Effects of perioperative red blood cell transfusion on systemic immune indicators and postoperative recovery in patients undergoing cesarean section: A propensity score matching analysis

Background Postpartum hemorrhage (PPH) is one of the leading causes of maternal death in all regions of the world. Intra-operative red blood cell transfusion is one of the most effective strategies to treat PPH. However, the effect of blood transfusion on patients undergoing cesarean section remains unclear. The aim of this study was to investigate the effects of a intra-operative blood transfusion on systemic immune indicators and postoperative recovery in patients undergoing cesarean section. Methods This retrospective study enrolled patients undergoing cesarean section between January 2016 and June 2020. We divided these patients into two groups according to whether they received a intraoperative blood transfusion. Baseline characteristics were compared between groups. After propensity score matching, the length of stay (LOS), perioperative systemic inammation-based scores, and postoperative complications were compared. Univariate and multivariable Cox proportional hazard models were used to evaluate the associations between covariates and outcomes.


Introduction
Postpartum hemorrhage (PPH) is one of the leading causes of maternal death in all regions of the world, and the de nition of PPH is controversial (1). Generally, blood loss of 500 ml or more, blood loss with related physical signs regardless of the mode of delivery, and symptoms of hypovolemia after delivery are considered indicators of PPH (2). In addition, the American College of Obstetricians and Gynecologists (ACOG) issued recommendations that emphasize the importance of an organized and systematic process to help coordinate the response and management of PPH (3). Red blood cell transfusion is one of the most important treatments for PPH (4).
However, the effect of intra-operative red blood cell transfusion on postoperative recovery remains unclear.
For patients undergoing cesarean section with bleeding and anemia, blood transfusion can save lives (5,6).Intraoperative strategies for red blood transfusions may affect short-term clinical outcomes in patients undergoing surgery (7). Several retrospective studies have shown that transfused patients have an increased risk of developing infectious and noninfectious postoperative complications after major surgery(8). In addition, the results from a recent randomized controlled trial (RCT) indicate that a restrictive approach in which transfusions were triggered by hemoglobin (Hb) levels lower than 7 g/dL increased the risk of postoperative complications.
Furthermore, some studies have found that the infusion of allogeneic blood inhibits the immune function of patients, which is not conducive to improving the prognosis in the perioperative period (9).
The association of intraoperative blood transfusion and postoperative recovery in patients suffering from PPH remains unclear.
Additionally, studies examining whether receiving blood transfusion during surgery will promote postoperative complications and the effect of the duration of hospitalization on obstetric patients are still lacking. Therefore, we conducted a retrospective study to assess the association between intraoperative blood transfusion and postoperative recovery in patients undergoing cesarean section. We hypothesized that intraoperative blood transfusion is associated with a longer length of stay. Furthermore, we investigated the association between perioperative blood transfusion and perioperative systemic in ammatory indicators, postoperative recovery parameters, postoperative complication rates and readmission rates.

Patients
Patients who underwent cesarean section at Shanghai First Maternity and Infant Health Hospital from January 2016 to June 2020 were enrolled in this study. This study was reviewed and approved by the Shanghai First Maternity Ethnics Committee (Protocol: #2020-049). All pregnant women or their relatives signed an informed consent form for research data use before undergoing cesarean section. The exclusion criteria were emergency surgery, twin pregnancy, patients treated with anti-in ammatory drugs or immunosuppressants for more than one month before cesarean section, patients with chronic in ammatory diseases, and incomplete medical records. Pregnancy characteristics, medical history, and operative details were reviewed and recorded.

Primary outcome
The primary outcome of this study was the LOS. LOS was de ned as the interval between the date of surgery and the date of discharge.

Secondary outcome
Secondary endpoints included perioperative systemic in ammation-based scores (NLR, LMR, and SII). The counts of neutrophils, lymphocytes, monocytes and platelets were recorded three days before surgery (pre-surgery), the rst day after surgery (POD1) and the third day after surgery (POD3). NLR was de ned as the neutrophil to lymphocyte ratio, LMR was de ned as the lymphocyte to monocyte ratio, and SII was de ned as neutrophil*platelet/lymphocyte ratio. Postoperative complications, wound complications, ICU admission rates and readmission rates within 30 days were reviewed and recorded.

