Discussion
This is a unique study that compared ED patients with AF for more than 48 hours’ duration who subsequently underwent TEE and then electrical cardioversion by emergency physicians back in the ED versus in the cardiology department. Successful cardioversion, defined at discharge, showed that over 90% of cardioversions were successful in both groups. Time to TEE was approximately one day for both groups. However, time to cardioversion, time kept in the hospital post-procedure, and LOS in the hospital were significantly shorter in the ED group versus the cardiology department group. This was thought to be due to several factors. All patients who were cardioverted underwent procedural sedation. Those who had the procedure performed in the ED had the sedation performed by a second ED physician which could be arranged relatively quickly. However, at the time of the study, the patients who underwent cardioversion in the cardiology department needed to wait for an anesthesiologist to perform the sedation. As there was no dedicated anesthesiologist, this could result in a significant waiting period. Also, there was a logistical need to arrange for a monitored observation area in the cardiology department. Some of the patients in the cardiology department underwent additional workups and treatment for a secondary issue such as anemia. Also, there may have been a trend to quickly discharge post-procedure patients from a busy urban ED. Interestingly there was no significant difference in the readmission rate to the hospital one-month post-hospital discharge and no known significant negative outcomes in the ED cardioversion group.
This study is the first to be done to examine ED cardioversion post-TEE performance in patients whose atrial fibrillation was more than 48 hours’ duration. Previous studies have described ED electrical cardioversions of less than 48 hours’ duration with success rates ranging from 85.5 to 97%.9 Burton et al. described a 7-day 10% return of patients, of whom 78% (25/32) returned due to a recurrence of AF.1 A recent multicenter randomized control trial assigned ED patients with recent-onset AF (<36 hours) who were symptomatic yet hemodynamically stable to be treated either with early cardioversion or with a “wait-and-see” approach. At 4-week follow-up, AF recurred in 49/164 (30%) in the delayed cardioversion group versus in 50/171 (29%) of the earlier cardioversion group.10 Scheuermeyer et al. performed a random 30-day follow-up chart review of 400/1233 AF patients who had undergone ED electrical cardioversion. Only 22 patients (5.5%) returned to the ED within one month for symptoms that were deemed potentially related to the index visit, with six patients (1.5%) requiring repeat cardioversion.8 The current study described an overall 94% success rate of rhythm control at discharge (141/151), with an overall 30-day return rate of 22% (32/151), of whom only 44% (14/32) returned due to recurrent AF, constituting 9.3% of the total study population. It is interesting to note that in the current study, the ED patients fared as well as those in the cardiology cohort. Clinical backgrounds and reasons for ED return also did not differ between the groups. Further investigation into the short-term and longer-term success of ED electrical cardioversion of AF of >48 hours’ duration after TEE performance is necessary. It seems, based on the results of the current study, that this approach may be at least as successful as some reports of ED cardioversion of AF of <48 hours’ duration.
It has been noted that patients with AF >48 hours’ duration who were treated with standard practice, i.e. weeks of anticoagulation before cardioversion, were not always adequately anticoagulated.11,12 TEE studies performed on patients with AF <48 hours have shown left atrial thrombi in 1.4% of patients, of whom 63% were receiving prior anticoagulation therapy, whereas patients not on prior anticoagulation medication had a 4% prevalence of left atrial thrombi.13 The data from these studies seem to point to the conclusion that timing of onset of AF should not be the only factor in determining a therapeutic approach to AF, and that performance of a TEE to rule out atrial thrombi may be warranted despite prior anticoagulation for even a short duration of AF. In a recent analysis of anticoagulation management with warfarin involving more than 120,000 patients in the Veterans Affairs health care system, the mean proportion of time in the therapeutic range was 58%, with significant variation across sites.14 With the new oral anticoagulants, this becomes less of an issue.15
The most prominent results of the current study were the differences in length of stay of patients based on where their cardioversion was performed. Time to TEE performance was not significantly different, but once it was performed, time to cardioversion, time to discharge, and inevitably length of hospital stay were all significantly shorter (p<0.001 for the latter three variables) for patients who were cardioverted in the ED versus those cardioverted in the cardiology ward.
Differences in in-patient versus ED lengths of stay in AF patients have previously been shown. Ptaszek et al. demonstrated a marked difference in in-patient LOS after the creation of an AF pathway (64 hours post-intervention versus 104 hours pre-intervention) with no significant corresponding reduction in ED stay.16 Other ED studies have cited shorter lengths of stay and higher patient satisfaction with ED AF cardioversion.7,9 As the frequency of AF related hospital visits is predicted to rise,2 it is necessary to create strategies to reduce the length of stay, thereby reducing hospital costs. The current study suggests that the performance of TEE in patients with AF of > 48 hours’ duration to rule out thrombi, and subsequent performance of electrical cardioversion in the ED reduces the length of hospital stay and is at least as safe as when performed on the cardiology ward.
Some EDs with fast turn-around time may choose to put such patients into an observation ward. While there is much literature on the use of an ED observation ward to manage patients with atrial fibrillation, none specifically look at the subcategory of those arriving more than 48 hours after onset.17–21 For those EDs that do have an observation unit, this treatment strategy may be appropriate.
On the other hand, many EDs have problems with crowding resulting in the boarding of patients.22-24 This strategy may also be advantageous in this setting, potentially resulting in earlier discharge from the hospital than having the patient wait to be treated on the cardiology ward.