[The usefulness of transcranial Doppler ultrasonography in the early phase of head injury]

Abstract

Objectives: 1) To know the transcranial Doppler (TCD) patterns in the early phase of the severe and moderate head injury and its prognostic implications. 2) To ascertain the TCD measurements concordance among different operators.

Material and methods: A prospective observational study in 275 consecutive intensive care unit patients with severe or moderate head injury was designed. Within 12 hrs. of the traumatic event a TCD study was done and a second one within the first 24 hrs. All patients were managed following a protocol made with the aim of promoting the early evacuation of the intracranial space occupying mass, preventing delayed brain damage and keeping the intracranial pressure (ICP) < 20 mm Hg and the cerebral perfusion pressure (CPP) > 60 mm Hg. The patient outcome was categorized with the Glasgow Outcome Score (GOS) at the discharge of the ICU. The mean velocity (MV) and the pulsatily index (PI) were measured in both middle cerebral arteries (MCA) and in the intracranial part of the internal carotid arteries in order to calculate the Lindegaard index. The presence of hypoperfusion, hyperaemia, vasospasm and cerebral circulatory arrest was registered. The presence of high intracranial pressure (HICP), the ICP peak and mode, the lowest CPP, the CPP mode, the minor medium arterial pressure (MAP) and the MAP mode were also recorded. A concordance analysis was made to ascertain the validity of the TCD data obtained by different operators.

Results: The concordance analysis among observers showed a kappa index of 0.7863 (p < 0.0001). The median stay in ICU was 7 days (Q1-Q3 of 3-15 days). The intra-ICU mortality was 20.72% (57/275) and the 53.44% of patients (147/275) showed favourable outcome (GOS 4-5) at the discharge of the ICU. The initial TCD showed a MV decreased and a PI increased in the MCA. Eighty of the e 275 patients showed a normal haemodynamic pattern whereas the pattern was abnormal in the remaining 195, degrees the hypoperfusion pattern predominated overall (181/275, 61.87%). HIPO was registered in 123 patients (58.01) an not in 89 (41.99%). A statistic significant correlation was found between the HICP and the DTC data. The early hypoperfusion pattern was related with a higher incidence of HICP (p > 0.05). The decrease in the MV and the increase in the PI were significantly associated with a higher mortality and a worse functional outcome (except for the MV within the first day). The PI was high within the first 24 h in the group of patients who died but was normalised among the survivals group.

Conclusions: Early TCD detects a cerebral hypoperfusion status in the severe and moderate head injury that may imply therapeutic considerations. This hypoperfusion strongly correlates with the severity of the injury, the incidence of HICP and the functional outcome at the ICU discharge. In experienced hands, TCD measurements are reliable when done by different operators.