The CHADS 2 (congestive heart failure, hypertension, age >75 years, diabetes mellitus, stroke or transient ischemic attack [2 points]) scoring scheme has been found to be a good predictor of stroke risk in patients with nonvalvular atrial fibrillation (AF). However, the value of the CHADS 2 scoring system in the risk stratification of patients with AF who undergo direct-current cardioversion has not yet been specifically investigated. In this study, a subgroup of 541 patients from the Assessment of Cardioversion Using Transesophageal Echocardiography (ACUTE) study who had AF for >48 hours and planned to undergo transesophageal echocardiography before direct-current cardioversion were enrolled. Each patient had a CHADS 2 score calculated. Of the patients with CHADS 2 scores of 0, 14 (10%) were found to have left atrial appendage thrombi on transesophageal echocardiography. After 6 months of follow up, patients with CHADS 2 scores of 3 to 6 showed a significantly higher mortality rate in comparison with patients with lower CHADS 2 scores (4.3% vs 0.5%, p = 0.004), despite their similar prevalence of left atrial appendage thrombus and stroke (thrombus: 13.4% vs 11.6%, p = 0.60; stroke: 0% vs 0.3%, p = 0.70). In conclusion, the CHADS 2 scoring system may be useful for predicting short-term mortality risk in patients with AF receiving elective direct-current cardioversion. However, in the preprocedural risk assessment of these patients, the CHADS 2 scoring system is not reliable in predicting risk for left atrial appendage thrombus formation, especially in patients with low CHADS 2 scores.
Recently, the CHADS 2 scoring system was found to be predictive of some echocardiographic findings, namely intracardiac thrombus, in patients with atrial fibrillation (AF) before pulmonary vein isolation (PVI). Therefore, a preprocedural CHADS 2 score may aid in the risk stratification of patients in AF who undergo direct-current cardioversion (DCC) and possibly correlate with the risk for intracardiac thrombus, thus allowing clinicians to use transesophageal echocardiography (TEE) selectively in high-risk patients before the procedure. In this study, we evaluated the role of a preprocedural CHADS 2 score in quantifying the risk for stroke, rate of mortality and the routine need for transesophageal echocardiographic screening before DCC.
Methods
Of the 1,074 patients enrolled from 1994 to 1999 in the initial Assessment of Cardioversion Using Transesophageal Echocardiography (ACUTE) study, 544 were randomized to the TEE arm and underwent this procedure before cardioversion. Of the resulting 544 patients, completed CHADS 2 scores were available in 541; therefore, the other 3 patients were excluded from this study. The CHADS 2 score consists of congestive heart failure (1 point), hypertension (1 point), age >75 years (1 point), diabetes mellitus (1 point), and history of stroke or transient ischemic attack or embolic event (2 points), and all the data in this regard were collected retrospectively. This study was approved by the institutional review board, and written informed consent was obtained from all patients.
Patients enrolled in the ACUTE trial were >18 years of age and had AF for ≥48 hours. Patients with atrial flutter who had documented histories of AF were also eligible and included in this study. Patients were excluded from the study if they had pure atrial flutter or other supraventricular tachycardias, were anticoagulated for >7 days, were hemodynamically unstable, had contraindications to TEE or warfarin, were women of childbearing potential, or needed anticoagulation to be discontinued because of an elective procedure. These criteria follow the general recommendations of American College of Chest Physicians at the time for patients with AF >2 days in duration who elected to undergo electrical cardioversion.
All patients selected underwent preprocedural TEE and subsequent cardioversion after index TEE. Patients in the TEE-guided arm of the study were started on warfarin or intravenous heparin at enrollment and underwent TEE and cardioversion upon maintenance of anticoagulation in the therapeutic range. After cardioversion, all patients were monitored weekly and maintained on 4 weeks of therapeutic warfarin.
