Page 38 - ANZCP Gazette APRIL 2022

Willcox et al.
3
Figure 1. CONSORT diagram.
for technology change over time such as reduced circuit size, we compared the groups a priori within two time periods: early (2007-2012; 53 accept vs. 53 refuse trans- fusion) and late (2013-2018; 65 accept v 65 refuse trans- fusion).
Statistical analysis
STATA, version 15.0 (StataCorp LP, College Station, TX, USA), was used for all analyses. A propensity score for refusal of blood transfusion within each time period was calculated using the following preoperative variables: hospital, age, severe left ventricular (LV) dysfunction, gender, emergency, body mass index, cerebrovascular disease, chronic obstructive pulmonary disease, myo- cardial infarction, reoperation, Hb level, diabetes, cre- atinine, hypertension, congestive heart failure, procedure type, Euroscore, CPB time and the year of operation. Patients were propensity matched in a 1:1 ratio using nearest-neighbour matching (caliper 0.01) without replacement of subjects resulting in 118 matched pairs.
Patient preoperative and perioperative characteris- tics were compared between cohorts according to an early (2007-2012) and late (2013-2018) time period and acceptance or refusal of transfusion. The Kolmogorov– Smirnov test was used to check for normality of contin- uous data according to skewness and kurtosis. Differences between groups were assessed using the Wilcoxon rank-sum test for continuous data and
Pearson’s chi-square test and Fisher’s exact test for cate- gorical variables. Categorical data are reported as % and continuous data as median (interquartile range). A p value of <0.05 was considered to be statistically signifi- cant.
Results
The propensity-matching process provided comparable patient cohorts of those transfused compared with those refusing transfusion based on the preoperative variables and procedure types reported in Table 1. The CPB inter- ventions for the two groups are shown overall in Table 1. The subgroup analysis of the early and late time periods is in Table 2. In patients accepting transfusion, 49% received a blood product of any kind (62% early vs. 38% late, p = 0.01), with red cell transfusion rates of 41% overall (54% early vs. 31% late, p = 0.012). The use of cell salvage was significantly higher for patients refusing transfusion over the entire period and within both the early and late time periods (70% vs. 22% p < 0.001, 67% vs. 25% p<0.001 and 72% vs. 20% p<0.001, respec- tively). Haemofiltration use, while low, was greater in patients refusing transfusion (8% vs. 4% p = 0.031). Tranexamic acid was given more frequently to patients refusing transfusion overall; however, equivalent rou- tine use was observed in the more recent period.
Of the CPB modifiable variables relating to haemodi- lution, there was no difference in circuit prime volume between groups within either the early or the late period.
35 APRIL 2022 | www.anzcp.org