Page 22 - ANZCP Gazette APRIL 2022
P. 22
Statistical analysis
The emboli data from each patient were exported into a Microsoft Excel (Microsoft Office®. Office Proofing Tools © 2012 Microsoft Corporation) spreadsheet for analysis. The data was not normally distributed, and hence they were reported as medians with interquartile ranges (IQR), and non-parametric tests were used for statistical analysis. The Wilcoxon Paired signed-rank test was used to compare gaseous and solid emboli counts. The two-sample Wilcoxon rank sum (Mann-Whitney) test was use for comparing the SACC and PASC. Statistical significance was defined with p-values <0.05 and <0.01 as highly significant. P-values ≥ 0.05 were defined as not significant. The statistical analysis was performed using Strata, version 16 (StataCorp LLC, TX, USA).
Results
size, and probable type II statistical error. When looking at average bypass times between the two groups during our phase of interest (Table 2), we can see that partial clamp use almost doubled the single clamp use average time at 24 minutes. This may also contribute to the high embolic counts during our study.
When comparing the median total emboli count for both techniques, medians with interquartile ranges were used as our data was not normally distributed.
Median total emboli count during aortic cross clamp removal to termination of bypass (our phase of interest) for SACC was 102.5 (46-197) and 181.5 (92-300) for PASC (Table 2). At first glance, we can see that the median embolic count in the PASC group was higher than the
Total average bypass time was 77.5 minutes and 80.15 minutes for SACC and PASC respectively.
Of interest, our results showed that there was little difference in median embolic counts prior to cross clamp removal (319.5 and 373.5 for SACC and PASC respectively; p-value 0.718), and therefore gives us a reliable baseline when comparing our results between the two techniques.
SACC, although this did not show any statistical significance
(p-value of 0.153). This could be due to the small sample
The proportion between gaseous and solid emboli between the two groups during our phase of interest are displayed in Table 3. Median gaseous emboli count was 84 (44-185) and 157.5 (80-254) for SACC and PASC respectively. Median solid emboli count was 12.5 (1-18) and 16 (10-39) for SACC and PASC respectively. Our data show that there were more gaseous and solid emboli generated with the PASC group, compared to the SACC group. Again, none of these comparisons show statistical significance (p-value 0.129 gaseous emboli v 0.321 solid emboli) for which is probably accounted due to the small sample size and probable type II error. Therefore, it is hard to make solid conclusions with our limited data set. However, it is important to point out, that although there is not statistical
19 APRIL 2022 | www.anzcp.org
