INTRODUCTION

Bleeding is a common and serious complication following coronary artery bypass grafting (CABG). Patients may enter a clinical picture compatible with hypovolemic and cardiogenic shock due to excessive blood loss. Some patients with postoperative bleeding need reoperation due to the development of cardiac tamponade. In addition, bleeding diathesis and disseminated intravascular coagulation secondary to excessive use of blood products may be encountered. In addition to these, problems such as the need for prolonged mechanical ventilator support, increased hospital infections, renal dysfunction, and prolonged intensive care unit (ICU) and hospital stay may occur^[1]. Therefore, predicting postoperative bleeding and taking necessary precautions will reduce morbidity and mortality rates and, also importantly, reduce hospital costs.

Recently, studies on the creation and use of more suitable protocols for improving hemostasis and reducing the use of blood products have gained popularity^[2]. Numerous risk factors have been described in the literature that predict postoperative bleeding after cardiac surgery. Preoperative use of antithrombotic agents and their cut-off times, some hematological diseases, comorbid diseases (diabetes mellitus [DM], hypertension [HT], chronic obstructive pulmonary disease [COPD], chronic kidney failure [CRF], etc.), new techniques, and equipment for cardiopulmonary bypass (CPB) are some of them^[3,4]. In addition, the WILL-BLEED Risk Scoring system was created using these risk factors to predict severe bleeding after CABG^[5].

In this study, we aimed to determine the relationship between plasma calcium level and C-reactive protein albumin ratio (CAR) and other demographic, clinical, and hematological markers in predicting severe bleeding after CABG.

METHODS

Statistical Analysis

Data were entered into the Statistical Package for the Social Sciences (IBM® SPSS Statistics for Windows, Version 23.0, Armonk, New York, United States of America) software package. Descriptive statistics were used, and quantitative variables were characterized using mean, maximum (max), and minimum (min) values; percentages were used for qualitative variables. Whether the distributions were normal or not was determined by Kolmogorov-Smirnov analysis. Normal distributions were reported as mean values. Student’s t-test was used for comparisons between groups. Pearson’s Chi-square test was used for comparative analysis of qualitative variables; however, Fisher’s exact test was used if the sample size was small (≤ 5). Nonparametric continuous variables were recorded as medians and compared using Mann-Whitney U tests. Interquartile range results were also given for the values recorded as median. P-value < 0.05 was considered statistically significant. Multivariate analysis was performed using the variables found to affect the drainage > 1000 ml in the univariate analysis. In the multivariate analysis, the area under the curve (AUC) was calculated by performing receiver operating characteristic (ROC) curve analysis for the nonparametric independent risk factors that were found to affect the drainage > 1000 ml (Figure 1). In addition, the best sensitivity and specificity values were accepted as the cut-off values.

Fig. 1 - Receiver operating characteristic curve analysis for parameters identified as independent risk factors in multivariate analysis.

Download

Fig. 1 - Receiver operating characteristic curve analysis for parameters identified as independent risk factors in multivariate analysis.

RESULTS

There were statistically significant differences between the groups with regards to CPB time, lymphocyte count, CRP, hemoglobin, calcium, albumin, and CAR. CPB time was longer in Group 2 than in Group 1. Lymphocyte count, hemoglobin, calcium, and albumin levels were higher in Group 1 compared to Group 2, while CRP levels and CAR were lower. The comparison between the groups in terms of demographic, clinical, and preoperative blood parameters is given in Table 1.

Table 1 - Grouping of patients according to the amount of drainage and comparison between the groups.

According to the multivariate analysis, the identified independent risk factors affecting postoperative severe bleeding were as follows: CRP (odds ratio [OR]=1.217, 95% confidence interval [CI]=1.020-1.452, P=0.02), calcium (OR=0.112, 95% CI=0.099-0.231, P<0.001), albumin (OR=0.695, 95% CI=0.497-0.972, P=0.03), and CAR (OR=831.0, 95% CI=5.844-11817.157), P=0.008) (Table 2). Afterwards, ROC curve analyzes were also performed to determine the optimum threshold values of the identified independent risk factors with sensitivity and specificity rates. It was observed that the parameter with the best AUC value was calcium (AUC=0.989, 95% CI=0.965-0.998). Threshold values were determined according to the best sensitivity and specificity values for all parameters (Table 3). Accordingly, patients were divided into groups with low and high values (Table 4).

Table 2 - Investigation of independent risk factors affecting drainage > 1000 ml by multivariate logistic regression analysis.

Table 3 - ROC curve analysis for parameters determined as independent risk factors affecting drainage > 1000 ml in multivariate analysis.

Table 4 - Grouping of patients according to the determined threshold values and comparison between these groups according to whether the drainage is > or < 1000 ml.

Group 1 patients were found to have statistically lower CRP values compared to Group 2 patients (OR=8.218, 95% CI=4.111-16.428, P<0.001). Group 2 patients were found to have lower calcium levels at a statistically higher rate than Group 1 patients (OR=0.003, 95% CI=0.001-0.013, P<0.001). Similarly, Group 1 patients were found to have albumin levels statistically higher than Group 2 patients (OR=0.109, 95% CI=0.048-0.245, P<0.001). Group 1 patients were found to have statistically lower CAR values than Group 2 patients (OR=12.670, 95% CI=6.110-26.274, P<0.001).

The comparison of the groups in terms of postoperative follow-up is given in Table 5. Statistically, Group 2 patients received more ES (P<0.001) and FFP (P=0.005) transfusions. Group 2 patients underwent more revisions (P=0.002) and also had longer ICU and hospital stays (P=0.007 and P=0.02, respectively). Statistically lower mortality was observed in Group 1 compared to Group 2 (P=0.01).

Table 5 - Comparison between the groups in terms of postoperative follow-up.

DISCUSSION

In this study, considering the hematological and biochemical parameters, the median lymphocyte, hemoglobin, calcium, and albumin values were found to be lower in the group with severe bleeding compared to the other group. In addition, the median CRP and CAR values were found to be higher in the group with severe bleeding compared to the other group. In multivariate analysis, CAR, CRP, calcium, and albumin values were found to be statistically significant, and it was concluded that these parameters independently predicted severe bleeding after CABG.

There are many factors affecting postoperative bleeding in open heart surgery. Hematological diseases, pharmacological agents used preoperatively, antiaggregant agents used perioperatively, and open heart surgery using CPB are some of them^[3,4]. Preoperative use of acetylsalicylic acid, clopidogrel, and other drugs may affect hemostatic functions and increase postoperative bleeding^[7,8]. In our study, aspirin and clopidogrel were discontinued seven days before surgery and warfarin five days before surgery in all patients. Low-molecular-weight heparin treatment was stopped 12 hours before surgery. Widespread microvascular bleeding may develop due to the use of CPB in coronary artery surgery. The use of heparin, development of systemic inflammation, and negative effects of hypothermia on the hemostatic system (decreased platelet number and dysfunction, fibrinogen level, and coagulation factor consumption, etc.) cause this situation^[3,9,10]. Contact of blood with CPB circuits induces coagulation activation. Thus, it causes more consumption of clotting factors and platelets in the circulation. In addition, while the crystalloid fluids used to expand volume in the CPB process cause dilution of clotting factors and platelets, the use of colloid fluids both inhibit platelet function and cause thrombocytopenia. As a result, it causes an increase in the amount of postoperative bleeding^[9,11].

Various risk factors associated with postoperative excessive bleeding following adult cardiac surgery were identified in a current integrative review study by Lopes et al^[3]. In that study, reviewing a total of 17 studies from seven databases, the predictors of severe bleeding after open heart surgery were classified as patient-related, procedure-related, and postoperative factors. Patient-related factors included male sex, DM, low BMI and left ventricular ejection fraction, high preoperative hemoglobin level, low preoperative platelet counts, and fibrinogen concentration, whereas perioperative-related factors included operating surgeon, CABG with three or more grafts, internal thoracic artery usage, increased cross-clamping, CPB, and total operation times, low intraoperative body temperature, and postoperative fibrinogen levels and metabolic acidosis. The authors consequently deduced that the mentioned predictors could be utilized for risk stratification of severe bleeding following open-heart surgery, and the evaluation of patients could be guided by knowing these factors, hence perioperative awareness can be prioritized. In addition, it was also expressed that timely determination and correction of the modifiable factors could be facilitated. With reference to this review study, only increased CPB time was identified as a common risk factor in both their and our studies. Furthermore, low preoperative hemoglobin level was defined as a risk factor in our study, whereas in the review study, on the contrary, high preoperative hemoglobin level was surprisingly expressed as a risk factor for severe bleeding after cardiac surgery.

Calcium plays an important role in the platelet aggregation and coagulation cascade. Thus, it takes part in ensuring hemostasis. Additionally, calcium is a cofactor in the enzymatic system, has an impact on the control of vasomotor tone, and has significant effects on the contractility of cardiac and striated muscle^[12,13]. In addition, in vitro studies have shown that there is a relationship between ionized calcium level and clot strength and concentration^[14]. Some studies have been carried out investigating the relationship between hypocalcemia and bleeding complications in different patient groups. In a systematic review by Vasudeva et al.^[13], it was reported that the amount of bleeding may be excessive because of coagulopathy developing due to hypocalcemia in adult patients with multitrauma, and this situation causes an increase in mortality due to massive blood transfusion. Epstein et al.^[15] showed that low calcium level in the postpartum period causes an increase in the amount of postpartum hemorrhage. In another study, they reported that low serum calcium level increased the size of hematoma in patients with intracerebral hemorrhage and therefore was associated with coagulopathy^[16]. In the ROC curve analysis of our study, we found that the parameter with the highest sensitivity and specificity in predicting severe postoperative bleeding was serum calcium level. Therefore, preoperative determination of serum calcium level may be important in predicting bleeding complications and taking necessary precautions.

CAR is developed as a novel sensitive marker that reflects the immune status of patients. Currently, predictive values of CAR have been widely studied in a wide range of diseases. Moreover, predictive values of CAR have also been examined for patients undergoing cardiac surgery. Karabacak et al.^[17] and Aksoy et al.^[18], in their studies to predict new-onset atrial fibrillation after CABG, showed that CAR indicates a higher inflammatory state and is better than CRP and albumin alone at detecting postoperative atrial fibrillation. Furthermore, CAR has been shown to be an independent predictor of saphenous vein graft disease and to have an association with atherosclerosis^[19,20]. Kahraman et al.^[21] described that CAR is strongly associated with rehospitalization rates, paravalvular leak, and increased mortality rates in patients undergoing aortic valve replacement due to isolated degenerative severe aortic stenosis. To the best of our knowledge, our study was the first to determine that CAR predicted severe postoperative bleeding in patients who underwent CABG, and there is no published research on the relationship between CAR and postoperative bleeding in the literature.

Albumin is a negative acute phase reactant that is inversely related to inflammation and oxidative stress. Hypoalbuminemia is associated with the occurrence of certain cardiovascular diseases. In patients with low albumin levels before CABG surgery, postoperative atrial fibrillation, acute kidney injury, slower recovery, readmission, and mortality rates were found to be high^[17,22]. Franco et al.^[23] showed that hypoalbuminemia is frequently seen after cardiac surgery, and low serum albumin levels increase the rates of sepsis, prolonged ICU stay, and in-hospital mortality due to bleeding complications. In addition, low serum albumin level is associated with endothelial damage, increased blood viscosity, and coronary artery narrowing due to platelet aggregation^[24]. As far as we know, our study was the first to determine that low albumin levels in the preoperative period predicted postoperative bleeding.

CRP is a major acute phase reactant that rises acutely and rapidly in stress, infection, and tissue damage. There are literature studies showing that it is predictive in many diseases. In a meta-analysis including 26 studies examining the prognostic value of coronary artery disease, high CRP level was reported to be an independent predictor of major adverse cardiovascular events, cardiovascular mortality, and all-cause mortality^[25]. In another meta-analysis in the literature, high CRP levels were reported to be a predictor of postoperative atrial fibrillation in patients undergoing coronary artery surgery^[26]. In our study, we determined that high CRP predicts severe postoperative bleeding in patients who underwent CABG.

CONCLUSION

To the best of our knowledge, the present study was the first one investigating whether serum calcium level and CAR, a novel biochemical marker, predicted severe bleeding after CABG. Our study demonstrated for the first time that higher serum CRP level and CAR and lower serum calcium and albumin levels were associated with severe bleeding after CABG, and these biochemical parameters could be useful for the prediction of postoperative severe bleeding following CABG. Further large-scale well-designed studies are required to support our findings and to obtain stronger scientific evidence.