Assessment of Cerebral Autoregulation Using Invasive and Noninvasive Methods of Intracranial Pressure Monitoring

Background: Pulse amplitude index (PAx), a descriptor of cerebrovascular reactivity, correlates the changes of the pulse amplitude of the intracranial pressure (ICP) waveform (AMP) with changes in mean arterial pressure (MAP). AMP relies on cerebrovascular compliance, which is modulated by the state of the cerebrovascular reactivity. PAx can aid in prognostication after acute brain injuries as a tool for the assessment of cerebral autoregulation and could potentially tailor individual management; however, invasive measurements are required for its calculation. Our aim was to evaluate the relationship between noninvasive PAx (nPAx) derived from a novel noninvasive device for ICP monitoring and PAx derived from gold standard invasive methods.

Methods: We retrospectively analyzed invasive ICP (external ventricular drain) and non-invasive ICP (nICP), via mechanical extensometer (Brain4Care Corp.). Invasive and non-invasive ICP waveform morphology data was collected in adult patients with brain injury with arterial blood pressure monitoring. The time series from all signals were first treated to remove movement artifacts. PAx and nPAx were calculated as the moving correlation coefficients of 10-s averages of AMP or non-invasive AMP (nAMP) and MAP. AMP/nAMP was determined by calculating the fundamental frequency amplitude of the ICP/nICP signal over a 10-s window, updated every 10-s. We then evaluated the relationship between invasive PAx and noninvasive nPAx using the methods of repeated-measures analysis to generate an estimate of the correlation coefficient and its 95% confidence interval (CI). The agreement between the two methods was assessed using the Bland-Altman test.

Results: Twenty-four patients were identified. The median age was 53.5 years (interquartile range 40-70), and intracranial hemorrhage (84%) was the most common etiology. Twenty-one (87.5%) patients underwent mechanical ventilation, and 60% were sedated with a median Glasgow Coma Scale score of 8 (7-15). Mean PAx was 0.0296 ± 0.331, and nPAx was 0.0171 ± 0.332. The correlation between PAx and nPAx was strong (R = 0.70, p < 0.0005, 95% CI 0.687-0.717). Bland-Altman analysis showed excellent agreement, with a bias of - 0.018 (95% CI - 0.026 to - 0.01) and a localized regression trend line that did not deviate from 0.

Conclusions: PAx can be calculated by conventional and noninvasive ICP monitoring in a statistically significant evaluation with strong agreement. Further study of the applications of this clinical tool is warranted, with the goal of early therapeutic intervention to improve neurologic outcomes following acute brain injuries.