Diagnostic and prognostic value of the optic nerve sheath diameter with respect to the intracranial pressure and neurological outcome of patients following hemicraniectomy

BACKGROUND: In cases showing cerebrospinal fluid (CSF) redistribution as a compensatory mechanism in acute intracranial hypertension, the optic nerve sheath diameter (ONSD) can be used to estimate intracranial pressure (ICP). However, it remains unclear whether the ONSD can be applied in patients with skull defects after a craniectomy, because the primary injury or surgical craniectomy may alter the dynamics of the CSF circulation or structure of the optical nerve sheath. This study explored the value of the ONSD in patients after a hemicraniectomy. METHODS: This prospective observational study enrolled patients after a hemicraniectomy. All patients underwent invasive ICP monitoring and ocular ultrasound within 6 h postoperatively. We followed the patients for 6 months and evaluated them using the Glasgow Outcome Score (GOS), classifying the outcome as favorable (GOS 4–5) or unfavorable (GOS 1–3). We evaluated the ONSD in both according to the ICP and neurological outcome. RESULTS: Of the 33 enrolled patients, 20 (60.6%) had an unfavorable outcome at 6 months. Disagreement was seen in the ONSD measurements between the eyes [craniectomy side (ONSDips) and opposite side (ONSDcon)]. The intraclass correlation coefficient between ONSDips and ONSDcon was 0.745 (p < 0.001). ONSD had no significant correlation with ICP in Spearman correlation analysis (ONSDips r = 0.205, p = 0.252; ONSDcon r = 0.164, p = 0.362). Receiver operator characteristic (ROC) curve analysis revealed that the GCS, Helsinki computed tomography (CT) score, pupil reaction, and ONSDcon measured after the craniectomy were significantly associated with a poor outcome. ONSDcon > 5.5 mm predicted a poor outcome, with an area under the ROC curve of 0.717 (95% confidence interval, 0.534–0.860, p = 0.02), 70% sensitivity, and 69.2% specificity. CONCLUSIONS: After hemicraniectomy, the ONSD measured on ultrasound was unreliable for evaluating ICP, but showed potential prognostic value for a poor neurological outcome.