The Frontal Bone Window for Transcranial Doppler Ultrasonography in Critically Ill Patients: Validation of a New Approach in the ICU.

ABSTRACT

BACKGROUND AND OBJECTIVE: The temporal bone window (TBW) for transcranial Doppler (TCD) often fails to insonate the anterior cerebral artery (ACA). The frontal bone window (FBW) has never been evaluated in intensive care units (ICU). The main objective was to determine the ability of the FBW to assess ACA velocities in critically ill patients. METHODS: A prospective study was conducted in two ICUs of the Montpellier University Hospital (France), between November 2014 and September 2016. Adult patients admitted to ICU for brain injury, with a Glasgow Coma Scale score ≤ 13, were enrolled within 3 days after admission. A first TCD examination was carried out bilaterally through the TBW and FBW by an intensivist expert in TCD, repeated by the same examiner, and 15 min later by an intensivist certified in TCD, designated as non-expert, blinded. The success of the FBW examinations was defined by the ability to measure the ACA velocities. Intra- and interobserver agreements were analyzed according to the Bland and Altman method. RESULTS: A total of 147 patients were analyzed. The FBW succeeded in insonating the ACA in 66 patients [45%, CI (37-53)], 45 bilaterally and 21 unilaterally. For 16 patients (11%), the FBW was the only way to measure ACA velocities. By combining the two techniques, the ACA success rate increased from 62% CI (54-70) to 73% CI (65-79) (P = 0.05). Intra- and interobserver mean biases and 95% limits of agreement for ACA systolic velocity measurements through the FBW were 1 (- 33 to 35) and 2 (- 34 to 38) cm s-1, respectively. For paired TBW and FBW measures of ACA velocities, mean biases (± SD) for ACA systolic, and mean and diastolic velocities were relatively close to zero, but negatives (- 7 ± 33, - 2 ± 19, - 1 ± 15 cm s-1, respectively), highlighting that ACA velocities were lower with the FBW (A2 segment) than TBW (A1 segment). The correlation coefficient for ACA systolic velocities measured by the FBW and TBW was R = 0.47, CI (0.28-0.62). No risk factors for failure of the FBW were identified. CONCLUSIONS: In ICU, the FBW was able to insonate the ACA in 45% of patients admitted for brain injury, without the use of contrast agents. The FBW could improve the detection of ACA vasospasms.