Prophylaxis for patients at Risk to Eliminate Post-operative Atrial Fibrillation (PREP-AF trial): a protocol for a feasibility randomized controlled study

The long-term impact of postoperative atrial fibrillation after cardiac surgery

OBJECTIVE

The literature has reported worse in-hospital outcomes for patients with atrial fibrillation. The objective of the following study is to provide detailed results on the long-term impact of postoperative atrial fibrillation on survival and hospital readmission in cardiac surgery.

METHODS

All patients undergoing open cardiac surgery were reviewed with the exclusion of preoperative atrial fibrillation or patients undergoing ventricular assist device, transplant, or Cox-Maze procedures. Propensity matching (1:1) was performed to ensure similar baseline characteristics. Multivariable analysis identified significant associations with mortality and readmission.

RESULTS

A total of 12,227 patients with cardiac disease were divided into 7927 patients (64.8%) without postoperative atrial fibrillation and 4300 patients (35.2%) with new-onset postoperative atrial fibrillation. Propensity matching (1:1) yielded 4275 risk-adjusted pairs. There was no difference between the nonpostoperative atrial fibrillation versus postoperative atrial fibrillation cohorts regarding operative mortality (4.61% vs 4.12%; P = .26) and stroke (2.32% vs 2.76%; P = .191). Patients with postoperative atrial fibrillation had higher rates of reoperation (12.12% vs 6.83%; P < .001), transfusion (43.42% vs 36.94%; P < .001), sepsis (1.99% vs 0.80%; P < .001), prolonged ventilation (15.88% vs 9.24% vs; P < .001), pneumonia (6.60% vs 2.36%; P < .001), renal failure (6.90% vs 3.37%; P < .001), and dialysis (4.94% vs 2.08%; P < .001). The postoperative atrial fibrillation cohort had a significantly higher incidence of atrial fibrillation on follow-up (11.74% vs 4.75%; P < .001). Postoperative atrial fibrillation was independently associated with mortality (hazard ratio, 1.21; 1.12-1.33; P < .001), all-cause readmissions (hazard ratio, 1.05; 1.01-1.1; P = .010), and heart failure-specific readmission (hazard ratio, 1.14; 1.04-1.26; P = .01).

CONCLUSIONS

Patients in the postoperative atrial fibrillation cohort had worse perioperative morbidity, lower survival, and more readmissions for heart failure on long-term follow-up.

Association of Mitral Regurgitation with Postoperative Atrial Fibrillation in Critically Ill Noncardiac Surgery Patients: A Prospective Cohort Study

Purpose

Atrial fibrillation (AF) is common in critically ill patients and can have serious consequences. Postoperative AF (POAF) in critically ill patients following noncardiac surgery has been understudied, contrary to cardiac procedures. Mitral regurgitation (MR) is associated with left ventricular dysfunction, which might contribute to the occurrence of AF in postoperative critically ill patients. This study aimed to investigate the association between MR and POAF in critically ill noncardiac surgery patients and establish a new nomogram for the prediction of POAF in critically ill noncardiac surgery patients.

Patients and Methods

A prospective cohort of 2474 patients who underwent thoracic and general surgery was enrolled in this study. Data on preoperative transthoracic echocardiography (TTE), electrocardiogram (ECG), and several commonly utilized scoring systems (CHA2DS2-VASc, HATCH, COM-AF, HART, and C2HEST) and baseline clinical data were collected. Independent predictors were selected by univariate and multivariable logistic regression analysis, and a nomogram was constructed for POAF within 7 days after postoperative intensive care unit (ICU) admission. The ability of the MR-nomogram and other scoring systems to predict POAF was compared by receiver operator characteristic (ROC) curve analysis and decision curve analysis (DCA). Additional contributions were evaluated by integrated discrimination improvement (IDI) and net reclassification improvement (NRI) analysis.

Results

A total of 213 (8.6%) patients developed POAF within 7 days after ICU admission. Compared to CHA2DS2-VASc, HATCH, COM-AF, HART, and C2HEST scoring systems, MR-nomogram showed better predictive ability for POAF with an area under the ROC curve of 0.824 (95% confidence interval: 0.805-0.842, p < 0.001). The improvement of the MR-nomogram in predictive value was supported by NRI and IDI analysis. The net benefit of the MR nomogram was maximal in DCA.

Conclusion

MR is an independent risk factor of POAF in critically ill noncardiac surgery patients. The nomogram predicted POAF better than other scoring systems.