Transesophageal echocardiographic findings noted were as follows: left atrial appendage (LAA) spontaneous echocardiographic contrast was defined as “smokelike” echoes with swirling motion seen at an optimal gain setting during the cardiac cycle. Thrombus was defined as a circumscribed and uniformly echo-dense intracavitary mass distinct from the underlying left atrium or LAA endocardium and pectinate muscles that was present in >1 imaging plane. All patients with thrombi were confirmed using an echocardiographic core laboratory.
The primary end point was a composite of stroke, transient ischemic attack, and systemic embolism. The secondary end point included a composite of major and minor bleeding, mortality, and successful return to and maintenance of sinus rhythm. Any bleeding complication that was fatal, needed transfusion, or could not be terminated without surgical procedure was considered major. Because previous studies have demonstrated that DCC-related embolic events are at highest risk in the initial 4 weeks after the procedure, a study period of 8 weeks was chosen to cover the period of atrial “stunning” and early recurrent AF. There was also a 6-month follow-up period to further assess outcomes such as embolic events, mortality, and recurrence of AF.
Descriptive statistics, including means, percentiles, and SDs, were used for the continuous variables in this study using SAS version 9.1.3 (SAS Institute Inc., Cary, North Carolina). Continuous variables were analyzed using simple linear regression, and the p value reported is the t statistic for significance of the regression coefficient on CHADS 2 . For the categorical variables, descriptive statistics including frequencies and percentages were used. Categorical variables were analyzed using the chi-square test, and the p value reported is for the Mantel-Haenszel chi-square test. Patients with high CHADS 2 scores (≥3) were combined into 1 group, compared to low CHADS 2 scores, and analyzed using Fisher’s exact test for primary and secondary end point statistics. A p value ≤0.05 was considered statistically significant.
Results
We analyzed 541 patients in this study, and patient characteristics are listed in Table 1 . CHADS 2 score points were assigned mostly for hypertension in 288 patients (53%) and congestive heart failure in 153 patients (28%) ( Table 1 ). The duration of AF before the index TEE, on the basis of patient history report, was close to 13 days on average (interquartile range 4 to 48).
| Characteristic | Value |
|---|---|
| Age (years) | 64.5 ± 13.4 |
| Men | 361 (67%) |
| Heart failure | 153 (28%) |
| Hypertension | 288 (53%) |
| Diabetes mellitus | 77 (14%) |
| History of stroke, transient ischemic attack, or systemic embolic events | 48 (8.9%) |
| CHADS 2 score | |
| 0 | 138 (26%) |
| 1 | 198 (37%) |
| 2 | 123 (23%) |
| 3 | 60 (11%) |
| 4–6 | 22 (4%) |
| Left atrial/LAA spontaneous echo contrast | 294 (54%) |
| Left atrial/LAA thrombus | 64 (11.8%) |
| AF | 468 (87%) |
| Atrial flutter | 27 (5%) |
| AF and atrial flutter | 32 (5.9%) |
| Paced with underlying atrial fibrillation | 12 (2.2%) |
| Aortic atheroma | |
| Aorta evaluated | 140 (26%) |
| Atheroma identified | 103 (74%) |
| New York Heart Association functional class | 413 (76%) |
| 0 | 22 (5.3%) |
| I | 165 (40%) |
| II | 134 (32%) |
| III | 78 (19%) |
| IV | 14 (3.4%) |
LAA thrombi on TEE were seen in 14 patients (10%) with CHADS 2 scores of 0. The higher risk group with CHADS 2 scores ≥2 accounted for 30 (47%) of all the LAA thrombi detected in this cohort of patients ( Table 2 ). Interestingly, the prevalence of LAA thrombus was not significantly different among patients with different CHADS 2 scores (p = 0.52). It is worth noting that 66.8% of patients received heparin infusion at the time of TEE because they were not being appropriately anticoagulated (international normalized ratios of 2 to 3).
