Dual intracranial pressure monitoring in neuro-ICU patients : A retrospective observational study.

Intracranial pressure (ICP) monitoring is a cornerstone of the management of patients with acute brain injury. External ventricular drains (EVDs) and intraparenchymal pressure sensors are the two most widely used invasive monitoring modalities. Although both techniques are routinely employed in neurocritical care, discrepancies between simultaneously recorded ICP values may occur because of technical limitations, sensor drift, or physiological intracranial pressuregradients. In this study, we evaluated dual ICP monitoring using EVD and intraparenchymal pressure sensor. The primary objective was to assess the frequency of signal mismatches. The secondary objective was to identify associations with patient diagnosis and device type. This was a single-center, retrospective observational study, from July 2013 to January 2015 and from December 2017 to April 2019. We included ICU patients with traumatic brain injury (TBI) or aneurysmal subarachnoid hemorrhage (aSAH). All patients had dual ICP monitoring with an intraparenchymal pressure sensor and an EVD. Paired ICP measurements were collected every three hours during a 15-day observation period. Three categories of discordance were analyzed: E1: absolute difference > 6 mmHg between EVD and intraparenchymal measurements E2 : EVD-positive / intraparenchymal-negative threshold discordance (EVD > 20 mmHg and intraparenchymal ICP < 20 mmHg) E3: intraparenchymal-positive / EVD-negative threshold discordance (intraparenchymal ICP > 20 mmHg and EVD < 20 mmHg). A total of 313 patients (169 traumatic brain injury and 144 aneurysmalsubarachnoid hemorrhage) were included, providing 22,294 valid paired ICP measurements. Overall, 84% (n = 267) of patients experienced at least one E1 event, whereas 15.7% (n = 49) experienced at least one E2 event and 28.1% (n = 88) at least one E3 event. Agreement between ventricular and intraparenchymal measurements remained acceptable across pathologies and device types. However, discordances occurred more frequently during episodes of intracranial hypertension. Discordance between ventricular and intraparenchymal ICP measurements was not uncommon in this cohort of neurocritical care patients. These findings underscore the complexity of ICP interpretation and highlight the limitations inherent to all invasive monitoring techniques. Although dual monitoring may help identify discordant measurements in selected situations, the clinical significance of these discrepancies and their impact on patient management and outcomes remain to be established.