Page 30 - ANZCP Gazette May 2023

Jabur et al.
3
(Mx). The derived parameter “Mx” reflects the strength of that correlation.18,19 Strong correlation (high Mx) between CBFv and MAP suggests absent autoregula- tion, and weak correlation (low Mx) suggests intact autoregulation. Single values of Mx were calculated as the averaged correlation coefficient between 30 consec- utive 10 s paired values of MAP and CBFv. This approach eliminates high-frequency heart rate and respiratory rate waves, which can confound measurements.18 All calculations were performed using a 300 s moving win- dow and updated every 10 s to smooth out the transition between the different time epochs.19 Rather than attempting to define an Mx threshold separating “autoregulated” and “dysautoregulated” states, the Mx metric was used as a continuous variable assumed to represent an index of cerebrovascular response to changing MAP. Emboli counting and Mx measurement began from the inception of CPB and continued until 30 min after removal of the aortic cross-clamp.
Artifact removal
Data were summarized according to group (OCS and CCS), both overall (pooled data for both cerebral hemi- spheres) and by side (right and left hemispheres), using means and standard deviations or medians and inter- quartile ranges, as appropriate.
The primary outcome measure was the change in Mx, defined as the difference between the mean Mx in the 30 min before and after cross-clamp removal (i.e., the pre−post change in mean Mx), both overall and within hemispheres. Comparisons were made between the OCS group and the CCS group and between the left and right hemispheres in both groups. Analysis was per- formed using a linear mixed model with a random intercept for participants to account for the two repeated (right/left) measures within each patient (allowing for both within-person and between-person variability). First, a model including the main effects of group (OCS vs CCS) and side (right vs left hemisphere) on the change in Mx was evaluated. A further model including an interaction between group and side was then assessed in order to ascertain whether the effect of group differed according to which side was being considered. Marginal mean differences between groups and by side were esti- mated using this model through post hoc tests. Finally, the relationship between total emboli count (from initi- ation of CPB to 30 min post-aortic cross-clamp removal) and change in Mx was evaluated separately for OCS and CCS groups.
Analysis was performed using Stata version 17 (StataCorp LLC, Texas), with a threshold for statistical significance set at p < 0.05.
Results
Forty subjects were recruited from an approximate pool of 1400 potential patients. Seventeen patients were ruled out of the study during the enrolment process (one patient declined consent, six patients were withdrawn because of technical monitoring issues, two patients did not have an adequate transtemporal window, and eight patients had their surgery cancelled). Characteristics of the relevant patient and surgical details in both study groups are described in Table 1.
A summary of the Mx and emboli counts in the study periods of interest is presented in Table 2. Emboli counts in patients undergoing OCS were substantially higher than CCS, particularly after the aortic cross-clamp was removed, as previously reported.17 The majority of emboli passed into the right cerebral hemisphere circu- lation in both patient groups.
In linear mixed model analyses, there was no signifi- cant overall effect of group (mean pre–post clamp removal difference (md): 0.08, 95% CI: −0.04, 0.21; p = 0.19) (Figure 1) or side (md: 0.01, 95% CI: −0.03, 0.05; p = 0.74) on the change in Mx. There was, however,
Raw intraoperative Mx measurements may be con- founded by artifactual CBFv and MAP data arising from interventional disturbances. Therefore, potential artifact- generating events were recorded in real-time and time- aligned with MAP and CBFv records to allow retrospective cleaning using three criteria: Criteria 1— removal of MAP data with an abrupt drop to zero or negative pressure associated (e.g.,) with zeroing of the pressure transducers; Criteria 2—removal of MAP data with an abrupt temporary (<3 s) increase >50 mmHg above the baseline associated with blood sampling or arterial line flushing; Criteria 3—removal of MCA flow velocity data indicating an abrupt increase (>15 cm/ sec) in the MCA flow velocities lasting more than 5 s (adjacent abrupt events are classified as a “block”) asso- ciated with knocking or movement of the TCD probes. The Mx was recalculated across the phases of interest after artifacts were removed.
Statistical analysis
After the removal of artifact, the Mx and emboli data from each patient were exported into a Microsoft Excel (Microsoft® Office Proofing Tools© 2012 Microsoft Corporation, Washington State) spreadsheet. To address our primary hypothesis, Mx data from 30 min before and after removal of the aortic cross-clamp were consid- ered, along with MCA emboli counts from initiation of CPB to cross-clamp removal and from cross-clamp removal to 30 min after the release of the aortic cross- clamp. The mean Mx (i) 30 min before the cross-clamp removal and (ii) 30 min after the cross-clamp removal was calculated for each patient’s left and right hemispheres.
27 MAY 2023 | www.anzcp.org