Traumatic brain injury is a major contributor to mortality and acquired morbidity in children across the world (1). In Germany, 458,844 children <18 years were hospitalized with TBI between 2014 and 2018, resulting in an incidence rate of 687/100,000 child years (CY) and mortality of 0.67/100,000 CY (0.1%) (2). Age <2 years, GCS <6, hypothermia, hyperglycemia, and coagulation disorders are associated with increased mortality (3). Due to the brain's low tolerance towards hypoxia, optimizing the timeframe between insult, diagnostics and targeted neuroprotective measures is particularly crucial to reduce long-term sequelae and mortality.

Adherence to specific therapy recommendations has reduced mortality and improved functional outcome after severe traumatic brain injury (sTBI) (4). Immediate protective intubation, normalization of vital parameters (5), and basic neuroprotective measures are the cornerstones of sTBI management (6). Additionally, invasive intracranial pressure (ICP) monitoring is suggested for comatose [Glasgow Coma Scale (GCS) value ≤8] or continuously sedated patients by international guidelines and in the 2022 update of the German national guideline with strong consensus but low level of evidence (4, 6, 7). The previous German guideline, which was in effect since 2011 and expired in 2017, acknowledged the potential benefits of ICP monitoring but declared it as individual case decision (8) (Figure 1).

In spite of the tentative national guideline recommendation, a survey in 2017 among 29 German pediatric intensive care units (PICU) found that 25 (86%) units declared to use ICP monitoring always or often in patients with severe TBI (9). Our study aimed to verify how many children <10 years with sTBI who fulfilled criteria for ICP measurement in fact received this treatment and describe the characteristics of these patients via a prospective nationwide surveillance study in Germany.

During the study period, 217 cases <10 years with GCS ≤8 or neurosurgery due to TBI were reported to ESPED. One hundred two cases had an initial GCS ≤8, of which 37 (36%) received ICP monitoring (Table 1). In 15 of 37 (41%) cases, information on the type of inserted ICP monitor was provided (n = 11 (73%) parenchymal probe, n = 4 (27%) external ventricular drainage). Almost all included children had been previously healthy (91%). The monitored patients were older, had a more balanced gender distribution and lower admission temperature. Regarding admission weight and blood pressure no differences were observed. Of infants <1 year, only 27 (11%) received ICP monitoring (data not shown). Sixty-nine children (68%) were initially on invasive ventilation and 11 (11%) received vasopressors/inotropes in the emergency unit. In the group with ICP monitoring, the primary care was performed more frequently in the highest level trauma centers (Table 1). These patients had noticeably more often GCS levels ≤5 compared to patients without ICP monitoring (Figure 2).

This study provides an overview of sTBI management in children <10 years between 2019 and 2022 in German children's hospitals. ICP monitoring was performed in approximately one third of cases. Mortality and adverse functional outcome were more frequent in the ICP monitored group, likely reflecting differences in baseline characteristics regarding patient age, intracranial pathologies, and treatment center.

ICP monitored children were older than non-monitored children. Particularly, infants below one year of age rarely received ICP monitoring, possibly due to the availability of transfontanellar ultrasound to monitor ventricular width, cerebral edema and perfusion. Of note, ICP monitored patients had lower initial GCS scores, presented with cerebral edema more often, were more often invasively ventilated, received vasopressors/inotropes and frequently suffered from multiple injuries. Thus, the indication to insert an ICP monitor seems to be driven by considerations beyond the state of consciousness, despite recommendations of current national and international guidelines (4, 6).

However, the performed ICP monitoring practice in Germany between 2019 and 2022 did not correspond to international sTBI guidelines that were in effect at that time. The German national TBI guideline did not clearly recommend ICP monitoring for patients with GCS ≤8, in contrast to the 2022 update (4, 8). This low monitoring rate is of interest, because it clearly deviates from the self-reported ICP monitoring rate of up to 86% in a nationwide survey (9). This reflects a gap between theoretical knowledge on the indication and the actual implementation.

Even though the guidelines have strong expert consensus regarding their recommendations on ICP monitoring in sTBI, the level of evidence underlying these recommendations is low (4, 6, 7) and it is still under debate whether ICP monitoring actually improves outcomes. While positive effects of diversion of cerebrospinal fluid on outcomes have been reported (12–14), other observational studies found no effect or even point in the opposite direction (15, 16). In the absence of sufficiently powered prospective randomized controlled trials, it remains unclear whether these findings truly reflect adverse effects of the interventions or the impossibility to completely adjust for confounding factors. The low level of evidence may cause uncertainty upon the decision whether to insert ICP monitoring or not, which may be driven by additional considerations, such as pending surgical or neurosurgical procedures and accompanying conditions.

It seems reasonable and has been reported in adults, that ICP monitored patient receive interventions to reduce intracranial pressure more often, both medical and surgical (15). This finding was also observed in our study and points towards important further considerations: ICP monitoring alone cannot improve patient outcomes, but has to be accompanied by prompt and adequate interventions to control intracranial hypertension (17). Further, potential complications such as infection, hemorrhage, mechanical disruption, malfunction and misplacement of the device have to be considered (18, 19). In clinical practice, non-invasive techniques to indirectly monitor ICP e.g., near-infrared spectroscopy, transcranial and transfontanellar Doppler ultrasound, fundoscopy and magnetic resonance imaging (MRI) are available to complement invasive ICP monitoring and may impact both the decision to insert ICP monitoring and the conservative management in the PICU.

This study is limited by the surveillance design, which covers only children's hospitals and relies on the compliance of participating centers. The small case number can be partially explained by the rarity of the condition, but it is evident that only a fraction of cases occurring during the study period were captured. With an estimated 8%, the completeness of this survey is at the lower range of literature reports (20) and lower than previously reported for ESPED surveys (21). Also, no data on treatment and device associated complications were available due to missing specific questions. No information was collected on the decision-making process of treating physicians to control for indication bias and information in the type of inserted ICP monitor was missing in more than half of cases.

ICP monitoring practice in severely head-injured children in Germany does not comply with current international guidelines. The reasons remain unclear but a contradictory expired national guideline may have hindered adherence to international guidelines. This guideline discrepancy is removed with the 2022 update of the German national TBI guideline, and a future study should elucidate whether subsequent changes in ICP monitoring practice occur. Educational interventions to promote the content of the updated guideline may further increase awareness of and compliance with current recommendations. For the future, randomized controlled studies to verify whether ICP monitoring in children with sTBI is beneficial and to untangle the effects of monitoring and ICP directed interventions to control ICP in monitored patients are indispensable.