Anesthesia Care
Upon entering the operating room, pregnancy was monitored according to the American Society of Anesthesiologists (ASA) monitoring standards. Spinal anesthesia was the rst choice for all cesarean sections in patients without contraindications. Patients were placed in the right recumbent position and punctured in L3-L4 or L2-L3 space. After observing the free ow of the cerebrospinal uid after subarachnoid puncture, 0.5% hyperbaric ropivacaine (2 mL of 0.75% hyperbaric ropivacaine and 1 mL of 10% glucose) was administered. An epidural indwelling catheter with a length of 4 cm was used in the epidural space. Five minutes later, the anesthetic plane was evaluated at the T4-T6 level. When the effect was not satisfactory, 5 ml of 0.5% ropivacaine were injected into the epidural space. When the patient presented with contraindications to spinal anesthesia (coagulation dysfunction, lumbar disease, shock, etc.), general anesthesia was administered. General anesthesia was induced with propofol (target-controlled infusion, effect-site concentration: 3.0-4.0 µg/ml), remifentanil (0.3-0.5 µg/kg), and rocuronium (0.6 mg/kg). The patients were then endotracheally intubated, and general anesthesia was maintained with propofol, sufentanil and remifentanil. Repeated bolus injections of sufentanil and rocuronium were administered as necessary throughout the operation.

Statistical Analysis
We compared different clinical factors between the blood transfusion group and the non-blood transfusion group. Descriptive statistics, including the means and standard deviations, were reported for continuous variables. Frequency counts and percentages were calculated for categorical variables such as the ASA physical status, gestational age at delivery, birthweight, operation time, blood loss, and pre-Hgb levels. Univariate Cox proportional hazards models were tted to evaluate the effects of continuous variables on time-to-event outcomes. Multivariable Cox proportional hazard models were used for the multivariate analysis to include important and signi cant covariates. We performed a propensity score matching analysis to reduce selection bias by building a matched group of patients to compare LOS between patients with a blood transfusion and those without a blood transfusion. Patients were matched using a 5-to-1 digit greedy match algorithm. Statistical analyses were performed with SPSS version 17.0 (SPSS Inc., Chicago, IL, USA).
P<0.05 was considered statistically signi cant.

Results
A total of 1426 patients who underwent cesarean section were enrolled in this study. After reviewing the exclusion criteria, 329 patients were included in the BT group, and 909 were included in the NBT group ( Figure 1). After propensity score matching, no signi cant differences in age (P=0.139), BMI (P=0.526), ASA status (P=0.811), gestational age at delivery (P=0.363), birthweight (P=0.262), primary mode of anesthesia (P=0.893), primary cause of PPH (P=0.993), and experience of the attending obstetrician (P=0.878) were observed between groups. However, signi cant differences in the operation time (P<0.001), blood loss (P<0.001) and pre-Hgb levels (P<0.001) were identi ed (Table 1).

Primary outcome
The LOS was de ned as the date of surgery to the date of discharge. The LOS in the NBT group was signi cantly shorter than that in the BT group (mean time (days): 4.2 vs. 7.1, P=0.026, Fig. 2A).

Secondary outcome
Regarding postoperative complications, the incidence of postoperative vomiting was higher in the BT group than in the NBT group (3.2% vs. 4.9%, P=0.032, Fig. 2B). The incidence of postoperative fever was higher in the BT group than in the NBT group (5.41% vs. 2.24%, P=0.032, Fig. 2B). For postoperative pneumonia, no signi cant differences were observed between the two groups (3.2% vs. 3.4%, P=0.563, Fig. 2B). Furthermore, the incidence of wound complications in the BT group was higher than that in the NBT group (15.44% vs. 10.45%, P=0.028, Fig. 2B). The incidence of intestinal obstructions in the BT group was signi cantly higher than that in the NBT group (5.88% vs. 2.75%, P=0.034, Fig. 2B). However, the NBT group had signi cantly lower 30-day and ICU admission rates than the BT group (4.40% vs. 7.14%, 1.70% vs. 4.10%, P=0.018 and P=0.034, respectively, Fig. 2C). According to the univariate analysis, birthweight, a longer operation time, low pre-Hb levels and blood transfusion were associated with a longer LOS. The multivariable analysis showed that birthweight, operation time, low pre-Hb levels and blood transfusion were associated with a longer LOS. After matching, blood transfusion was still associated with a longer LOS.
After matching, no signi cant differences in the preoperative NLR, LMR or SII were observed between the two groups (Fig. 3). Compared with preoperative assessments, the NLR and SII in the two groups increased signi cantly and the LMR was signi cantly decreased in all patients on POD1 and POD3 (P<0.001, Fig. 3B). The NLR and SII values were slightly lower on POD3 than on POD1 but were still higher than the preoperative values (P<0.001, Fig. 3A-C). The NLR and SII values in the NBT group on POD1 and POD3 were signi cantly lower than those in the BT group (P<0.001, Fig. 2A-C), whereas a signi cantly higher LMR was detected in the NBT group than in the BT group (P <0.001, Fig. 3B).

