Transcranial Doppler Pulsatility Index for Initial Management of Brain-Injured Patients

Abstract

To the Editor: We read with interest the article by Behrens et al,1 concerning the lack of accuracy of transcranial Doppler (TCD) to predict intracranial pressure (ICP). The authors measured the pulsatility index (PI) in 8 patients with idiopathic normal pressure hydrocephalus (INPH) and communicating CSF systems. The ICP-PI relationship was then explored through the lumbar infusion technique to get a range of ICP values between 0 and 50 mm Hg in each individual. The authors found a large interindividual variation in the slopes of ICP-PI correlations that should preclude any predictability of the ICP as measured by PI. Although original in its approach, this study has several limitations. First, it is probably premature to draw firm conclusions from data collected in 8 patients. The message could have been more appropriate if circumvented to patients with INPH only. A good estimation of ICP was recently shown with PI measurements in 45 ventilated head-injured patients with actually measured ICP below 40 mm Hg.2 That study confirmed previous data with an estimation of ICP using PI in 81 patients with various intracranial disorders.3 Second, there is large evidence that TCD measurements must be interpreted according to potential confounding factors, such as Paco2, blood hematocrit, arterial blood pressure, and heart rate. This point was extensively addressed in animals subjected to manipulation of cerebrovascular resistance, ie, CO2 challenge, arterial hypotension, and intracranial hypertension.4 The simulation model used in the present study took arterial blood pressure only in spontaneously breathing patients, which should represent an important limitation to the interpretation of data. Third, in patients with severe head injury, abnormal PI measurements on admission prompted interventions to restore cerebral perfusion pressure.5 We found PI as one TCD measurements to distinguish between patients at risk for secondary neurological deterioration after minor-to-moderate brain trauma.6 With no evidence of hypocapnia, elevated PI values can thus be viewed as the result of reduced compliance throughout the entire brain, including increased rigidity of the cerebral arteries and elevated ICP. It is obvious that TCD provides only a crude estimation of brain hemodynamics and ICP and it cannot be substituted for the actually measured ICP. However, its permanent accessibility in the emergency setting should promote the use of TCD to improve the initial management of brain-injured patients. Unfortunately, the present study does not provide enough evidence to abandon the use of TCD. Pierre Bouzat Gilles Francony Bertrand Fauvage Jean-François Payen Grenoble, France