- Email: [email protected]
An emergency department observation unit protocol for acute-onset atrial fibrillation is feasible
ORIGINAL CONTRIBUTION
An Emergency Department Observation Unit Protocol for Acute-Onset Atrial Fibrillation Is Feasible
From the Department of Emergency Medicine, William Beaumont Hospital, Royal Oak, MI*; and the Department of Emergency Medicine, Wayne State University School of Medicine, Detroit, MI.‡ Author contributions are provided at the end of this article. Received for publication February 12, 2001. Revision received November 19, 2001. Accepted for publication December 17, 2001. Presented at the Society of Academic Emergency Medicine annual meeting, San Francisco, CA, May 2000. Address for reprints: Michael A. Ross, MD, 3601 West Thirteen Mile Road, Royal Oak, MI 48073-6769; 248-551-3080 or 248-551-2016; E-mail [email protected]. Copyright © 2002 by the American College of Emergency Physicians. 0196-0644/2002/$35.00 + 0 47/1/122785 doi:10.1067/mem.2002.122785
Benjamin O. Koenig, MD* Michael A. Ross, MD*‡ Raymond E. Jackson, MD, MS*‡
Study objective: We sought to describe the feasibility of an emergency department observation unit (EDOU) treatment protocol for the management of uncomplicated acute-onset atrial fibrillation (AAF). Methods: This descriptive case series took place at a major suburban, university-affiliated teaching hospital. Patients were prospectively enrolled in an EDOU treatment protocol if they had uncomplicated AAF that failed initial ED attempts to convert to sinus rhythm. In the EDOU, patients underwent ECG monitoring, serial creatine kianse MB measurements, and further rate control with optional electrical cardioversion. Primary outcomes measured were EDOU rate of conversion to sinus rhythm, rate of discharge home, length of stay, positive diagnostic outcomes, complications of AAF, and 7-day return visits. Results: Sixty-seven patients were studied. Patients were symptomatic for a median of 4.0 hours, had mean initial ED pulse rates of 137±23 beats/min, and spent 4.7±2.2 hours in the ED before transfer to the EDOU. While in the EDOU, 55 (82%) patients converted to sinus rhythm. Five (7%) patients were admitted because of positive test results: 2 for myocardial infarction, 2 for fever, and 1 for ventricular tachycardia. Twelve (18%) patients remained in atrial fibrillation, with 9 admitted and 3 discharged. Overall, 81% of patients were discharged in 11.8±7.0 hours, and 19% were admitted after 17.6±9.5 hours of observation. Three discharged patients returned within 7 days, 2 for uncomplicated recurrent AAF and 1 for chest pain subsequently found to be noncardiac in origin. There were no major complications attributable to the EDOU protocol. Conclusion: Selected patients with AAF for whom initial ED management fails can subsequently be managed in an EDOU with a high short-term conversion and discharge rate. [Koenig BO, Ross MA, Jackson RE. An emergency department observation unit protocol for acute-onset atrial fibrillation is feasible. Ann Emerg Med. April 2002;39:374-381.]
3 7 4
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
INTRODUCTION
M AT E R I A L S A N D M E T H O D S
Atrial fibrillation is the most common cardiac arrhythmia managed in the emergency department.1 It results from the chaotic depolarization of atrial tissue.2 It is associated with a reduction in cardiac output3 and an increased risk of stroke. Atrial fibrillation accounts for 30,000 to 40,000 strokes per year in the United States.4 Acute-onset atrial fibrillation (AAF; or new-onset, paroxysmal atrial fibrillation) is defined as atrial fibrillation in which symptoms have lasted for less than 48 to 72 hours. Because atrial fibrillation leads to changes within the heart that perpetuate it (atrial fibrillation begets atrial fibrillation), cardioversion to sinus rhythm is more likely to be successful the sooner it is conducted.5 For this reason, patients presenting to the ED with AAF benefit from timely restoration of sinus rhythm. Traditionally, when initial conversion to sinus rhythm fails, these patients are admitted to the hospital to achieve rate control, to restore sinus rhythm, and to monitor for comorbidities or complications, such as myocardial infarction or embolic events. The incidence of serious cardiovascular complications within this group may be as high as 20%.6 It has been reported that hospital lengths of stay for atrial fibrillation are markedly greater than for any other arrhythmia.7,8 However, a subset of patients with AAF that may not require inpatient admission has been described. Recent studies suggest that approximately one third of admitted patients with AAF may not have required inpatient admission.9 Others have proposed that stable patients with recent-onset atrial fibrillation may avoid hospitalization by using a combination of initial rate control and various therapies to encourage conversion to sinus rhythm.10 Studies in Canada and Australia have suggested that many patients with AAF may be safely managed in the ED and discharged.11-13 In the United States, there has been growing interest in the ED observation unit (EDOU) as an alternative to inpatient admission. In this setting, patients with other conditions are aggressively managed with accelerated treatment or diagnostic protocols, resulting in high discharge rates, lower costs, and improved satisfaction.14-18 It would seem that the EDOU would be an ideal alternative to inpatient admission for the management of stable patients with AAF presenting to the ED. However, there are currently no studies of AAF managed in an EDOU setting. In this study, we report the outcomes of an EDOU treatment protocol for the management of patients with AAF in whom initial ED therapy has failed.
This is a prospective descriptive case series of patients enrolled in an EDOU treatment protocol for the management of AAF. This study was conducted in the ED at the William Beaumont Hospital, a high-volume, universityaffiliated, suburban teaching hospital. The study was granted an institutional review board exemption at the study hospital. The EDOU has been operational since 1995 and complies with the American College of Emergency Physicians’ guidelines for the management of an EDOU.19,20 Board-certified emergency physicians manage all patients in the EDOU 24 hours a day. The unit is staffed with a patient/nurse ratio of approximately 5:1. Staffing patterns in the unit have previously been reported.21 A disposition must be made on all patients in the EDOU within 18 hours. The unit does not take patients who require continuous infusions of vasoactive medications, such as diltiazem drips. Intravenous diltiazem may be used in the main ED to achieve initial rate control and then discontinued before transfer to the unit. Preceding the study, a consensus group of members from the departments of cardiology and emergency medicine developed the EDOU accelerated treatment protocol for AAF on the basis of a literature review (Figure 1). The intent of the EDOU protocol was to provide an alternative to prolonged inpatient admission of patients with uncomplicated AAF. Use of the EDOU for patients with uncomplicated AAF was not compulsory for the physician. The protocol included the following components: inclusion and exclusion criteria, treatment plan, and criteria for discharge or admission from the EDOU. The basic treatment steps included initial rate control and screening for EDOU eligibility in the ED, followed by continued rate control, clinical evaluation, and discretionary cardioversion in the EDOU. Because the group found there to be a variety of acceptable medications, with none clearly superior, the protocol directed the basic steps that would occur rather than the specific medications to be used for rate control, cardioversion, and anticoagulation. This approach to the management of AAF has been suggested but not tested.10 Patients were eligible for inclusion if the onset of symptoms for their atrial fibrillation was clearly less than 48 hours. Patients had to be hemodynamically stable, which was defined as having a systolic blood pressure of greater than 100 mm Hg. Initial pulse rate control was achieved in the ED before transfer to the observation unit by using standard medications. A ventricular response of less than 110 beats/min for at least 1 hour after initial ED
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 7 5
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
treatment was required. Patients were excluded if they continued to have chest discomfort suggestive of ischemia despite initial rate control. They were also excluded if they showed evidence of significant comorbidities, including ECG evidence of acute myocardial ischemia, congestive heart failure, pulmonary embolism, sepsis, stroke, or peripheral embolic complications. Once initial ED rate control and testing were completed, the emergency physician contacted the patients’ private physicians or cardiologists regarding the patients’ planned transfer to the EDOU. Consensus regarding specific medications and electrocardioversion was reached at this time. If the patients’ primary care physicians preferred inpatient admission, the patient was admitted instead of being sent to the EDOU. These patients were excluded from this study. The EDOU protocol included both diagnostic and therapeutic components. While in the EDOU, patients received continuous ECG and ST-segment monitoring; a
Figure 1.
