Optic nerve sheath diameter and pulsatility index for the diagnosis and follow-up in pediatric traumatic brain injury: a prospective observational cohort study.

ABSTRACT

PURPOSE: Invasive neuromonitoring could be difficult in children with traumatic brain injury (TBI). This study aimed to determine whether noninvasive intracranial pressure (nICP), calculated via pulsatility index (PI) and optic nerve sheath diameter (ONSD) had correlated with each other and patient outcome. METHODS: All moderate-severe TBI patients were eligible. Patients with a diagnosis of intoxication that did not affect the mental status or cardiovascular system were enrolled as controls. The PI measurements were routinely performed bilaterally on the middle cerebral artery. A software (QLAB's Q-Apps) was used to calculate PI, which further placed the ICP equation of Bellner et al. Linear probe with a 10 MHz frequency transducer to measure ONSD, which further placed the ICP equation of Robba et al. All measurements were performed by a point-of-care ultrasound certified pediatric intensivist under the supervision of a neurocritical care specialist, before and 30 min after a hypertonic saline (HTS) infusion for every 6 h when the patient's mean arterial pressure, heart rate, body temperature, hemoglobin, and blood CO2 levels were within normal ranges. The secondary outcome was the effect of hypertonic saline (HTS) on nICP. Delta-sodium values of each HTS infusion were calculated as a difference between pre- and post-measurements. RESULTS: Twenty-five TBI patients (200 measurements) and 19 controls (57 measurements) were included. Median nICP-PI and nICP-ONSD on admission were significantly higher in the TBI group (11.03 (9.98-12.63), p = 0.004, and 13.14 (12.27-14.64), p < 0.001, respectively). Median nICP-ONSD of severe TBI patients were higher than moderate TBI patients (13.58 (13.14-15.71) and 12.30 (9.83-13.14), respectively, p = 0.013). The median nICP-PI was the same across the type of injury (falls and motor vehicle accidents), while the median nICP-ONSD of the motor vehicle accident group was higher than falls. The first nICP-PI and nICP-ONSD measurements in PICU and admission pGCS were negatively correlated (r = - 0.562, p = 0.003 and r = - 0.582, p = 0.002, respectively). The mean nICP-ONSD during the study period and admission pGCS and GOS-E peds score significantly correlated. However, the Bland-Altman plots showed significant bias between the two methods of ICP except after 5th dose of HTS. All nICP values significantly decreased in time, and it was most obvious after the 5th dose of HTS. No significant correlations were found between delta sodium levels and nICP. CONCLUSION: Noninvasive estimation of ICP is helpful for the management of pediatric severe TBI patients. nICP driven by ONSD is more consistent with clinical findings of increased ICP but not useful as a follow-up tool in acute management because of slow circulation of CSF around the optic sheath. The correlation between admission GCS scores and GOS-E peds score favors ONSD as a good candidate for determining disease severity and predicting long-term outcomes.