Abstract
Objective:
Minimal invasive extracorporeal circuits (MiECC) have been associated with a significant reduction in the incidence of postoperative atrial fibrillation (AF). Nevertheless, AF remains one of the most common complications following elective primary coronary artery bypass grafting (CABG). The aim of this study was to identify the predictors of AF persisting beyond the hospital stay in elective primary CABG patients.
Methods:
We conducted a retrospective analysis for the predictors of AF that persisted beyond discharge between all patients who received an elective isolated CABG in our institution between 2009 and 2014. Patients with a positive history for intermittent or persistent AF were excluded from the analysis. Almost all patients were discharged to a rehabilitation facility where they stayed for 3 to 4 weeks postoperatively. At rehab approximately 91% of them received Holter monitoring at least once prior to their discharge.
Results:
A total of 770 patients were included in the analysis of which 763 patients survived the in-hospital stay. The incidence of AF at hospital discharge was 4.2% (32/763) while that on Holter monitor at Rehab was 1.5% (10/685). Age and the type of extracorporeal circuit (ECC) utilized were the only significant predictors for both AF at discharge (p < 0.01 both) and on Holter monitor in rehab (p < 0.01 and 0.02, respectively). This was also confirmed on multivariate analysis.
Conclusion:
Our findings show that the benefits of MiECC regarding the incidence of postoperative AF persist beyond hospital discharge. They may thus positively influence the outcomes of patients beyond the early postoperative period.
Keywords
Introduction
The term minimal invasive extracorporeal circuits (MiECC), was recently introduced 1 to emphasize the role these circuits play in preserving homeostasis during cardiopulmonary bypass. Their positive impact on the incidence of postoperative atrial fibrillation (AF) has been shown in a multi-centric prospective randomized trial 2 and confirmed by a recently published meta-analysis 3 comparing the concept to conventional extracorporeal circuits. Nevertheless, AF remains one of the most frequent postoperative complications of elective primary coronary artery bypass grafting (CABG). Furthermore, its persistence beyond discharge has been shown to have a negative impact on long term patient outcome. It was hence our aim to identify the predictors of AF persisting beyond discharge following elective primary CABG.
Materials and methods
We conducted a retrospective single center analysis of 770 consecutive elective patients who underwent primary isolated CABG in our institution from 2009 to 2014. This time period was chosen as during that period a standardized protocol for prophylaxis against AF had been implemented and as we still did not have enough MiECC systems to cover 100% of our CABG patients, hence there were still enough patients undergoing elective CABG with a conventional extracorporeal circuit (CECC) to be included in the analysis.
The local ethics committee (Medizinische Hochschule Hannover Nr. 8485_BO_K_2019) approved the study. Individual patient’s consent was waivered because of the study’s retrospective design and data collection from routine care. Excluded were patients with a history for AF or supraventricular arrhythmias.
Procedural strategy
Surgery was performed through a median sternotomy. All circuits were primed solely with crystalloid solution as retrograde autologous priming was employed in most patients. Myocardial protection was achieved with intermittent warm blood cardioplegia in the MiECC (Type IV) patients and Bretschneider cardioplegia in the conventional ECC patients. Further details of the ECC were published elsewhere. 2
Anesthesiologists, surgeons and perfusionists as well as the anesthetic management were the same for all patients.
Postoperative management of atrial fibrillation
Prophylaxis against AF included continuing the patient́s beta blocker up to the morning of the operation, followed by metoprolol 50 mg twice daily (once the patients were hemodynamically stable) up to discharge. Additionally, patients received magnesium 5 mmol per.os (p.o.) 1 day prior to surgery, 12.6 mmol/24 h intravenously via an infusion pump for the first 48 h after surgery and 5 mmol p.o. daily up to the fifth postoperative day. Furthermore, potassium was substituted to keep the serum potassium level above 4 mmol/l.
An electrocardiogram (ECG) was recorded on arrival to the Intensive care unit (ICU), on the first postoperative day, at discharge and when any arrhythmia was detected. During their ICU stay and in the intermediate care unit the patients had a continuous monitoring of their ECG.
