Postoperative atrial fibrillation (POAF) is the most common complication following cardiac surgical procedures [1]. Its incidence ranges from 20% to 50%, depending on factors such as the type of surgery, patient population, and diagnostic criteria used [2,3]. According to data from the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database, POAF occurs in 25% of patients undergoing isolated coronary artery bypass grafting (CABG), 30% of those undergoing isolated valve procedures, and 40% to 50% of those undergoing combined CABG/valve surgeries [4].

In the United Kingdom, approximately 32,000 adult patients across 35 centers undergo cardiac surgery annually, and 30% develop POAF. Even a single episode of POAF extends ICU stay by 12–24 h and hospital stay by 2–5 days [5,6], and nearly triples in-hospital mortality (from 0.5% to 3.3%) and one-year mortality (from 3.7% to 9.9%) [7,8]. POAF increases healthcare costs by approximately £10,000 ($12,500) per patient in the first postoperative year [9]. For these reasons, the prevention of POAF was identified as one of the top ten research priorities by the James Lind Alliance Cardiac Surgery Priority Setting Partnership in 2019, involving patients, caregivers, and healthcare professionals [10].

Multiple studies have confirmed that patients who develop POAF face increased risks of morbidity, mortality [11,12], and complications such as stroke, heart failure, renal dysfunction, and prolonged hospitalization [12,13]. Reducing its incidence may improve clinical outcomes, increasing interest in identifying risk factors that could enable the implementation of preventive strategies.

The pathophysiology of new-onset POAF remains poorly understood. Recent evidence supports an inflammatory mechanism in its development [14,15]. Transfusions of red blood cells (RBCs) or whole blood (WB) modulate the inflammatory response to cardiac surgery by altering various inflammatory mediators and enhancing systemic inflammation [16,17]. Currently, the most effective and safest method to reduce the need for transfusion and its potential complications is intraoperative autologous blood salvage, especially in cardiac surgeries (Class I; Level A evidence) [18]. This technique is a cornerstone of patient blood management [19]. Autologous cell salvage may improve red blood cell viability, preserve their biconcave shape, enhance oxygen delivery, and attenuate the inflammatory response to surgery [20,21].

This retrospective cohort study was designed to assess a potential association between POAF, autologous blood salvage, and transfusion as the primary outcome, and mortality risk as a secondary outcome.

This retrospective observational cohort study collected data from 26 hospitals in four Brazilian states, evaluating 357 adult patients who underwent cardiac surgery with intraoperative autologous blood salvage between January 2021 and December 2023.

The study was conducted in accordance with the Declaration of Helsinki (2008 revision). Due to the retrospective nature of the study and the use of anonymized data, consent was considered unnecessary.

Nearly all types of cardiac surgeries were represented: CABG, valve replacement or repair (VR), aortic aneurysm repair (AAR), the Bentall-de Bono procedure (BB), redirection of blood flow (RBF), myectomy, cardiac tumor resection, and combined procedures (e.g., CABG + VR, CABG + AAR, or VR + AAR). Preoperative and intraoperative variables (volume and number of autologous units recovered) were assessed for potential associations with an increased risk of transfusion.

Patients were divided into two groups: one with 161 patients (45%) who only received their recovered autologous blood, and the other with 196 patients (55%) who also required transfusion of homologous blood components (Table 1). Patients who received homologous blood were more frequently female (p-value <0.001), had lower body weight (p-value = 0.002), and exhibited lower preoperative levels of hemoglobin, hematocrit, and platelets (p-value <0.001).

Regarding autologous blood recovery, the median volume was 505 mL (1.3 units) in the group that required homologous transfusion, compared to 462 mL (1.2 units) in the group that did not require homologous components (p-value <0.001) (Table 2).

The most commonly used blood component in the operating room was packed red blood cells (PRBCs; n = 107), with a median of 2 (IQR = 1.5) units. PRBCs were also the most frequently used component in the ICU.

With respect to surgical type, 163 patients (45.7%) underwent CABG, 129 (36.1%) underwent valve surgery, and 34 (9.5%) underwent aortic aneurysm repair. Other surgical procedures were less frequent.

Autologous blood transfusion was not associated with POAF (Odds ratio [OR] = 0.75; 95% CI: 0.45–1.11). The only factor independently associated with POAF in the multivariate analysis was older age (OR 1.04; 95% CI: 1.01–1.08) (Table 3).