Protocol for the EDOU treatment of AAF. BP, Blood pressure; PR, pulse rate; MI, myocardial infarction; CHF, congestive heart failure; PE, pulmonary embolism; CVA, cerebral vascular accident (stroke); NPO, nothing by mouth. EDOU inclusion criteria Systolic BP >100 mm Hg, PR under 110 beats/min for 1 h (with treatment) No chest pain with rate controlled No evidence of acute comorbidites: MI, CHF, PE, CVA, etc Onset known to be less than 48 h Cardiologist and emergency physician agree with plan to observe EDOU interventions Cardiac and ST-segment monitoring Vital signs every 2 h Anticoagulation: aspirin, discretionary heparin (SC or IV) Rate control as needed with calcium channel blockers, β-blockers, or digoxin Testing: CK-MB at 3, 6, and 9 h from arrival in ED Educate patient on cardioversion (medical or electrical) if initial EDOU treatment fails within 12 h NPO at 12 h for electrical cardioversion to occur outside EDOU EDOU disposition parameters Home from the EDOU Patient converts and remains in sinus rhythm for 1 h Negative CK-MB rule-out Stable condition Discuss home medication therapy with cardiologist Admit (inpatient) from the EDOU Failure to maintain rate control Positive CK-MB test results Unstable condition Time frame Less than 18 h
3 7 6
repeat ECG at 9 hours, at the time of conversion to sinus rhythm, or both; vital signs measurement every 2 hours; and creatine kinase (CK) MB testing at 3, 6, and 9 hours from arrival in the ED. Treatment in the EDOU included further medications to control ventricular response and discretionary use of heparin and chemical cardioversion agents. If cardioversion did not occur after 12 hours in the EDOU, the patients were given a nothing-by-mouth order, and electrical cardioversion was considered. Electrical cardioversion occurred outside the observation unit in an adjacent ED bed. Discharge from the EDOU was considered if the patient converted and remained in sinus rhythm for 1 hour, had negative serial CK-MB test results, and had normal vital signs. The cardiologist or primary care physician was contacted at the time of discharge to discuss home medication therapy and to arrange office follow-up. Inpatient admission from the EDOU was considered if the patient did not convert to sinus rhythm within 18 hours, had a persistent rapid ventricular response, was found to have significant comorbidities on serial testing, or was in unstable clinical condition. The study period began in January 1997 and included all EDOU AAF protocol patients enrolled over 34 consecutive months. The study is limited to patients presenting to the ED with AAF who were subsequently managed in the EDOU. All patients had initial data prospectively entered into an EDOU database (Access; Microsoft, Redmond, WA). This database included patient identification number, EDOU diagnosis, date and time of arrival in the ED and the EDOU, date and time of departure from the EDOU, and final disposition (admission or discharge). From this database, all patients managed for atrial fibrillation in the EDOU were identified. Further clinical information was obtained from patient records and hospital databases. This included additional patient demographic data, specific treatments provided in both the ED and the EDOU, clinical outcomes, and information on return visits. Available information was reviewed by one author (BOK) to verify that all patients had AAF. Patients were excluded if they were observed for complications of chronic atrial fibrillation, if their initial ECG showed sinus rhythm, or if their duration of symptoms was greater than 48 hours or could not be determined from the medical record. Patients with ECGs consistent with atrial flutter were included in the analysis. Key outcomes measured were the rates of conversion to sinus rhythm, rates of discharge home, lengths of stay in the ED and EDOU, and ED return visits within 7 days
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
of the index visit. Return visit data were collected for the primary study hospital and a second local community hospital. Data were also collected for positive clinical outcomes of the study protocol, including positive serial CKMB or ECG results, evidence of sepsis, thromboembolic complications, major hemodynamic events (including sustained ventricular tachycardia), ventricular fibrillation, persistent high-grade bradydysrhythmias, shock, and death. Standard descriptive methods were used to report the data, including means and SDs. A list of patients treated at our institution with a principle International Classification of Diseases, ninth edition (ICD-9) code of atrial fibrillation or flutter (ICD-9 codes 427.31 and 427.32, respectively) was obtained to estimate the total number of patients seen at our hospital for AAF during the study period. From this group, a detailed chart review was conducted on 100 consecutive patients to profile the percentage of patients seen for the treatment of chronic atrial fibrillation versus AAF and the percentage of these patients managed as inpatients, observation patients, or patients discharged directly from the ED. R E S U LT S
Over the study period, 259,175 patients were seen in the ED. Of these, 2,750 (1.06%) had the ICD-9 code for atrial fibrillation coded in their medical record. On chart review,
Table 1.
65% (1,788) were seen for treatment of chronic atrial fibrillation and 35% (125) for AAF, yielding an estimated 962 patients with AAF. Of patients with AAF, 13% converted in the ED and were discharged home, 80% (770) were admitted as an inpatient, and 7% (67 patients) were treated in the EDOU and are detailed in this study. During the study period, 102 patients were identified as having been treated for a primary diagnosis of atrial fibrillation in the EDOU. Sixteen of these patients were excluded from analysis because they were identified in the ED as having chronic atrial fibrillation, 14 were excluded because their initial ED ECG showed conversion to sinus rhythm, 3 were excluded because the duration of symptoms exceeded 48 hours, and 2 were excluded for an indeterminate duration of symptoms. The remaining 67 patients were included in this study. Patient characteristics, presenting complaints, medical histories, and clinical presentations are presented in Table 1. The average age was 58 years, with 58% (36) of the patients being men. By far the most frequent presenting symptom was palpitations (feeling of strong or rapid heartbeat), followed by chest pain, dizziness, and shortness of breath. The average duration of symptoms was 11.6 hours, with a median duration of 4 hours. The average pulse rate on arrival in the ED was 137±23 beats/min. Five (7%) patients presented with pulse rates of less than 100 beats/min, and 4 of these were taking β-blockers. Six (9%) patients showed atrial flutter on their initial ECG. The ED management before transfer to the EDOU is described in Table 2. Seventy-three percent (49) of the
EDOU AAF study patient characteristics. Characteristic
Patients (N=67)
Mean±SD age, y Age range, y Male sex Presenting symptoms (%) Palpitations Chest pain Dizziness Shortness of breath Weakness Mean duration of symptoms, h Median duration of symptoms, h Previous atrial fibrillation episode (%) Cardiac medications at home (%) Digoxin β-Blockers Calcium channel blockers Propafenone Mean±SD pulse rate at ED arrival, beats/min Patients in atrial flutter at ED arrival (%)
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
58±17 20–90 36 (54) 57 (85) 15 (22) 12 (18) 12 (18) 5 (7) 11.6 4 36 (54) 13 (19) 13 (19) 8 (12) 9 (13) 137±23 6 (9)
Table 2.
