Evaluation of Optic Nerve Sheath Diameter and Transcranial Doppler As Noninvasive Tools to Detect Raised Intracranial Pressure in Children

Abstract

Objectives: To compare the diagnostic accuracy of the ultrasonography-guided optic nerve sheath diameter with transcranial Doppler-guided middle cerebral artery flow indices against the gold standard invasive intraparenchymal intracranial pressure values in children. Design: A single-center prospective cohort study. Setting: PICU of a tertiary care teaching hospital in North India. Patients: Eligible children (2–12 yr) are admitted to ICU and are undergoing intracranial pressure monitoring using an intraparenchymal catheter. Observations with a parallel measured intracranial pressure greater than or equal to 20 mm Hg were included as case-observations. Children with an invasive intracranial pressure of less than or equal to 15 mm Hg were taken as neurologic-control-observations and healthy children served as healthy-control-observations. Interventions: The horizontal and vertical diameters of the optic nerves were measured, and averages were calculated and compared. Middle cerebral artery flow indices (pulsatility index and resistive index) were measured bilaterally and averages were calculated and compared in the three groups. Twenty-two measurements of optic nerve sheath diameter were assessed by two different observers in quick succession for interrater reliability. Measurements and Main Results: A total of 148 observations were performed in 30 children. Four observations were excluded (intracranial pressure between 16 and 19 mm Hg). Of the 144 observations, 106 were case-observations and 38 were neurologic-control-observations. Additional 66 observations were healthy-control-observations. The mean optic nerve sheath diameter was 5.71 ± 0.57 mm in the case-observations group, 4.21 ± 0.66 mm in the neurologic-control-observations group, and 3.71 ± 0.27 mm in the healthy-control-observations group (p < 0.001 for case-observations vs neurologic-control-observations/healthy-control-observations). The mean pulsatility index in case-observations was 0.92 ± 0.41 compared with controls 0.79 ± 0.22 (p = 0.005) and the mean resistive index was 0.56 ± 0.13 in case-observations compared with 0.51 ± 0.09 (p = 0.007) in controls (neurologic-control-observations and healthy-control-observations). For the raised intracranial pressure defined by intracranial pressure greater than or equal to 20 mm Hg, the area under the curve for optic nerve sheath diameter was 0.976, while it was 0.571 for pulsatility index and 0.579 for resistive index. Furthermore, the optic nerve sheath diameter cutoff of 4.0 mm had 98% sensitivity and 75% specificity for raised intracranial pressure, while the pulsatility index value of 0.51 had 89% sensitivity and 10% specificity by middle cerebral artery flow studies. The sensitivity and specificity of 0.40 resistive index value in the raised intracranial pressure were 88% and 11%, respectively. Kendall correlation coefficient between intracranial pressure and optic nerve sheath diameter, pulsatility index, and resistive index was 0.461, 0.148, and 0.148, respectively. The Pearson correlation coefficient between two observers for optic nerve sheath diameter, pulsatility index, and resistive index was 0.98, 0.914, and 0.833, respectively. Conclusions: Unlike transcranial Doppler-guided middle cerebral artery flow indices, ultrasonography-guided optic nerve sheath diameter was observed to have a good diagnostic accuracy in identifying children with an intracranial pressure of greater than or equal to 20 mm Hg.

Predictive Value of Optic Nerve Sheath Diameter for Diagnosis of Intracranial Hypertension in Children With Severe Brain Injury

“…The ONSD values were significantly higher in patients compared to those of controls. This result was already found in a usual way, both in adults ( 8 , 19 , 22 , 23 ) and in children ( 1 , 13 , 24 – 26 ).…”

“…A recent paediatric meta-analysis has reviewed the diagnostic performances of ONSD and found sensitivities between 39 and 100% and specificities between 22 and 100% ( 9 ). Some studies found a good correlation between ONSD and ICP, such as Kerscher ( 10 ) or Padayachy ( 11 ), whereas others do not find this correlation, such as Biggs ( 12 ) and Sharawat ( 13 ). It is noteworthy that few paediatric studies used invasive measurement of ICP ( 10 , 11 , 13 , 14 ).…”

“…Some studies found a good correlation between ONSD and ICP, such as Kerscher ( 10 ) or Padayachy ( 11 ), whereas others do not find this correlation, such as Biggs ( 12 ) and Sharawat ( 13 ). It is noteworthy that few paediatric studies used invasive measurement of ICP ( 10 , 11 , 13 , 14 ).…”

The Role of Transcranial Grayscale and Doppler Ultrasound Examination in Diagnosis of Neonatal Hypoxic-ischemic Encephalopathy

“…In line with these findings, multiple recent studies evaluated the diagnostic and prognostic validity of TCD in varied cases, where Cardim et al [27] demonstrated agreement between invasive and noninvasive ICP measurements using TCD and optic nerve sheet diameter ultrasound and suggested the potential role for a non-invasive technique for detection of ICH. Thereafter, Sharawat et al [28] assured the use of TCD as a non-invasive diagnostic modality for increased ICP and using Pearson correlation coefficient found the coefficients for optic nerve sheath diameter, pulsatility index, and resistive index were 0.98, 0.914, and 0.833, respectively. Moreover, Roldán et al [29] conducted a meta-analysis for the most effective techniques for the non-invasive multimodal monitoring for traumatic brain injury and concluded that the near-infrared spectroscopy and TCD are the two most prominent and widely used technologies for non-invasive monitoring in TBI.…”

Transcranial Doppler Ultrasound, a Review for the Pediatric Intensivist

“…Therefore, limited data suggest the TCD-derived PI may be helpful in the detection of intracranial hypertension early in the course of severe TBI. Conversely, the PI was recently investigated within a heterogenous cohort of 30 children with ICP monitors placed (67% infectious, 10% metabolic, 17% immune-mediated, and 7% other) [ 30 ]. Although children with intracranial hypertension had an elevated PI (0.92 ± 0.41) compared to those measured points without intracranial hypertension (0.69 ± 0.27, p < 0.001), the suggested cut-off for a PI value reflecting ICP ≥ 20 mmHg was 0.51 with 89% sensitivity, 10% specificity, 50% positive predictive value, and 47% negative predictive value.…”