Introduction: Transcranial Dopplers (TCDs) have been used to monitor cerebral blood flow velocities in subarachnoid hemorrhage (SAH).The purpose of our two-part study was to compare the reliability of relative increases in flow velocities with conventionally used absolute flow velocity indices and to correct for hyperemia-induced flow velocity change.
Methods: Part 1: Charts of 50 patients admitted to Hahnemann University Hospital with aneurismal SAH were reviewed. Mean middle cerebral artery maximum flow velocities (MCA-MFV) were reviewed for initial flow velocities (IFVs) and maximal flow velocities (MFVs) that were reached during hospital course. Correlating flow velocities (SFVs) were noted in patients who developed symptomatic vasospasm. MFV/IFV and SFV/IFV ratios were calculated to evaluate relative changes in flow velocity. Part 2: Correction for hyperemia was derived from Lindegaards Ratio using extracranial internal carotid artery (ICA) flow velocity ratio (corrected MCA-MFV/observed MCA-MFV = EC-ICAFV (day1)/observed EC-ICAFV).
Results: Part 1: All 10 patients who developed symptomatic vasospasm exhibited a twofold increase (SFV/IFV: >2) in flow velocities prior to developing symptomatic vasospasm, and 5 patients had a threefold increase (SFV/IFV: >3). Of the 40 patients who did not develop symptomatic vasospasm, 33 patients did not have a twofold increase in their flow velocities at any time. The positive predictive value for MFV/IFV greater than 3 (n = 6) and SFV/IFV greater than 3 (n = 5) was 100%. The negative predictive value for MFV/IFV less than 2 (n = 33) was 100%. Data using relative changes (twofold increase) in flow velocity was more sensitive (100 to 90%), specific (83 to 70%), and predictive (positive predictive value [PPV]: 59 to 45%; negative predictive value [NPV]: 100 to 97%) for symptomatic vasospasm than absolute flow velocity indices using MCA-MFV greater than 120 even in combination with Lindegaards Ratio (MCA/ICA greater than 3). Part 2: Correction for hyperemia by modifying Lindegaard's Index in the 32 patients where data was available improved the PPV of absolute flow velocities from 44 to 62%. In this population, the application of this equation while evaluating relative change in flow velocities improved PPV of twofold increase from 57 to 73%.
Conclusion: Relative changes in flow velocities in patients with aneurysmal SAH correlated better with clinically significant vasospasm than absolute flow velocity indices. Correction for hyperemia improved predictive value of TCD in vasospasm.