Our standard treatment for new-onset AF included raising the postoperative metoprolol dose as well as a single intravenous dose of 5–10 mg metoprolol and/or 150–300 mg of Amiodaron depending on the patient’s condition and the preference of the attending physician. If the AF persisted for more than 24–48 h. An electric cardioversion was attempted (rhythm control being the institutional strategy for treatment of AF during that time period) before the patient was discharged from hospital.
AF persisting at discharge was defined according to the following electrocardiographic feature in the ECG performed 1 day prior to patient discharge: irregular R-R intervals when atrioventricular conduction is present, absence of distinct and repeat P wave and irregular atrial activity.
Post-discharge monitoring of AF
It is general policy in Germany that all the patients who undergo CABG are discharged to a rehabilitation facility where they stay for another 3–4 weeks before they are discharged. During rehabilitation there is further monitoring of their rhythm by repeated standard ECG but also by Holter monitoring in most facilities.
Study hypothesis and research question
The hypothesis was that postoperative AF results from a combination of predisposing factors that are intrinsic to the patient and triggering factors that the patient is subjected to perioperatively. However, as it is well known that most episodes of postoperative AF are either self-limited or resolve under therapy within the first few days, most patients are discharged from hospital with a stable sinus rhythm. We therefore postulated that the magnitude of the combined effect of the predisposing and the triggering factors could play a role in the incidence of new onset postoperative AF that persisted beyond discharge and questioned if the type of ECC utilized might play a role in this setting.
Data collection and statistical evaluation
Patient data were retrospectively extracted from the department’s database and medical records. The patients’ discharge letter from the rehabilitation facilities are also stored in these records. Variables were defined according to the European System of Cardiac Operative Risk Evaluation (EuroSCORE). 4
Results are reported as mean ± standard deviation or percentages as appropriate. Continuous variables were compared by unpaired Student’s t-test while Fisher’s exact test was used to compare discrete variables. For all statistical tests, a p value of <0.05 was deemed statistically significant.
Binary logistic regression using a forward stepwise selection method was carried out to test for independent predictors of AF. The model included all variables with a p value of <0.2 on univariate analysis. We performed a receiver operating characteristic (ROC) analysis and reported the area under the ROC curve for this model. The corresponding confidence interval was computed using the DeLong method. 5
To compare the results between both groups of ECC we applied a 1:2 nearest neighbor propensity score matching considering ECC time and age. A maximum caliper width of 0.2 times the standard deviation of the logit of the propensity score was used for matching the groups.
All analyses were performed with the SPSS (Version26) statistical software (IBM Corporation, Armonk, New York, USA) as well as the statistical programming environment R (version 3.6.0) using the R-packages MatchIt (version 3.0.2, Ho et al. 6 ), tableone (version 0.12.0), and pROC (version 1.15.3, Robin et al. 7 ) for propensity score matching, the computation of the SMD and the ROC analysis, respectively.
Results
770 Patients with a preoperatively documented sinus rhythm (SR) were included in the study, of those 7 died in-hospital and were excluded from the analysis. Among the 763 survivors whose data are presented in Table 1, 32 patients (4.2%) were discharged with an AF. Supplemental Table S1 depicts the differences among patients discharged in SR and those discharged with an AF. It shows that those in the AF group were significantly older (71 ± 8 years vs 66 ± 9 years; p <0.01) and were more frequently operated by a conventional extracorporeal circuit (CECC) namely (41% vs 17%; p = 0.01). Furthermore, the incidence of AF at discharge was more than threefold higher among those patients who underwent cardiopulmonary bypass with a CECC (9.4%) than those in whom a MiECC was utilized (3.0%).
Demographics, perioperative characteristics, and results.
MiECC: minimal invasive extracorporeal circuit; CECC: conventional extracorporeal circuit; COPD: chronic obstructive pulmonary disease; IDDM: insulin dependant diabetes mellitus; LVEF: left ventricular ejection fraction; ECC: extracorporeal circulation; AF: atrial fibrillation; Renal dysfunction, pulmonary hypertension and neurological dysfunction are defined in accordance with the EuroSCORE 1.
based on a subgroup analysis for the 685 patients who received Holter monitoring during rehab. Data presented as mean ± standard deviation unless otherwise mentioned.