Initial treatment in the ED for patients with AAF. Characteristic Rate-control medications (%) Diltiazem Verapamil Metoprolol Digoxin Any of the above Chemical cardioverting medications (%) Propafenone Procainamide Flecainide Amiodarone Sotalol Any of the above Mean±SD ED length of stay, h Mean±SD pulse rate at EDOU transfer, beats/min
Patients (N=67)
16 (24) 12 (18) 12 (18) 26 (39) 49 (73) 41 (61) 5 (7) 1 (1) 1 (1) 1 (1) 49 (73) 4.7±2.2 92±18
3 7 7
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
patients were treated with medication in the ED to control their pulse rate. Most patients (61%) were also given an initial dose of propafenone (Rhytmol) in the ED; the dosage varied between 150 and 600 mg. Eight (12%) patients were given other antiarrhythmics. Overall, 73% (49) of patients received initial medication in the ED to encourage chemical cardioversion. The length of stay in the ED before transfer to the observation unit was 4.7±2.2 hours. At the time of transfer to the EDOU, the mean pulse rate was 92±18 beats/min. The clinical results of treatment in the EDOU are summarized in Figure 2. Fifty-one (76%) patients converted to sinus rhythm while in the EDOU before the use of electrical cardioversion. Of the 16 (24%) patients who did not convert, electrical cardioversion was attempted on 5 (7%) patients, with 4 successful conversions. Of the 12 (18%) patients remaining in atrial fibrillation, 9 were admitted
Figure 2.
Observation unit AAF clinical outcomes. AF, Atrial fibrillation; SR, sinus rhythm; EC, electrical cardioversion; LOS, length of stay in the EDOU.
and 3 were discharged on appropriate medications and scheduled for follow-up with a cardiologist. Four (6%) patients were admitted on the basis of the results of positive diagnostic outcomes, despite having converted to sinus rhythm (Table 3). Two of these patients had fevers in the observation unit, and 2 patients were ruled in for non–Q-wave myocardial infarction on serial CK-MB testing. A fifth patient had a positive diagnostic outcome but also did not convert from atrial fibrillation. This patient had a 10-beat run of ventricular tachycardia after treatment with both propafenone and electrical cardioversion. During inpatient admission, he had electrophysiologic testing, an automatic implantable cardioverter defibrillator was placed, and he did well. No patients experienced major complications (death, cardiac arrest, shock, thromboembolic events, or high-grade atrioventricular block) as a result of the treatment protocol
67 patients treated for AF
51 converted to SR without use of EC
16 remained in AF
5 had EC attempted
4 successful conversions
1 failed conversion (Table 3, patient 5)
55 (82%) converted to SR
51 discharged in SR
11 EC not attempted
12 (18%) remained in AF
4 admitted in SR (Table 3, patients 1-4)
3 discharged in AF
54 (81%) discharged to home: average EDOU LOS=11.8±7.0 h
9 admitted in AF
13 (13%) admitted to hospital: average EDOU LOS=17.6±9.5 h
3 patients returned (Table 3): • Two with recurrent AAF—cardioverted and discharged in 2-3 d • One with chest pain—found to be noncardiopulmonary chest pain
3 7 8
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
while in the EDOU. No patients experienced a delay in therapy for the aforementioned positive clinical outcomes as a result of care in the unit. Three (6%) of the 54 discharged patients returned to the primary hospital ED within 7 days: 1 because of chest pain and 2 because of symptomatic recurrent atrial fibrillation. The patient with chest pain subsequently had negative serial cardiac marker results, followed by a negative sestimibi exercise stress test result and no clinical evidence of other acute pathology. The 2 patients with recurrent AAF were chemically cardioverted and discharged in 2 to 3 days without complications. None of these 3 patients had been treated with cardioverting medications during their EDOU visit. No patients returned to the proximate community hospital within 7 days of discharge. Overall, 54 (81%) patients were discharged in 11.8±7.0 hours, and 13 (19%) patients were admitted after 17.6±9.5 hours in the EDOU. This was in addition to the 4.7 hours of treatment in the ED before transfer to the EDOU. DISCUSSION
We found that, for a selected subset of ED patients with uncomplicated AAF, management in an EDOU was a feasible alternative to inpatient admission. We describe a basic protocol for the management of these patients that contains both therapeutic and diagnostic components. By using this protocol, high rates of conversion to sinus rhythm and discharge can be expected, and latent occult pathology may be identified. No previous studies have addressed the treatment of AAF in an EDOU. A previous
retrospective review of inpatient charts suggested that a subset of patients admitted for this diagnosis may not require inpatient admission.9 Others described a subset of patients who can be managed exclusively in an ED with high conversion and discharge rates.11-13 In a review of atrial fibrillation, Falk10 suggested a protocol similar to the one we used, but it was not applied to an EDOU setting. Because patients remaining in AAF are generally not sent home and conversion to sinus rhythm may require more time than a busy ED can offer, the EDOU may be an ideal alternative setting for uncomplicated cases. There were no EDOU patients with AAF who had major complications as a result of the treatment protocol. However, the protocol did allow the physician to identify 5 (7%) patients with significant latent or occult complications of AAF that were not initially apparent in the ED. Two (3%) patients were ruled in for myocardial infarction on serial CK-MB testing despite cardioversion, and 2 (3%) patients had fevers. This raises the question of the need for short-term monitoring of all patients with AAF, even if cardioversion occurs at or just after arrival in the ED. One patient had nonsustained ventricular tachycardia and was subsequently found to require an internal defibrillator. On review, this was found to be a result of the patient’s underlying disease process and not the EDOU protocol used. However, this emphasizes the importance of having necessary ACLS equipment and resuscitative skills in the EDOU for such patients. We found that diagnostic evaluation with cardiac monitoring, serial vital signs measurement, and cardiac markers measurement to be most helpful.
Table 3.
EDOU patients with AAF and positive diagnostic outcomes and 7-day returns. Patient No.
Age (y)
Sex
Complication
1
61
Male
Non–Q-wave MI
2
54
Female
Non–Q-wave MI
3 4 5 6 7 8
90 77 74 52 74 67
Female Male Male Male Female Female
Fever in EDOU Fever in EDOU Ten beats of ventricular tachycardia Returned for recurrent atrial fibrillation Returned for recurrent atrial fibrillation Returned for chest pain
Outcome Cardiac catheterization showed mild-to-moderate coronary narrowing and was treated medically, patient was discharged in 3 d Admitted but signed out against medical advice 24 h later without workup, alive at 6-mo follow-up UTI found and treated, discharged in 12 d No cause found, discharged in 4 d Admitted, electrocardiologic testing, AICD placed, home in 6 d Chemically cardioverted and discharged in 2 d Chemically cardioverted and discharged in 3 d Negative serial CK-MB and graded exercise stress test results with myocardial perfusion imaging, discharged in 1 d for noncardiac chest pain
MI, Myocardial infarction; UTI, urinary tract infection; AICD, automatic implantable cardioverter defibrillator.
