This kind of anonymised, registry-driven methodology is increasingly recognised as essential for studying rare outcomes in small populations, where individual case series from single centres remain the only realistic route to actionable insight.

For clinicians, the message is that penetrating injury and violence must now feature prominently in paediatric trauma preparedness, even in regions where blunt trauma from road traffic collisions has historically dominated planning assumptions. For policymakers, the series provides evidence that prevention strategies aimed at interpersonal violence among adolescents are not merely social programmes but trauma-prevention interventions with measurable mortality implications. And for the broader research community, the work demonstrates that when deaths are rare, the appropriate scientific response is not to abandon the question but to build the registries, sustain the follow-up, and dissect every case with the rigour that each of the 28 children in this series deserves.

Subject of Research: Paediatric trauma mortality and violence-related penetrating injury at a level 1 trauma centre

Article Title: Paediatric trauma mortality at a level 1 trauma centre: a retrospective case series at Karolinska University Hospital, Stockholm

Article References: Yaqoob, A., & Akram, S. (2026). Paediatric trauma mortality at a level 1 trauma centre: a retrospective case series at Karolinska University Hospital, Stockholm. BMC Pediatrics, 26(1), Article 845. https://doi.org/10.1186/s12887-026-07618-4

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07618-4

Keywords: paediatric trauma, trauma mortality, penetrating injury, interpersonal violence, injury severity, case series, Paediatric, trauma, mortality, level, centre, retrospective