After the aneurysm was secured with the patient in stable clinical condition, continuous bilateral TCDs were performed using a PMD 100 Transcranial Doppler System and a 2-MHz transducer fixed in a standard head frame (Spencer Technologies, Seattle, USA). The middle cerebral artery (MCA) was insonated bilaterally through temporal bone windows. Tests were performed in a standard hospital bed with the head elevated at 30°. Once the temporal windows were located, the proper transducer position and angulation were noted and the bilateral probes were fixed on the head frame to prevent motion. The MCA was identified using established TCD criteria: depth from 30 to 60 mm, flow direction toward the transducer, flow velocities from 46 to 86 cm/s, and the anatomic relationship with the internal carotid artery (ICA) bifurcation.

End-tidal CO₂ (P _ET CO₂) values were precisely targeted for cerebrovascular reactivity (CVR) measurements using the method of prospective end-tidal targeting previously described by Slessarev et al. [7]. Blood pressure cuffs or arterial lines, when available, were used to record the blood pressure. After a short acclimatization period, subjects’ P _ET CO₂ levels were set to 40 mmHg for 2 min (normocapnia; Step 1), 45 mmHg for 2 min (hypercapnia; Step 2), and then back to 40 mmHg (normocapnia; Step 3). The fraction of inspired oxygen (FiO₂) was kept at 100 %.

MCAv was recorded bilaterally for each phase of the test. A global MCAv was generated averaging the means of both hemispheres. MCAv, pulsatility index (PI), end-tidal CO₂ and O₂ levels, heart rate (HR), blood pressure (BP), and intracranial pressure (ICP), when available, were recorded. The absolute change and percentage change for P_ETCO₂ and MCAv between each step in the protocol were calculated. The CVR index was calculated by dividing the percentage change in MCAv by the absolute change in P _ET CO₂ (CVR = % ∆ MCAv/1 mmHg ∆ CO₂). Separate calculations were made for the right and left side for each step of the test, and both sides were averaged to generate a global CVR (gCVR) index. The mean plus and minus one standard deviation of the control group data were used to define a “normal CVR,” against which the SAH data was compared. The groups were also compared using the definition of “normal CVR” by Marshall et al. [8] as an increase in MCAv of at least 2 % per mmHg of CO₂ change. The Mann-Whitney test was used to analyze the differences between MCAv and CO₂ levels between the phases and the CVR index between the two groups. To test the association between mean MCAv and CO₂ level, a general linear mixed model (CO₂ group – SAH versus Control – and interaction between CO₂ group) was used. A random intercept model was used to account for variability between patients at baseline and a random slope to account for the variability between patients in the association between MCAv and P _ET CO₂.