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 7 9
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
We did not identify any thromboembolic events in our treatment group. The incidence of these has been reported to be as high as 5% to 7% in the cardioversion of atrial fibrillation without anticoagulation.22 However, Weigner and Caulfield23 have shown a lower risk of thromboembolic events (0.8%) if symptoms have been present for less than 48 hours. Consequently, elective cardioversion without anticoagulation is only recommended for patients with true AAF. If symptoms have been present for longer than 48 to 72 hours or if the duration of symptoms is unclear, it is suggested that patients receive anticoagulation for 1 to 3 weeks before attempting cardioversion.24 Although our protocol called for admission of patients with persistent atrial fibrillation, 3 (4%) patients were discharged in atrial fibrillation. These patients were discharged on oral anticoagulants, with planned short-term outpatient follow-up and delayed cardioversion. These patients were seen by a cardiologist while in the EDOU, and follow-up arrangements were made. Thus, these exceptions were believed to be safe. On follow-up, no delayed complications were identified in these 3 patients. Traditional practice has been to admit patients with AAF in whom initial ED attempts to convert to sinus rhythm failed or to discharge patients once they converted to sinus rhythm and were without complications. In our EDOU study, 81% of patients with AAF were discharged after an average of 11.8 hours of treatment in the EDOU. Others have found that, compared with the accelerated care available in an EDOU setting, inpatient lengths of stay and costs are greater for chest pain and asthma.14,16,17 Because the focus of this preliminary study was the feasibility of this approach, we did not attempt a randomized comparison of inpatient and EDOU management of AAF. However, on the basis of our results, we believe that such a study may be warranted. Restricting the drugs used in our protocol may have allowed us to study which drugs worked best in our protocol. However, our purpose was to study the feasibility of this protocol and not the specific medications. We did not believe that the literature was compelling enough to mandate the use of any one drug or combination of drugs to be used in our protocol. For this reason, we standardized the basic treatment steps rather than the specific drugs used. A high percentage of patients treated only with medication to control ventricular response spontaneously converted to sinus rhythm. This raises the question of which is safer, rate control and a limited period of observation or the use of an antiarrhythmic, such as ibutilide, which has an 8.3% incidence of polymorphic ven-
3 8 0
tricular tachycardia.25 Propafenone was preferentially used for cardioversion at our institution because it was believed to have a superior safety profile relative to other cardioverting drugs of similar efficacy. We speculate that our conversion rates may have been slightly improved if all patients had been treated with chemical or electrical cardioversion. However, this study was not designed to answer that question. Again, future studies of AAF in an EDOU setting might explore the nature of the optimal treatment strategy. This case series has several inherent limitations. Because we used a case series of patients treated in an EDOU, we did not detail the clinical characteristics of the entire population of patients presenting to the ED with AAF. We believed that this would be beyond the focus of this case series. However, we did provide a general profile of this population in our review of ICD-9 data. Not all eligible patients were enrolled in the EDOU protocol because physician preference was allowed regarding admission to the EDOU or an inpatient bed. This may have introduced a referral bias, perhaps selecting a healthier subset of patients. However, we are aware of no setting in which the use of an EDOU is compulsory for all patients who might otherwise be admitted as an inpatient. Making the protocol compulsory for all patients with uncomplicated persistent AAF might similarly introduce a selection bias that would not reflect actual practice. Our intent was to assess the feasibility of a protocol that allowed discretionary enrollment, similar to what occurs in clinical practice. On the basis of our findings, we believe that a larger, prospective, randomized controlled trial comparing this EDOU protocol with inpatient treatment of all eligible patients with uncomplicated AAF would help to better define other outcomes, such as cost and patient satisfaction. In summary, we found that it is feasible for patients with uncomplicated AAF to be managed with a simple EDOU protocol, with a majority being successfully cardioverted and discharged in less than 12 hours. A small but significant number of patients may be found to have positive diagnostic findings. Further study of the management of uncomplicated atrial fibrillation in an EDOU setting regarding its effect on cost and patient satisfaction is warranted. ADDENDUM
The recent decision by the Centers for Medicare and Medicaid Services to separately fund observation services for only the 3 most well-studied conditions (chest pain, asthma, and congestive heart failure) underscores the
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
need to further study other conditions that may be amenable to EDOU treatment. A copy of the Federal Register comments made by the Centers for Medicare and Medicaid Services can be accessed at http://www.hcfa.gov/regs/Pp44671.pdf.26 Author contributions: MAR and BOK conceived the study, and REJ worked with them to refine its design. BOK conducted data collection. MAR, REJ, and BOK all analyzed and interpreted the data. MAR and BOK drafted the manuscript, while MAR, BOK, and REJ all contributed to revisions. MAR and BOK take responsibility for the paper as a whole.
22. Mancini GBJ, Foldberger AL. Cardioversion of atrial fibrillation: consideration of embolization, anticoagulation, prophylactic pacemaker and long term success. Am Heart J. 1982;104:617-621. 23. Weigner MJ, Caulfield TA. Risk for clinical thromboembolism associated with conversion to sinus rhythm in patients with atrial fibrillation lasting less than 48 hours. Ann Intern Med. 1997;126:615-620. 24. Laupacis A, Albers G, Dolen J, et al. Antithrombotic therapy in atrial fibrillation. Chest. 1995;108:352S-359S. 25. Stambler BS, Wood MA, Ellenbogen KA, et al. Efficacy and safety of repeated intravenous doses of ibutilide for rapid conversion of atrial flutter or fibrillation. Circulation. 1996;94:16131621. 26. Federal Register. Vol. 66, No. 165. Proposed rules, p. 44,690-44,692. Available at: http://www.hcfa.gov/regs/Pp44671.pdf. Accessed on August 24, 2001.
REFERENCES 1. Huagui L, Easley A, Barrington W, et al. Evaluation and management of atrial fibrillation in the emergency department. Emerg Med Clin North Am. 1998;16:289-303. 2.
Narayan SM, Cain ME, Smith JM. Atrial fibrillation. Lancet. 1997;350:943-950.
3. Upshaw C. Hemodynamic changes after cardioversion of chronic atrial fibrillation. Arch Intern Med. 1997;157:1070-1076. 4.
Brickner ME. Cardioembolic stroke. Am J Med. 1996;100:465-474.
5. Wijffells MC, Kirchof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation: a study in awake chronically instrumented goats. Circulation. 1995;92:1954-1968. 6. Friedman HZ, Scot FG, Bonema JD, et al. Acute complications associated with new-onset atrial fibrillation. Am J Cardiol. 1991;67:437-439. 7. Prystowsky EN, Benson DW, Fuster V, et al. Management of patients with atrial fibrillation; a statement for healthcare professionals from the Subcommittee on Electrocardiography and Electrophysiology, American Heart Association. Circulation. 1996;93:1262-1277. 8. Bialy D, Lehmann MH, Schumacher DN, et al. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol. 1992;19:41A. 9. Mulcahy B, Coates WC, Hennerman PL, et al. New-onset atrial fibrillation: When is admission medically justified? Acad Emerg Med. 1996;3:114-119. 10. Falk R. Medical progress: atrial fibrillation. N Engl J Med. 2001;344:1067-1078. 11. Michael JA, Stiell IG, Agarwal S, et al. Cardioversion of paroxysmal atrial fibrillation in the emergency department. Ann Emerg Med. 1999;33:379-387. 12. Joseph AP, Ward MR. A prospective, randomized controlled trial comparing the efficacy and safety of sotalol, amiodarone, and digoxin for the reversion of new-onset atrial fibrillation. Ann Emerg Med. 2000;36:1-9. 13. Innes GD, Vertesi L, Dillon EC, et al. Effectiveness of verapamil-quinidine versus digoxinquinidine in the emergency department treatment of paroxysmal atrial fibrillation. Ann Emerg Med. 1997;29:126-134. 14. Rydman RJ, Isola MI, Roberts RR, et al. Emergency department observation unit versus hospital inpatient care for a chronic asthmatic population: a randomized trial of health status outcome and cost. Med Care. 1998;36:599-609. 15. Rydman RJ, Roberts RR, Albrecht GL, et al. Patient satisfaction with an emergency department asthma observation unit. Acad Emerg Med. 1999;6:178-183. 16. Gomez MA, Jefferey LA, Karagounis LA, et al. An emergency department based protocol for rapidly ruling myocardial ischemia reduces hospital time and expense: results of a randomized study (ROMIO). J Am Coll Cardiol. 1996;28:25-33. 17. Farkouh ME, Smars PA, Reeder GS, et al. A clinical trial of a chest-pain observation unit for patients with unstable angina. N Engl J Med. 1998;339:1882-1888. 18. Rydman RJ, Zalenski RJ, Roberts RR, et al. Patient satisfaction with an emergency department chest pain observation unit. Ann Emerg Med. 1997;29:109-116. 19. Brillman J, Mathers-Dunbar L, Graff L, et al. Management of observation units. Ann Emerg Med. 1995;25:823-830. 20. ACEP policy statement. Emergency department observation units. Ann Emerg Med. 1995;25:863-864. 21. Ross MA, Naylor S, Compton S, et al. Maximizing use of the emergency department observation unit: a novel hybrid design. Ann Emerg Med. 2001;37:267-274.