Of the 763 survivors 685 patients (90%) received a Holter monitoring at least once during their stay in a rehabilitation facility (about 4 weeks postoperatively). In 10 of those patients (1.5%) AF was still recorded. It is worth mentioning that all patients who were discharged with AF had a Holter monitoring at least twice during their stay at rehab. 48 patients who went to rehab had no Holter monitoring but at least two ECGs during their stay. Hence 723 patients (95%) had a follow up with a documentation of their rhythm during rehab, of those 12 patients (1.7%) still had an AF beyond the 30th postoperative day. Comparing the patients in AF with those in SR during Rehab (Supplemental Table S2) shows that age and the type of extracorporeal circuit utilized during their CABG remain the only statistically significant differences between both groups. Again the incidence of AF on Holter monitoring was more than three-fold higher among patients in whom a CECC was used (4% vs 0.9%; Table1).
A comparison between the patient operated by both types of ECC (Table1) revealed that although both groups showed statistically significant differences in both the incidence of AF on discharge as well as on Holter monitor during their rehabilitation they also showed significant differences in both age and bypass time. Following propensity score matching the differences regarding the incidence of AF on discharge as well as on Holter monitoring however remained significantly different (Table 2).
Demographics, perioperative characteristics, and results propensity score matched patients.
MiECC: minimal invasive extracorporeal circuit; CECC: conventional extracorporeal circuit; PSM: propensity score matched; COPD: chronic obstructive pulmonary disease; IDDM: insulin dependent diabetes mellitus; LVEF: left ventricular ejection fraction; ECC: extracorporeal circulation; AF: atrial fibrillation.
Renal dysfunction, pulmonary hypertension, and neurological dysfunction are defined in accordance with the EuroSCORE 1.
Data presented as mean ± standard deviation unless otherwise mentioned.
In a sub-analysis depending on the extracorporeal circuits (ECC) utilized, age (p = 0.03) was the only predictor for AF persisting beyond discharge among the 639 patients in the MiECC group while both age and ECC time (p = 0.03 and 0.02, respectively) were the only predictors among the 138 patients in the CECC group (Supplemental Table S3).
Multivariate analysis for the predictors of AF persisting at discharge, on Holter monitoring and at rehab as a whole, all confirm age and the type of circuit utilized as independent risk factors (Table 3). Further, they confirm extracorporeal circulation time as an independent predictor of AF at discharge for the patients in whom a CECC was used. The discriminatory ability of the prediction model for predicting postoperative atrial fibrillation persisting at discharge was moderate, with an AUC of 0.68 (95% CI: 0.57–0.78; Figure 1).
Independent predictors of AF persisting at discharge and in rehab.
AF: atrial fibrillation; ECC: extracorporeal circulation; CECC: conventional extracorporeal circulation.
ROC curve for postoperative atrial fibrillation persisting at discharge after elective coronary artery bypass grafting, based on preoperative and surgical data.
Discussion
MiECC were concepted to minimize the adverse effects attributed to ECC. Over more than two decades studies have evolved to support this hypothesis. However, as most of these studies involved low patient numbers their impact on the guidelines supporting the use of MiECC is limited.8,9 On the other hand, a prospective randomized multicenter trial published in 2011 2 showed that among 500 patients who received a CABG and/or an aortic valve replacement (AVR), there was a significantly lower incidence of postoperative major adverse events as well as a lower incidence of new onset atrial fibrillation (NOAF) (16.3% vs 24.2%; p = 0.03) among MiECC patients. In addition, the type of circuit employed was the only modifiable independent risk factor for sustaining AF (Odd ratio of 1.7), while age and the type of operation were the only other independent risk factors. A meta-analysis of prospective randomized trials comparing MiECC to CECC published in 2015 3 confirmed a reduced incidence of AF among MiECC patients (19.2% vs 27.6%; p <0.001; RR of 0.76).
In a subgroup analysis of an observational study 10 following the concept of utilizing a modular MiECC as a universal heart lung machine for all commers we could show that among 64 patients who underwent elective, urgent und emergent CABG and or AVR only 16% sustained AF and that all of them were on SR—demonstrated by Holter monitoring—during their postoperative stay at a rehabilitation facility. It was this finding that gave us the idea for the current study.
