Introduction: Current guidelines suggest the use of intracranial pressure (ICP) monitors in pediatric patients with severe traumatic brain injury (TBI). Surveys indicate that ICP monitoring practices vary by hospital. We aimed to determine whether ICP monitoring is associated with decreased mortality among pediatric patients with severe TBI. Methods: This retrospective cohort extracted clinical and demographic data from the Virtual Pediatric Systems (VPS) database. Included patients were < 18 years with STAR codes for TBI and lowest recorded GCS of ≤ 8 between 2010-2024. Patients from centers without ICP monitors were excluded. ICP monitoring practices and patient demographics were described using summary statistics. 1:1 nearest-neighbor propensity matching with clustering on hospital was performed and the average effect of ICP monitor placement on mortality was reported. Results: There were 9,335 included admissions across 96 PICUs between 1/1/2010-12/31/2024 in which 2,449/9,335 (26.2%) had ICP monitors. Center-level ICP monitoring ranged from 0.5% to 61.5%. There was no correlation between center-level ICP monitoring and unadjusted center-level mortality (p = 0.41). Patients with ICP monitors were older (8 3-14 versus 6 2-13 years, p < 0.001) and had lower GCS (3 3 – 6 versus 6 3 – 7, p < 0.001), a lower proportion of reactive pupils (60.2% versus 68.4%, p < 0.001), and a higher PIM2 probability of death (4.9% 3.3% – 23.1% versus 4.1% 3.0% – 34.3%, p < 0.001). 4,188/9,254 (45.2%) encounters were matched. The overall standardized mean difference before and after matching was 0.64 and 0.02, respectively. The marginal risk difference for death in patients treated with ICP monitors was -2.9% (95% CI -0.7%, -5.0%), and the marginal risk ratio was 0.91 (95% CI 0.84, 0.98). However, after matching, patients treated with ICP monitors had a higher proportion of new neurologic disability (Pediatric Cerebral Performance Category 2-5; 59.6% 407/683 versus 33.3% 234/701, p < 0.001). Conclusions: ICP monitoring was associated with reduced mortality in severe TBI in this study, but a higher proportion of new neurologic disability. Future studies should examine heterogeneity of treatment effect in subpopulations by age, mechanism, and imaging findings.
Johnson-Nibling et al. (Sun,) studied this question.