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 8 1
An Emergency Department Observation Unit Protocol for Acute-Onset Atrial Fibrillation Is Feasible
From the Department of Emergency Medicine, William Beaumont Hospital, Royal Oak, MI*; and the Department of Emergency Medicine, Wayne State University School of Medicine, Detroit, MI.‡ Author contributions are provided at the end of this article. Received for publication February 12, 2001. Revision received November 19, 2001. Accepted for publication December 17, 2001. Presented at the Society of Academic Emergency Medicine annual meeting, San Francisco, CA, May 2000. Address for reprints: Michael A. Ross, MD, 3601 West Thirteen Mile Road, Royal Oak, MI 48073-6769; 248-551-3080 or 248-551-2016; E-mail [email protected]. Copyright © 2002 by the American College of Emergency Physicians. 0196-0644/2002/$35.00 + 0 47/1/122785 doi:10.1067/mem.2002.122785
Benjamin O. Koenig, MD* Michael A. Ross, MD*‡ Raymond E. Jackson, MD, MS*‡
Study objective: We sought to describe the feasibility of an emergency department observation unit (EDOU) treatment protocol for the management of uncomplicated acute-onset atrial fibrillation (AAF). Methods: This descriptive case series took place at a major suburban, university-affiliated teaching hospital. Patients were prospectively enrolled in an EDOU treatment protocol if they had uncomplicated AAF that failed initial ED attempts to convert to sinus rhythm. In the EDOU, patients underwent ECG monitoring, serial creatine kianse MB measurements, and further rate control with optional electrical cardioversion. Primary outcomes measured were EDOU rate of conversion to sinus rhythm, rate of discharge home, length of stay, positive diagnostic outcomes, complications of AAF, and 7-day return visits. Results: Sixty-seven patients were studied. Patients were symptomatic for a median of 4.0 hours, had mean initial ED pulse rates of 137±23 beats/min, and spent 4.7±2.2 hours in the ED before transfer to the EDOU. While in the EDOU, 55 (82%) patients converted to sinus rhythm. Five (7%) patients were admitted because of positive test results: 2 for myocardial infarction, 2 for fever, and 1 for ventricular tachycardia. Twelve (18%) patients remained in atrial fibrillation, with 9 admitted and 3 discharged. Overall, 81% of patients were discharged in 11.8±7.0 hours, and 19% were admitted after 17.6±9.5 hours of observation. Three discharged patients returned within 7 days, 2 for uncomplicated recurrent AAF and 1 for chest pain subsequently found to be noncardiac in origin. There were no major complications attributable to the EDOU protocol. Conclusion: Selected patients with AAF for whom initial ED management fails can subsequently be managed in an EDOU with a high short-term conversion and discharge rate. [Koenig BO, Ross MA, Jackson RE. An emergency department observation unit protocol for acute-onset atrial fibrillation is feasible. Ann Emerg Med. April 2002;39:374-381.]
3 7 4
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
INTRODUCTION
M AT E R I A L S A N D M E T H O D S
Atrial fibrillation is the most common cardiac arrhythmia managed in the emergency department.1 It results from the chaotic depolarization of atrial tissue.2 It is associated with a reduction in cardiac output3 and an increased risk of stroke. Atrial fibrillation accounts for 30,000 to 40,000 strokes per year in the United States.4 Acute-onset atrial fibrillation (AAF; or new-onset, paroxysmal atrial fibrillation) is defined as atrial fibrillation in which symptoms have lasted for less than 48 to 72 hours. Because atrial fibrillation leads to changes within the heart that perpetuate it (atrial fibrillation begets atrial fibrillation), cardioversion to sinus rhythm is more likely to be successful the sooner it is conducted.5 For this reason, patients presenting to the ED with AAF benefit from timely restoration of sinus rhythm. Traditionally, when initial conversion to sinus rhythm fails, these patients are admitted to the hospital to achieve rate control, to restore sinus rhythm, and to monitor for comorbidities or complications, such as myocardial infarction or embolic events. The incidence of serious cardiovascular complications within this group may be as high as 20%.6 It has been reported that hospital lengths of stay for atrial fibrillation are markedly greater than for any other arrhythmia.7,8 However, a subset of patients with AAF that may not require inpatient admission has been described. Recent studies suggest that approximately one third of admitted patients with AAF may not have required inpatient admission.9 Others have proposed that stable patients with recent-onset atrial fibrillation may avoid hospitalization by using a combination of initial rate control and various therapies to encourage conversion to sinus rhythm.10 Studies in Canada and Australia have suggested that many patients with AAF may be safely managed in the ED and discharged.11-13 In the United States, there has been growing interest in the ED observation unit (EDOU) as an alternative to inpatient admission. In this setting, patients with other conditions are aggressively managed with accelerated treatment or diagnostic protocols, resulting in high discharge rates, lower costs, and improved satisfaction.14-18 It would seem that the EDOU would be an ideal alternative to inpatient admission for the management of stable patients with AAF presenting to the ED. However, there are currently no studies of AAF managed in an EDOU setting. In this study, we report the outcomes of an EDOU treatment protocol for the management of patients with AAF in whom initial ED therapy has failed.
This is a prospective descriptive case series of patients enrolled in an EDOU treatment protocol for the management of AAF. This study was conducted in the ED at the William Beaumont Hospital, a high-volume, universityaffiliated, suburban teaching hospital. The study was granted an institutional review board exemption at the study hospital. The EDOU has been operational since 1995 and complies with the American College of Emergency Physicians’ guidelines for the management of an EDOU.19,20 Board-certified emergency physicians manage all patients in the EDOU 24 hours a day. The unit is staffed with a patient/nurse ratio of approximately 5:1. Staffing patterns in the unit have previously been reported.21 A disposition must be made on all patients in the EDOU within 18 hours. The unit does not take patients who require continuous infusions of vasoactive medications, such as diltiazem drips. Intravenous diltiazem may be used in the main ED to achieve initial rate control and then discontinued before transfer to the unit. Preceding the study, a consensus group of members from the departments of cardiology and emergency medicine developed the EDOU accelerated treatment protocol for AAF on the basis of a literature review (Figure 1). The intent of the EDOU protocol was to provide an alternative to prolonged inpatient admission of patients with uncomplicated AAF. Use of the EDOU for patients with uncomplicated AAF was not compulsory for the physician. The protocol included the following components: inclusion and exclusion criteria, treatment plan, and criteria for discharge or admission from the EDOU. The basic treatment steps included initial rate control and screening for EDOU eligibility in the ED, followed by continued rate control, clinical evaluation, and discretionary cardioversion in the EDOU. Because the group found there to be a variety of acceptable medications, with none clearly superior, the protocol directed the basic steps that would occur rather than the specific medications to be used for rate control, cardioversion, and anticoagulation. This approach to the management of AAF has been suggested but not tested.10 Patients were eligible for inclusion if the onset of symptoms for their atrial fibrillation was clearly less than 48 hours. Patients had to be hemodynamically stable, which was defined as having a systolic blood pressure of greater than 100 mm Hg. Initial pulse rate control was achieved in the ED before transfer to the observation unit by using standard medications. A ventricular response of less than 110 beats/min for at least 1 hour after initial ED
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 7 5
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
treatment was required. Patients were excluded if they continued to have chest discomfort suggestive of ischemia despite initial rate control. They were also excluded if they showed evidence of significant comorbidities, including ECG evidence of acute myocardial ischemia, congestive heart failure, pulmonary embolism, sepsis, stroke, or peripheral embolic complications. Once initial ED rate control and testing were completed, the emergency physician contacted the patients’ private physicians or cardiologists regarding the patients’ planned transfer to the EDOU. Consensus regarding specific medications and electrocardioversion was reached at this time. If the patients’ primary care physicians preferred inpatient admission, the patient was admitted instead of being sent to the EDOU. These patients were excluded from this study. The EDOU protocol included both diagnostic and therapeutic components. While in the EDOU, patients received continuous ECG and ST-segment monitoring; a
Figure 1.
Protocol for the EDOU treatment of AAF. BP, Blood pressure; PR, pulse rate; MI, myocardial infarction; CHF, congestive heart failure; PE, pulmonary embolism; CVA, cerebral vascular accident (stroke); NPO, nothing by mouth. EDOU inclusion criteria Systolic BP >100 mm Hg, PR under 110 beats/min for 1 h (with treatment) No chest pain with rate controlled No evidence of acute comorbidites: MI, CHF, PE, CVA, etc Onset known to be less than 48 h Cardiologist and emergency physician agree with plan to observe EDOU interventions Cardiac and ST-segment monitoring Vital signs every 2 h Anticoagulation: aspirin, discretionary heparin (SC or IV) Rate control as needed with calcium channel blockers, β-blockers, or digoxin Testing: CK-MB at 3, 6, and 9 h from arrival in ED Educate patient on cardioversion (medical or electrical) if initial EDOU treatment fails within 12 h NPO at 12 h for electrical cardioversion to occur outside EDOU EDOU disposition parameters Home from the EDOU Patient converts and remains in sinus rhythm for 1 h Negative CK-MB rule-out Stable condition Discuss home medication therapy with cardiologist Admit (inpatient) from the EDOU Failure to maintain rate control Positive CK-MB test results Unstable condition Time frame Less than 18 h
3 7 6
repeat ECG at 9 hours, at the time of conversion to sinus rhythm, or both; vital signs measurement every 2 hours; and creatine kinase (CK) MB testing at 3, 6, and 9 hours from arrival in the ED. Treatment in the EDOU included further medications to control ventricular response and discretionary use of heparin and chemical cardioversion agents. If cardioversion did not occur after 12 hours in the EDOU, the patients were given a nothing-by-mouth order, and electrical cardioversion was considered. Electrical cardioversion occurred outside the observation unit in an adjacent ED bed. Discharge from the EDOU was considered if the patient converted and remained in sinus rhythm for 1 hour, had negative serial CK-MB test results, and had normal vital signs. The cardiologist or primary care physician was contacted at the time of discharge to discuss home medication therapy and to arrange office follow-up. Inpatient admission from the EDOU was considered if the patient did not convert to sinus rhythm within 18 hours, had a persistent rapid ventricular response, was found to have significant comorbidities on serial testing, or was in unstable clinical condition. The study period began in January 1997 and included all EDOU AAF protocol patients enrolled over 34 consecutive months. The study is limited to patients presenting to the ED with AAF who were subsequently managed in the EDOU. All patients had initial data prospectively entered into an EDOU database (Access; Microsoft, Redmond, WA). This database included patient identification number, EDOU diagnosis, date and time of arrival in the ED and the EDOU, date and time of departure from the EDOU, and final disposition (admission or discharge). From this database, all patients managed for atrial fibrillation in the EDOU were identified. Further clinical information was obtained from patient records and hospital databases. This included additional patient demographic data, specific treatments provided in both the ED and the EDOU, clinical outcomes, and information on return visits. Available information was reviewed by one author (BOK) to verify that all patients had AAF. Patients were excluded if they were observed for complications of chronic atrial fibrillation, if their initial ECG showed sinus rhythm, or if their duration of symptoms was greater than 48 hours or could not be determined from the medical record. Patients with ECGs consistent with atrial flutter were included in the analysis. Key outcomes measured were the rates of conversion to sinus rhythm, rates of discharge home, lengths of stay in the ED and EDOU, and ED return visits within 7 days
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
of the index visit. Return visit data were collected for the primary study hospital and a second local community hospital. Data were also collected for positive clinical outcomes of the study protocol, including positive serial CKMB or ECG results, evidence of sepsis, thromboembolic complications, major hemodynamic events (including sustained ventricular tachycardia), ventricular fibrillation, persistent high-grade bradydysrhythmias, shock, and death. Standard descriptive methods were used to report the data, including means and SDs. A list of patients treated at our institution with a principle International Classification of Diseases, ninth edition (ICD-9) code of atrial fibrillation or flutter (ICD-9 codes 427.31 and 427.32, respectively) was obtained to estimate the total number of patients seen at our hospital for AAF during the study period. From this group, a detailed chart review was conducted on 100 consecutive patients to profile the percentage of patients seen for the treatment of chronic atrial fibrillation versus AAF and the percentage of these patients managed as inpatients, observation patients, or patients discharged directly from the ED. R E S U LT S
Over the study period, 259,175 patients were seen in the ED. Of these, 2,750 (1.06%) had the ICD-9 code for atrial fibrillation coded in their medical record. On chart review,
Table 1.
65% (1,788) were seen for treatment of chronic atrial fibrillation and 35% (125) for AAF, yielding an estimated 962 patients with AAF. Of patients with AAF, 13% converted in the ED and were discharged home, 80% (770) were admitted as an inpatient, and 7% (67 patients) were treated in the EDOU and are detailed in this study. During the study period, 102 patients were identified as having been treated for a primary diagnosis of atrial fibrillation in the EDOU. Sixteen of these patients were excluded from analysis because they were identified in the ED as having chronic atrial fibrillation, 14 were excluded because their initial ED ECG showed conversion to sinus rhythm, 3 were excluded because the duration of symptoms exceeded 48 hours, and 2 were excluded for an indeterminate duration of symptoms. The remaining 67 patients were included in this study. Patient characteristics, presenting complaints, medical histories, and clinical presentations are presented in Table 1. The average age was 58 years, with 58% (36) of the patients being men. By far the most frequent presenting symptom was palpitations (feeling of strong or rapid heartbeat), followed by chest pain, dizziness, and shortness of breath. The average duration of symptoms was 11.6 hours, with a median duration of 4 hours. The average pulse rate on arrival in the ED was 137±23 beats/min. Five (7%) patients presented with pulse rates of less than 100 beats/min, and 4 of these were taking β-blockers. Six (9%) patients showed atrial flutter on their initial ECG. The ED management before transfer to the EDOU is described in Table 2. Seventy-three percent (49) of the
EDOU AAF study patient characteristics. Characteristic
Patients (N=67)
Mean±SD age, y Age range, y Male sex Presenting symptoms (%) Palpitations Chest pain Dizziness Shortness of breath Weakness Mean duration of symptoms, h Median duration of symptoms, h Previous atrial fibrillation episode (%) Cardiac medications at home (%) Digoxin β-Blockers Calcium channel blockers Propafenone Mean±SD pulse rate at ED arrival, beats/min Patients in atrial flutter at ED arrival (%)
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
58±17 20–90 36 (54) 57 (85) 15 (22) 12 (18) 12 (18) 5 (7) 11.6 4 36 (54) 13 (19) 13 (19) 8 (12) 9 (13) 137±23 6 (9)
Table 2.
Initial treatment in the ED for patients with AAF. Characteristic Rate-control medications (%) Diltiazem Verapamil Metoprolol Digoxin Any of the above Chemical cardioverting medications (%) Propafenone Procainamide Flecainide Amiodarone Sotalol Any of the above Mean±SD ED length of stay, h Mean±SD pulse rate at EDOU transfer, beats/min
Patients (N=67)
16 (24) 12 (18) 12 (18) 26 (39) 49 (73) 41 (61) 5 (7) 1 (1) 1 (1) 1 (1) 49 (73) 4.7±2.2 92±18
3 7 7
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
patients were treated with medication in the ED to control their pulse rate. Most patients (61%) were also given an initial dose of propafenone (Rhytmol) in the ED; the dosage varied between 150 and 600 mg. Eight (12%) patients were given other antiarrhythmics. Overall, 73% (49) of patients received initial medication in the ED to encourage chemical cardioversion. The length of stay in the ED before transfer to the observation unit was 4.7±2.2 hours. At the time of transfer to the EDOU, the mean pulse rate was 92±18 beats/min. The clinical results of treatment in the EDOU are summarized in Figure 2. Fifty-one (76%) patients converted to sinus rhythm while in the EDOU before the use of electrical cardioversion. Of the 16 (24%) patients who did not convert, electrical cardioversion was attempted on 5 (7%) patients, with 4 successful conversions. Of the 12 (18%) patients remaining in atrial fibrillation, 9 were admitted
Figure 2.
Observation unit AAF clinical outcomes. AF, Atrial fibrillation; SR, sinus rhythm; EC, electrical cardioversion; LOS, length of stay in the EDOU.
and 3 were discharged on appropriate medications and scheduled for follow-up with a cardiologist. Four (6%) patients were admitted on the basis of the results of positive diagnostic outcomes, despite having converted to sinus rhythm (Table 3). Two of these patients had fevers in the observation unit, and 2 patients were ruled in for non–Q-wave myocardial infarction on serial CK-MB testing. A fifth patient had a positive diagnostic outcome but also did not convert from atrial fibrillation. This patient had a 10-beat run of ventricular tachycardia after treatment with both propafenone and electrical cardioversion. During inpatient admission, he had electrophysiologic testing, an automatic implantable cardioverter defibrillator was placed, and he did well. No patients experienced major complications (death, cardiac arrest, shock, thromboembolic events, or high-grade atrioventricular block) as a result of the treatment protocol
67 patients treated for AF
51 converted to SR without use of EC
16 remained in AF
5 had EC attempted
4 successful conversions
1 failed conversion (Table 3, patient 5)
55 (82%) converted to SR
51 discharged in SR
11 EC not attempted
12 (18%) remained in AF
4 admitted in SR (Table 3, patients 1-4)
3 discharged in AF
54 (81%) discharged to home: average EDOU LOS=11.8±7.0 h
9 admitted in AF
13 (13%) admitted to hospital: average EDOU LOS=17.6±9.5 h
3 patients returned (Table 3): • Two with recurrent AAF—cardioverted and discharged in 2-3 d • One with chest pain—found to be noncardiopulmonary chest pain
3 7 8
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
while in the EDOU. No patients experienced a delay in therapy for the aforementioned positive clinical outcomes as a result of care in the unit. Three (6%) of the 54 discharged patients returned to the primary hospital ED within 7 days: 1 because of chest pain and 2 because of symptomatic recurrent atrial fibrillation. The patient with chest pain subsequently had negative serial cardiac marker results, followed by a negative sestimibi exercise stress test result and no clinical evidence of other acute pathology. The 2 patients with recurrent AAF were chemically cardioverted and discharged in 2 to 3 days without complications. None of these 3 patients had been treated with cardioverting medications during their EDOU visit. No patients returned to the proximate community hospital within 7 days of discharge. Overall, 54 (81%) patients were discharged in 11.8±7.0 hours, and 13 (19%) patients were admitted after 17.6±9.5 hours in the EDOU. This was in addition to the 4.7 hours of treatment in the ED before transfer to the EDOU. DISCUSSION
We found that, for a selected subset of ED patients with uncomplicated AAF, management in an EDOU was a feasible alternative to inpatient admission. We describe a basic protocol for the management of these patients that contains both therapeutic and diagnostic components. By using this protocol, high rates of conversion to sinus rhythm and discharge can be expected, and latent occult pathology may be identified. No previous studies have addressed the treatment of AAF in an EDOU. A previous
retrospective review of inpatient charts suggested that a subset of patients admitted for this diagnosis may not require inpatient admission.9 Others described a subset of patients who can be managed exclusively in an ED with high conversion and discharge rates.11-13 In a review of atrial fibrillation, Falk10 suggested a protocol similar to the one we used, but it was not applied to an EDOU setting. Because patients remaining in AAF are generally not sent home and conversion to sinus rhythm may require more time than a busy ED can offer, the EDOU may be an ideal alternative setting for uncomplicated cases. There were no EDOU patients with AAF who had major complications as a result of the treatment protocol. However, the protocol did allow the physician to identify 5 (7%) patients with significant latent or occult complications of AAF that were not initially apparent in the ED. Two (3%) patients were ruled in for myocardial infarction on serial CK-MB testing despite cardioversion, and 2 (3%) patients had fevers. This raises the question of the need for short-term monitoring of all patients with AAF, even if cardioversion occurs at or just after arrival in the ED. One patient had nonsustained ventricular tachycardia and was subsequently found to require an internal defibrillator. On review, this was found to be a result of the patient’s underlying disease process and not the EDOU protocol used. However, this emphasizes the importance of having necessary ACLS equipment and resuscitative skills in the EDOU for such patients. We found that diagnostic evaluation with cardiac monitoring, serial vital signs measurement, and cardiac markers measurement to be most helpful.
Table 3.
EDOU patients with AAF and positive diagnostic outcomes and 7-day returns. Patient No.
Age (y)
Sex
Complication
1
61
Male
Non–Q-wave MI
2
54
Female
Non–Q-wave MI
3 4 5 6 7 8
90 77 74 52 74 67
Female Male Male Male Female Female
Fever in EDOU Fever in EDOU Ten beats of ventricular tachycardia Returned for recurrent atrial fibrillation Returned for recurrent atrial fibrillation Returned for chest pain
Outcome Cardiac catheterization showed mild-to-moderate coronary narrowing and was treated medically, patient was discharged in 3 d Admitted but signed out against medical advice 24 h later without workup, alive at 6-mo follow-up UTI found and treated, discharged in 12 d No cause found, discharged in 4 d Admitted, electrocardiologic testing, AICD placed, home in 6 d Chemically cardioverted and discharged in 2 d Chemically cardioverted and discharged in 3 d Negative serial CK-MB and graded exercise stress test results with myocardial perfusion imaging, discharged in 1 d for noncardiac chest pain
MI, Myocardial infarction; UTI, urinary tract infection; AICD, automatic implantable cardioverter defibrillator.
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 7 9
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
We did not identify any thromboembolic events in our treatment group. The incidence of these has been reported to be as high as 5% to 7% in the cardioversion of atrial fibrillation without anticoagulation.22 However, Weigner and Caulfield23 have shown a lower risk of thromboembolic events (0.8%) if symptoms have been present for less than 48 hours. Consequently, elective cardioversion without anticoagulation is only recommended for patients with true AAF. If symptoms have been present for longer than 48 to 72 hours or if the duration of symptoms is unclear, it is suggested that patients receive anticoagulation for 1 to 3 weeks before attempting cardioversion.24 Although our protocol called for admission of patients with persistent atrial fibrillation, 3 (4%) patients were discharged in atrial fibrillation. These patients were discharged on oral anticoagulants, with planned short-term outpatient follow-up and delayed cardioversion. These patients were seen by a cardiologist while in the EDOU, and follow-up arrangements were made. Thus, these exceptions were believed to be safe. On follow-up, no delayed complications were identified in these 3 patients. Traditional practice has been to admit patients with AAF in whom initial ED attempts to convert to sinus rhythm failed or to discharge patients once they converted to sinus rhythm and were without complications. In our EDOU study, 81% of patients with AAF were discharged after an average of 11.8 hours of treatment in the EDOU. Others have found that, compared with the accelerated care available in an EDOU setting, inpatient lengths of stay and costs are greater for chest pain and asthma.14,16,17 Because the focus of this preliminary study was the feasibility of this approach, we did not attempt a randomized comparison of inpatient and EDOU management of AAF. However, on the basis of our results, we believe that such a study may be warranted. Restricting the drugs used in our protocol may have allowed us to study which drugs worked best in our protocol. However, our purpose was to study the feasibility of this protocol and not the specific medications. We did not believe that the literature was compelling enough to mandate the use of any one drug or combination of drugs to be used in our protocol. For this reason, we standardized the basic treatment steps rather than the specific drugs used. A high percentage of patients treated only with medication to control ventricular response spontaneously converted to sinus rhythm. This raises the question of which is safer, rate control and a limited period of observation or the use of an antiarrhythmic, such as ibutilide, which has an 8.3% incidence of polymorphic ven-
3 8 0
tricular tachycardia.25 Propafenone was preferentially used for cardioversion at our institution because it was believed to have a superior safety profile relative to other cardioverting drugs of similar efficacy. We speculate that our conversion rates may have been slightly improved if all patients had been treated with chemical or electrical cardioversion. However, this study was not designed to answer that question. Again, future studies of AAF in an EDOU setting might explore the nature of the optimal treatment strategy. This case series has several inherent limitations. Because we used a case series of patients treated in an EDOU, we did not detail the clinical characteristics of the entire population of patients presenting to the ED with AAF. We believed that this would be beyond the focus of this case series. However, we did provide a general profile of this population in our review of ICD-9 data. Not all eligible patients were enrolled in the EDOU protocol because physician preference was allowed regarding admission to the EDOU or an inpatient bed. This may have introduced a referral bias, perhaps selecting a healthier subset of patients. However, we are aware of no setting in which the use of an EDOU is compulsory for all patients who might otherwise be admitted as an inpatient. Making the protocol compulsory for all patients with uncomplicated persistent AAF might similarly introduce a selection bias that would not reflect actual practice. Our intent was to assess the feasibility of a protocol that allowed discretionary enrollment, similar to what occurs in clinical practice. On the basis of our findings, we believe that a larger, prospective, randomized controlled trial comparing this EDOU protocol with inpatient treatment of all eligible patients with uncomplicated AAF would help to better define other outcomes, such as cost and patient satisfaction. In summary, we found that it is feasible for patients with uncomplicated AAF to be managed with a simple EDOU protocol, with a majority being successfully cardioverted and discharged in less than 12 hours. A small but significant number of patients may be found to have positive diagnostic findings. Further study of the management of uncomplicated atrial fibrillation in an EDOU setting regarding its effect on cost and patient satisfaction is warranted. ADDENDUM
The recent decision by the Centers for Medicare and Medicaid Services to separately fund observation services for only the 3 most well-studied conditions (chest pain, asthma, and congestive heart failure) underscores the
ANNALS OF EMERGENCY MEDICINE
39:4 APRIL 2002
OBSERVATION UNIT PROTOCOL FOR ATRIAL FIBRILLATION Koenig, Ross & Jackson
need to further study other conditions that may be amenable to EDOU treatment. A copy of the Federal Register comments made by the Centers for Medicare and Medicaid Services can be accessed at http://www.hcfa.gov/regs/Pp44671.pdf.26 Author contributions: MAR and BOK conceived the study, and REJ worked with them to refine its design. BOK conducted data collection. MAR, REJ, and BOK all analyzed and interpreted the data. MAR and BOK drafted the manuscript, while MAR, BOK, and REJ all contributed to revisions. MAR and BOK take responsibility for the paper as a whole.
22. Mancini GBJ, Foldberger AL. Cardioversion of atrial fibrillation: consideration of embolization, anticoagulation, prophylactic pacemaker and long term success. Am Heart J. 1982;104:617-621. 23. Weigner MJ, Caulfield TA. Risk for clinical thromboembolism associated with conversion to sinus rhythm in patients with atrial fibrillation lasting less than 48 hours. Ann Intern Med. 1997;126:615-620. 24. Laupacis A, Albers G, Dolen J, et al. Antithrombotic therapy in atrial fibrillation. Chest. 1995;108:352S-359S. 25. Stambler BS, Wood MA, Ellenbogen KA, et al. Efficacy and safety of repeated intravenous doses of ibutilide for rapid conversion of atrial flutter or fibrillation. Circulation. 1996;94:16131621. 26. Federal Register. Vol. 66, No. 165. Proposed rules, p. 44,690-44,692. Available at: http://www.hcfa.gov/regs/Pp44671.pdf. Accessed on August 24, 2001.
REFERENCES 1. Huagui L, Easley A, Barrington W, et al. Evaluation and management of atrial fibrillation in the emergency department. Emerg Med Clin North Am. 1998;16:289-303. 2.
Narayan SM, Cain ME, Smith JM. Atrial fibrillation. Lancet. 1997;350:943-950.
3. Upshaw C. Hemodynamic changes after cardioversion of chronic atrial fibrillation. Arch Intern Med. 1997;157:1070-1076. 4.
Brickner ME. Cardioembolic stroke. Am J Med. 1996;100:465-474.
5. Wijffells MC, Kirchof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation: a study in awake chronically instrumented goats. Circulation. 1995;92:1954-1968. 6. Friedman HZ, Scot FG, Bonema JD, et al. Acute complications associated with new-onset atrial fibrillation. Am J Cardiol. 1991;67:437-439. 7. Prystowsky EN, Benson DW, Fuster V, et al. Management of patients with atrial fibrillation; a statement for healthcare professionals from the Subcommittee on Electrocardiography and Electrophysiology, American Heart Association. Circulation. 1996;93:1262-1277. 8. Bialy D, Lehmann MH, Schumacher DN, et al. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol. 1992;19:41A. 9. Mulcahy B, Coates WC, Hennerman PL, et al. New-onset atrial fibrillation: When is admission medically justified? Acad Emerg Med. 1996;3:114-119. 10. Falk R. Medical progress: atrial fibrillation. N Engl J Med. 2001;344:1067-1078. 11. Michael JA, Stiell IG, Agarwal S, et al. Cardioversion of paroxysmal atrial fibrillation in the emergency department. Ann Emerg Med. 1999;33:379-387. 12. Joseph AP, Ward MR. A prospective, randomized controlled trial comparing the efficacy and safety of sotalol, amiodarone, and digoxin for the reversion of new-onset atrial fibrillation. Ann Emerg Med. 2000;36:1-9. 13. Innes GD, Vertesi L, Dillon EC, et al. Effectiveness of verapamil-quinidine versus digoxinquinidine in the emergency department treatment of paroxysmal atrial fibrillation. Ann Emerg Med. 1997;29:126-134. 14. Rydman RJ, Isola MI, Roberts RR, et al. Emergency department observation unit versus hospital inpatient care for a chronic asthmatic population: a randomized trial of health status outcome and cost. Med Care. 1998;36:599-609. 15. Rydman RJ, Roberts RR, Albrecht GL, et al. Patient satisfaction with an emergency department asthma observation unit. Acad Emerg Med. 1999;6:178-183. 16. Gomez MA, Jefferey LA, Karagounis LA, et al. An emergency department based protocol for rapidly ruling myocardial ischemia reduces hospital time and expense: results of a randomized study (ROMIO). J Am Coll Cardiol. 1996;28:25-33. 17. Farkouh ME, Smars PA, Reeder GS, et al. A clinical trial of a chest-pain observation unit for patients with unstable angina. N Engl J Med. 1998;339:1882-1888. 18. Rydman RJ, Zalenski RJ, Roberts RR, et al. Patient satisfaction with an emergency department chest pain observation unit. Ann Emerg Med. 1997;29:109-116. 19. Brillman J, Mathers-Dunbar L, Graff L, et al. Management of observation units. Ann Emerg Med. 1995;25:823-830. 20. ACEP policy statement. Emergency department observation units. Ann Emerg Med. 1995;25:863-864. 21. Ross MA, Naylor S, Compton S, et al. Maximizing use of the emergency department observation unit: a novel hybrid design. Ann Emerg Med. 2001;37:267-274.
APRIL 2002
39:4
ANNALS OF EMERGENCY MEDICINE
3 8 1