When autologous salvaged blood volume and homologous components were evaluated together, homologous transfusion was associated with higher risk of POAF (OR 2.51; 95% CI 1.18–5.65), whereas number of autologous units was not (OR 0.66; 95%CI 0.39–1.01). This analysis was adjusted for pre-operative hemoglobin (Table 4).

Regarding mortality, intraoperative homologous blood transfusion was associated with a significantly increased risk of death (HR: 5.67; 95% CI: 1.64–19.62). This association remained consistent in the multivariate analysis (HR: 5.17; 95% CI: 1.44–18.56), adjusted for age, sex, preoperative hemoglobin, and body weight (Table 5). Transfusion at any time point (intraoperative, ICU, or ward) was also associated with an increased risk of death (HR: 5.15; 95% CI: 1.16–22.94).

Both a higher volume of autologous blood recovered and the use of homologous blood transfusion were associated with an increased risk of death. The number of autologous units recovered was associated with a higher risk of mortality (HR: 1.83; 95% CI: 1.39–2.41), as was older age (HR: 1.04; 95% CI: 1.01–1.08). Conversely, higher preoperative hemoglobin levels were associated with a reduced risk of death (HR: 0.80; 95% CI: 0.67–0.96) (Table 6). Patients who needed homologous blood transfusion had lower survival (log-rank - p-value <0.01, Figure 1).

Fig. 1.

Survival analysis stratified according to use of homologous blood products Patients who received homologous blood products (HBP) had lower survival (log-rank - p-value <0.01).

The transfusion-related findings in this study underscore the urgent need to implement Patient Blood Management, as emphasized by the World Health Organization (WHO) in 2021. The WHO highlighted that anemia has a global prevalence, affecting billions of individuals worldwide. Preoperative anemia in surgical patients is more prevalent than in the general population and is associated with significantly increased morbidity, mortality, and hospital length of stay across most surgical procedures [22].

When comparing baseline characteristics between groups, preoperative hematocrit, hemoglobin, and platelet levels indicated that the group requiring homologous blood transfusions already presented with anemia at admission (hematocrit median = 37%; IQR: 8.755 and hemoglobin median = 12 g/dL; IQR: 3.2 g/dL; p-value <0.001). In this study, the risk-adjusted impact of preoperative hematocrit on morbidity and mortality is consistent with prior publications in cardiac surgery patients. Multiple studies have demonstrated a significant correlation between preoperative or intraoperative anemia and adverse cardiac surgical outcomes [23–25].

LaPar et al. investigated the relative contribution of preoperative hematocrit and packed red blood cell (PRBC) transfusion to postoperative outcomes following CABG. In this multi-institutional analysis of over 33,000 patient records, the key findings showed that PRBC transfusions had a greater impact than preoperative hematocrit on postoperative mortality, stroke, and renal failure [26]. Therefore, transfusion may be more of a marker of a critically ill patient than the direct cause of death.

In the present cohort, the anemic group also had significantly lower platelet counts (p-value = 0.047), which may explain the higher incidence of reported ‘increased intraoperative bleeding’ (20 patients – 10.2%), the need for homologous blood components in the operating room (158 patients – 80.6%), and the higher volume of recovered autologous red blood cells (median: 505 mL; IQR: 365.75 mL - p-value <0.001). In contrast, this complication was reported in only two patients (1.2%) in the non-transfused group.

Some authors have suggested that perioperative blood transfusions may induce systemic inflammatory changes, potentially increasing the likelihood of developing POAF in transfused patients during cardiac surgery [14,27,28]. However, the findings of this study are supported by a previous study that showed that intraoperative autologous blood transfusion was not associated with POAF (OR: 1.13; 95% CI: 0.49–2.62), and the number of autologous blood units recovered was also not associated with increased risk of POAF (OR: 0.75; 95% CI: 0.45–1.11) [29]. These results remained consistent after adjustment for age, weight, sex, and preoperative hemoglobin level.

The findings of this multi-institutional retrospective cohort study of cardiac surgical patients indicate no association between POAF, intraoperative autologous blood recovery, and blood transfusion. The only factor associated with POAF was advanced age. Preoperative anemia and blood transfusion independently increased the risk of postoperative morbidity and mortality. Urgent implementation of patient blood management strategies is warranted to improve surgical outcomes.

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The data that support the findings of this study are available from the corresponding author upon reasonable request.