| Variable | Total Events | CHADS 2 Score | ||||
|---|---|---|---|---|---|---|
| 0 (n = 138) | 1 (n = 198) | 2 (n = 123) | 3 (n = 60) | 4–6 (n = 22) | ||
| LAA thrombus | 64 | 14 (10%) | 20 (10%) | 19 (15%) | 9 (15%) | 2 (9.1%) |
| At 8 weeks | ||||||
| Major/minor bleed | 15 | 3 (2.2%) | 4 (2%) | 3 (2.4%) | 5 (8.3%) | 0 |
| Stroke/transient ischemic attack | 5 | 0 | 4 (2%) | 1 (0.8%) | 0 | 0 |
| Death | 13 | 0 | 7 (3.5%) | 2 (1.6%) | 4 (6.7%) | 0 |
| At 6 months | (n = 111) | (n = 163) | (n = 106) | (n = 47) | (n = 16) | |
| Normal sinus rhythm | 270 | 71 (64%) | 102 (63%) | 60 (57%) | 28 (60%) | 9 (56%) |
| (n = 114) | (n = 176) | (n = 114) | (n = 52) | (n = 18) | ||
| Stroke/transient ischemic attack | 1 | 0 | 1 (0.6%) | 0 | 0 | 0 |
| Death | 5 | 0 | 2 (1.1%) | 0 | 0 | 3 (17%) |
The incidence of thromboembolic events, namely stroke and transient ischemic attack, at 8 weeks and 6 months was relatively similar among all groups (p = 0.30). Patients with CHADS 2 scores ≤2 had a low incidence of stroke (0.3%) and death (0.5%) at 6 months. The rates of LAA thrombus and stroke were similar in patients with CHADS 2 scores ≤2 and ≥3. However, patients with CHADS 2 scores of 3 to 6 had a significantly higher rate of all-cause mortality compared to patients with CHADS 2 scores of 0 to 2 (4.3% vs 0.5%, p = 0.004; Table 3 ). Patients with higher CHADS 2 scores were more likely to experience bleeding complications than patients with lower CHADS 2 scores (6.1% vs 2.2%, p = 0.05).
| Variable | Total Events | CHADS 2 Score | p Value ⁎ | |
|---|---|---|---|---|
| 0–2 (n = 459) | 3–6 (n = 82) | |||
| LAA thrombus | 64 | 53 (11.6%) | 11 (13.4%) | 0.60 |
| At 8 weeks | ||||
| Major/minor bleed | 15 | 10 (2.2%) | 5 (6.1%) | 0.05 |
| Stroke/transient ischemic attack | 5 | 5 (1.1%) | 0 | 0.30 |
| Death | 13 | 9 (2%) | 4 (4.9%) | 0.10 |
| At 6 months | (n = 380) | (n = 63) | ||
| Normal sinus rhythm | 270 | 233 (61%) | 37 (59%) | 0.70 |
| (n = 404) | (n = 70) | |||
| Stroke/transient ischemic attack | 1 | 1 (0.3%) | 0 | 0.70 |
| Death | 5 | 2 (0.5%) | 3 (4.3%) | 0.004 |
⁎ Because of low cell frequencies, p values were obtained using Fisher’s exact test.
Of the 14 patients with CHADS 2 scores of 0 and LAA thrombi, none developed a stroke or embolic event within the 8 weeks or 6 months of follow up. No patient with a CHADS 2 score of 0 had a stroke or transient ischemic attack. In addition, maintenance of sinus rhythm in patients with CHADS 2 scores of 0 was more likely to be observed in patients without LAA thrombus on TEE (p = 0.05; Table 4 ). Few echocardiographic parameters had a significant trend across all CHADS 2 scores; most notably, the left ventricular ejection fraction decreased with increasing CHADS 2 score. Finally, in a comparison between patients with CHADS 2 scores of 0 and LAA thrombi versus none, we observed significant differences in the frequency of few echocardiographic findings, such as spontaneous echocardiographic contrast (higher in patients with thrombi; p = 0.002) and tricuspid valve regurgitation (lower in patients with thrombi; p = 0.007) ( Table 4 ).