Discussion
In this large retrospective cohort study, our results showed that the hospital stay of patients who received a blood transfusion was signi cantly longer, the incidence of postoperative vomiting and wound complications was higher than that of patients without a blood transfusion, and the postoperative in ammatory index uctuated more drastically in patients who received a blood transfusion.
The main bleeding site in postpartum hemorrhage is the uterus(10). When postpartum hemorrhage occurs, the uterine muscle tissue is hypoxic and does not contract well, and the sensitivity of oxytocin receptors decreases. Repeated hemostasis operations increase further damage to the uterine muscles (10). At the same time, the sinusoids on the dissected surface of the placenta cannot be closed, and brinogen deposition and thrombosis do not play a role (10). Therefore, obstetric hemorrhage has the characteristics of a fast speed, large volume, di cult evaluation, and di culty of stopping bleeding quickly.
Blood transfusion is the most common method for the clinical treatment of patients with obstetric hemorrhage. It effectively solve the patient's anemia caused by excessive blood loss, helps maintain blood volume and blood pressure, exerts a certain preventive effect on patients with shock, and improves the prognosis of patients (11,12). However, a large number of clinical practices have found that after patients with obstetric hemorrhage receive a large transfusion of blood, although their life and safety are guaranteed, they also experience certain side effects, mainly manifested as abnormal blood coagulation (13). Due to the long-term storage of blood, platelets and coagulation factors are destroyed to a certain extent. After entering the patient's body, the active coagulation factors will be reduced, which will lead to coagulation dysfunction.
In terms of postoperative complications, a large number of allogeneic blood transfusions will signi cantly increase the incidence of postoperative complications, such as blood transfusion-related lung injury (14) and surgical site infections (15). In the present study, the incidence of postoperative complications in patients receiving blood transfusion increased signi cantly during the perioperative period. Zhang et al reported that a perioperative blood transfusion potentially increases the risk of postoperative complications and is associated with a prolonged LOS (9). Furthermore, another study reported that perioperative blood transfusion may be associated with worse outcome in patients undergoing surgery(16). After blood transfusion, the immune system changes signi cantly. These changes have been shown to be potential mechanisms that increase the risk of infection, graft-versus-host disease and postoperative complications after blood transfusion (17). According to recent studies, the suppression of perioperative immune function may promote postoperative metastasis and recurrence in patients with cancer(18). A large number of studies have also shown that certain markers of blood components, such as NLR, LMR and SII, comprehensively indicate the balance of the host immune system and have been considered indicators related to the prognosis (19,20). An increased NLR or SII and decreased LMR are associated with a poor prognosis in patients undergoing surgery (21,22). Based on our results, patients undergoing a blood transfusion exhibit greater uctuations in these systemic in ammatory markers after surgery. Thus, blood transfusion is related to the strong suppression of immune function and/or in ammatory response disorders, such as a higher complication rate and longer hospital stay, in patients receiving a blood transfusion. (23) Our research has limitations.The limitations include the following: the study was a retrospective study rather than a random controlled trial, and the data were obtained from a single center. Second, even if we perform a propensity score matching analysis, considering the various factors related to postpartum hemorrhage, we still cannot avoid the possibility of biased results due to unmeasured confounding factors.

Conclusion
In summary, the transfusion of a large amount of blood can save the life and ensure the safety of patients with PPH, but it will exert certain effects on the levels of in ammatory factors and postoperative recovery of the patients.   Figure 1 Flowchart of patients enrolled in the study