Our study shows that the age and type of circuit employed were predictive for the incidence of AF persisting beyond discharge and during Rehabilitation. Further, we could demonstrate that the incidence of AF persisting beyond discharge and at rehabilitation was more than three-fold higher among the patients who were operated with a CECC. As age has been one of the most constantly reported predictors of AF in the literature this finding is not surprising. However, to our knowledge this is the first report showing that the type of the ECC employed during CABG had an influence on the incidence of AF persisting beyond discharge.
Contrary to the literature 11 in which the bypass time has been another frequently mentioned predictor of postoperative AF, we could not confirm it as a predictor for AF persisting beyond discharge in our cohort. This led to a sub-analysis for the predictors of AF depending on the type of circuit utilized, which showed that the duration of cardiopulmonary bypass was a predictor for AF persisting on discharge only for the patients in whom a CECC was employed. This finding supports the idea that MiECC confer their benefits through a better preservation of homeostasis thereby allowing for longer exposure of the patients to the circuits without triggering a pathophysiology that may induce the arrythmia.
The fact that different types of cardioplegia were used in both groups is most probably the biggest leanness of this study but is a shortcoming of almost all retrospective studies in this field. At the same time the largest randomized controlled trial 12 (1440 consecutive CABG patients) comparing blood cardioplegia to crystalloid cardioplegia has found no significant differences in the incidence of postoperative AF between both groups, not even in a subgroup analysis of high-risk patients. Furthermore, the largest prospective randomized study comparing MiECC to CECC has shown significant differences in the incidence of AF although the same type of cardioplegia was utilized in bothpatient groups. 2
In a prospective randomized trial 13 evaluating the preventive value of bi-atrial pacing on the incidence of late-phase postoperative AF up to 30 days after CABG or AVR the prevalence of AF lay approximately between 5% and 11% at day 7–10 postoperatively and around 5% at 30 days postoperatively in the group of patients who did not receive bi-atrial pacing. These findings are quite similar to the results found in our CECC group with 9% and 4% respectively.
In an analysis of time-related parametric risk factors for NOAF after heart surgery Melby et al. 14 identified different risk factors for different time intervals following surgery with a Phase I peaking immediately after the procedure and a Phase II peaking 48 h after the procedure, in both intervals age was one of the strongest risk factors. A further finding of that study was that cross-clamping time only played a significant role in the early phase. It is hence not surprising that age was confirmed as one of the main predictors in our study at the different time points as well as in the subgroup analysis. It supports the hypothesis that postoperative AF is related both to factors that are intrinsic to the patient and triggered pathophysiological changes that may be initiated by the operation.10,15 The notion that the MiECC is less invasive and hence constitutes a milder trigger could be assumed from our findings and could explain why AF persists more frequently when a conventional ECC is used. Whether this is related to milder inflammatory response, a milder ischemic-reperfusion injury or other mechanisms however remains to be shown.
Limitations
The retrospective nature of this study compromises its value. However, the variables of main interest in this study namely age, type of extracorporeal circuit used during CABG, duration of cardiopulmonary bypass, AF persisting at discharge, Holter monitoring or ECG at rehab are well documented variables that are seldomly misinterpreted or incorrectly documented. Furthermore, neither the patients nor the staff at the rehabilitation facility knew which extracorporeal system had been used so that this could not have biased the results.
An additional limitation lies in the fact that the timing of the discharge as well as the Holter monitoring were not uniform for all patients however, one has to point out that the population involved were all low risk patients presenting for primary elective CABG and that only survivors were involved so that the time points were presumably quite similar.
Conclusion
This study confirms that the benefits of MiECC regarding the incidence of AF persists beyond hospital discharge and may thus positively influence the outcomes of patients beyond the early postoperative period.
Supplemental Material
sj-pdf-1-prf-10.1177_0267659120978647 – Supplemental material for Predictors of postoperative atrial fibrillation persisting beyond hospital discharge after coronary artery bypass grafting
Supplemental material, sj-pdf-1-prf-10.1177_0267659120978647 for Predictors of postoperative atrial fibrillation persisting beyond hospital discharge after coronary artery bypass grafting by Aschraf El-Essawi, Ahmed Abdelhalim, Steffen Groeger, Ingo Breitenbach, Rene Brouwer, Fabian Kück and Wolfgang Harringer in Perfusion
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Aschraf El-Essawi: Speaker’s honoraria and consultation for Terumo and Medtronic. The other authors have none to declare.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
