108 H. M. da Rocha Coutinho et al.: J Extra Corpor Technol 2025, 57, 105 – 112
Table 2. Summary of descriptive findings.
Variables |
Mean ± standard deviation |
Minimum |
Percentile |
Maximum |
|
|
|
25 |
50 |
75 |
|
Age |
59.9 ± 12.4 years |
23 |
55 |
58 |
70 |
81 |
Arterial flow |
5.16 ± 0.50 L / min |
3.7 |
4.9 |
5.1 |
5.4 |
6.2 |
CPB time |
102.8 ± 40.5 min |
56 |
69 |
85 |
127 |
200 |
Aortic clamping time |
80 ± 33 min |
42 |
59.2 |
69 |
91.25 |
171 |
Initial hemoglobin level |
10.57 ± 2.49 mg / dL |
5.2 |
9.3 |
10.3 |
12.5 |
16.3 |
Final hemoglobin level |
10 ± 4 mg / dL |
5.7 |
8.3 |
9.8 |
10.2 |
30.6 |
Initial hematocrit level |
31.48 ± 8.5 % |
19 |
28 |
32 |
37 |
48 |
Final hematocrit level |
28.68 ± 4.7 % |
17 |
25 |
30 |
31 |
38 |
Plasma free hemoglobin before CPB |
0.02 ± 0.17 mg / dL |
0.01 |
0.01 |
0.02 |
0.02 |
0.09 |
Plasma free hemoglobin after CPB |
0.08 ± 0.48 mg / dL |
0.02 |
0.037 |
0.075 |
0.10 |
0.21 |
Plasma free hemoglobin 24 h after CPB |
0.03 ± 0.15 mg / dL |
0.01 |
0.02 |
0.02 |
0.04 |
0.07 |
Blood samples for analysis of hemoglobin, hematocrit and free plasma hemoglobin were collected before CPB, at the end of CPB and 24 h post-CPB.
Table 3. Summary of surgical procedures.
Variables |
Statistics |
n(%) |
Mitral valve replacement |
No |
26( 83.9) |
Yes |
5( 16.1) |
Tricuspid valve repair |
No |
30( 96.8) |
Yes |
1( 3.2) |
Aneurysm repair |
No |
29( 93.5) |
Yes |
2( 6.5) |
Closure of interatrial communication |
No |
30( 96.8) |
Yes |
1( 3.2) |
Coronary artery bypass grafting |
No |
9( 28.8) |
Yes |
22( 70.9) |
Aortic valve replacement |
No |
10( 32.3) |
Yes |
21( 67.7) |
Categorical variables are described as numbers( percentages).
not recorded. No adverse events were reported in the study population.
The effect size used in the comparison of continuous variables, such as age tested by Student’ s t-test, was Cohen’ s d. This statistical measure can be used to quantify the magnitude of the difference between two groups in terms of standard deviations. Interpretation of Cohen’ s d is as follows: 0.2 denotes a small effect, 0.5 indicates a medium effect, and 0.8, a large effect [ 17 ]. Therefore, the effect size obtained in the analysis confirms the absence of age difference between the two groups.
The association between group and gender was analyzed using the chi-square test of independence. Phi was employed to calculate the effect size, ranging from 0 to 1, with higher values indicating a stronger association; therefore, as can be seen, the effect size was negligible.
As mentioned in Materials and methods section, the effect size used in the comparisons by the Mann-Whitney test was the r biserial point. This measure of effect size indicates the strength of association between a dichotomous variable and a continuous variable. Like other correlation values, its range spans from �1to1, withvaluescloserto1indicatingastronger association. According to Cohen [ 14 ], the cut-off points for interpretation are as follows: small: r = 0.10 to 0.29; medium: r = 0.30 to 0.49; large: r = 0.50 and above.
No significant differences were observed for the variables tested in Table 5; in fact, the effect size for all variables was small.
The boxplots presented in Figures 2 – 4 illustrate the distribution of hemolysis markers across the study groups. Notably, Figure 2 includes extreme outliers that significantly expand its scale compared to Figures 1 and 3. To preserve the integrity of each dataset, the Y-axis scales were not standardized across figures, as doing so would artificially compress the distribution of Figures 1 and 3, making it difficult to discern variability and quartile distribution in these groups.
This visualization approach ensures that the true dispersion and data trends remain accurately represented, preventing misinterpretation of hemolysis levels. While some outliers in Figure 2 suggest increased variability, they do not affect the median and interquartile range observed in the other groups. Future studies may explore alternative statistical normalization techniques; however, in this analysis, retaining distinct scales provides a clearer and more clinically relevant interpretation of SafeCEC Ò performance in preventing retrograde flow-induced hemolysis.
The results of this study suggest that the use of SafeCEC Ò in CPB with a centrifugal pump is safe and does not elevate the hemolysis normally generated by CPB.
Discussion
Hemolysis is found in all surgical procedures using extracorporeal circuits. Several studies have identified rising levels of plasma free hemoglobin [ 18, 19 ], designating it as a marker of blood trauma. Hemolysis can occur in three distinct ways: