In recent years, the nomogram has been used extensively in clinical research, and shows more accurate advantages than the traditional scoring system, which can predict the prognosis of some diseases(9). In view of the importance of POAF in the prognosis of patients, there is great clinical interest in preventing this arrhythmia. However, undifferentiated conventional drug treatment will cause drug side effects in 60%-80% of patients(10). Therefore, it is particularly important to quantify the risk of POAF and clarify the indications of common antiarrhythmic drugs.
Present study validated the ability of novel nomogram for predicting the incidence of POAF in patients undergoing OPCAB. In previous researchers, the risk factors for POAF included age, diabetes, cerebrovascular accident, PCI history, chronic heart failure, left ventricular ejection fraction, left atrial enlargement, and electrolyte disturbance(11, 12). Our research indicated that age, diabetes, hypertension, previous PCI, and intraoperative potassium were related to the occurrence of POAF. And the authors compared it with C2HEST, CHADS2, and CHA2DS2-VASc scores. According to the DCA curves, the novel nomogram performed better than the C2HEST, CHADS2, and CHA2DS2-VASc scores. This nomogram demonstrated a predictive performance with a good discriminative ability (C-index in training and validation cohort respectively: 0.809 and 0.886) and calibration.
The C2HEST score has shown promising results as a simple practical tool for predicting incident AF based on clinical risk factors(13, 14). Correspondently, serious studies found that CHADS2 and CHA2DS2-VASc score, were independent predictors of POAF(15–17). The clinical practice of these risk scores has been proven in different participants, however, the population of these studies concentrated on inhomogeneous patients undergoing different types of cardiac surgery. And an easy-to-remember bedside tool is especially needed in a busy clinical setting. Therefore, the authors development a novel nomogram to visualize scores items. To our knowledge, as of the writing of this article, this study was the first to construct a quantitative nomogram to predict the probability of POAF in a Chinese population undergoing isolated OPCAB.
According to researchers, apart from a history of AF, age is the single best predictor for AF, regardless of surgical or non-surgical AF(18, 19). Degeneration of the atrial myocardium with aging may lead to a loss of side-to-side electrical coupling between muscle fibers, slowing down electrical conduction of the sinoatrial and atrioventricular nodes and atria, thereby providing an anatomic or electrophysiologic substrate for arrhythmogenesis(20). This study also confirmed that diabetes is an independent risk factor for predicting POAF. Earlier studies showed that diabetes has increased the risk of atrial fibrillation by 35%(21). The underlying mechanism of atrial fibrillation induced by diabetes is that the increase in blood glucose results in the transformation of epicardial adipose tissue, inducing endothelial dysfunction and myocardial fibrosis induced by proinflammatory mediators, leading to atrial remodeling and electrophysiological remodeling(22).
The current research results show that hypertension was associated with POAF. Notably, hypertension may induce atrial stretch and cardiomyopathy, which ultimately lead to structural and electrophysiological remodeling conducive to POAF(23). Hypertension, fluid overload, pathological activation of the renin-angiotensin-aldosterone system (RAAS) with subsequent enhanced myocardial fibrosis may be the underlying mechanisms by which renal dysfunction increases the likelihood of AF(24).
Deepak L. Bhatt and others believed that if patients are suitable for PCI or CABG at the same time, PCI will lead to a higher rate of repeated revascularization, and the probability of repeat PCI or CABG will be greatly increased, which will increase the incidence of postoperative complications(25). This study found that previous percutaneous coronary intervention (PCI) was an independent risk factor for OPCAB-POAF. According to the review of literature, the possible reasons are as follows: First, most patients with branch PCI have right coronary artery lesions, and the right coronary artery is the main blood supply source of the sinoatrial node and right atrium. PCI is not conducive to the physiology of coronary artery- and distal endothelial-mediated vasodilation, which easily causes myocardial ischemia and induces arrhythmia (26). Secondly, the incidence of myocardial infarction and coronary restenosis after PCI is 1.5%~6%. Insufficient myocardial blood supply easily leads to cardiac rhythm disorders and atrial fibrillation(26). Thirdly, most patients with PCI are more likely to have deterioration of cardiac function after repeat CABG. Heart failure increases CABG of the left ventricle and the pressure load of the left atrium. These pathophysiological changes may be the inducing factors of OPCAB-POAF in patients after PCI(3).
It is well known that blood potassium concentration plays an important role in cardiac electrophysiology. Intraoperative blood potassium concentration is related to adverse prognosis such as cardiovascular events(27). This study showed that low blood potassium concentration in the last intraoperative blood collection was an independent risk factor for OPCAB-POAF. The blood potassium index was obtained by blood gas analysis of the patient's arterial blood for the last time before the patient left the operating room. Raymond et al. clarified that low serum potassium concentration is related to the risk of atrial fibrillation. Serum K+ concentration is usually low in patients undergoing cardiac surgery, and the incidence of atrial fibrillation increases when K+ concentration is lower than 3.5 mmol/L(28). It is worth noting that our study found that even if the blood potassium concentration of most patients remained at 3.5-4.0 mmol/L before leaving the operating room, the probability of postoperative POAF is still very high. Combined with clinical and previous literature analysis, the possible reasons are as follows: First, the blood potassium level changes dynamically. Cardiac patients continue to pump dopamine and adrenaline after the operation, which further reduce blood potassium. Second, because patients need to continue mechanical ventilation for a period after cardiac surgery, they cannot consume food for an extended period, and the maintenance of blood potassium depends on intravenous potassium supplementation. If blood potassium is not detected and intravenous potassium supplementation is not timely after cardiac surgery, the potassium levels will continue to decline. Third, the arterial blood potassium concentration is inconsistent with that of myocardial extracellular fluid, and the value obtained by blood gas analysis is usually low. Many centers around the world believe that efforts should be made to maintain the serum K+ concentration in the “normal high” range (4.5–5.5 mmol/L), rather than intervene only when potassium drops below its “normal” lower threshold. Although there is no evidence that this association is causal, it is still considered as a routine method in many centers to prevent atrial fibrillation(18, 29). In addition, abnormal electrical remodeling of K+ channels can change the electrophysiological characteristics of the atria, change the duration and characteristics of the refractory period, and induce ectopic pacing or arrhythmia of atrioventricular cells(30); thus, it indirectly shows the effect of potassium ions on cardiac electrophysiology and arrhythmia.
Limitations: First, the retrospective study is not randomized but naturally grouped according to different factors related to levels of exposure, resulting in an imbalance between groups and bias. Second, the case data of the nomogram established in this study was derived from only a single center, and the sample sizes were limited, which makes the statistical results unreliable and produces a large standard error. Therefore, data acquired from more research centers and from larger sample sizes are needed to further demonstrate the model in the future. Third, this study is a retrospective study, and there are limitations in data collection. In the future, it is necessary to design a reasonable prospective study to demonstrate the model. Fourth, this study uses an internal validation method to verify the model. However, the performance of models that perform well in a single case data set is not necessarily satisfactory in other data sets. Therefore, future prospective cohort studies are needed to externally verify the prediction model in a new case data set.