Key points

Introduction

The first emergency department (ED) observation units (EDOUs) focused on chest pain and potential acute coronary syndromes. However, most EDOUs now cover multiple other conditions that lend themselves to protocolized, aggressive diagnostic and therapeutic regimens. In this article, the authors discuss the management of 4 cardiovascular conditions that have been successfully deployed in EDOUs around the country.

Introduction

The first emergency department (ED) observation units (EDOUs) focused on chest pain and potential acute coronary syndromes. However, most EDOUs now cover multiple other conditions that lend themselves to protocolized, aggressive diagnostic and therapeutic regimens. In this article, the authors discuss the management of 4 cardiovascular conditions that have been successfully deployed in EDOUs around the country.

Atrial fibrillation

Patient selection

A successful EDOU protocol for AF relies on an expedited algorithm incorporating anticoagulation, rate and rhythm control, and mechanisms to ensure rapid outpatient follow-up. Managing patients with AF in the EDOU relies on appropriate patient selection. The ideal patients are stable, without cardiac ischemia, and with onset of AF of less than 48 hours duration. The systolic blood pressure (SBP) should be 90 mm Hg or greater and heart rate less than 110 beats per minute (bpm) at the time of disposition to the EDOU, after administration of any required intravenous medications for rate control. Patients with AF secondary to an underlying cause, such as sepsis, pulmonary embolism, gastrointestinal (GI) bleed, or other life-threatening conditions that require hospital admission, are not candidates for EDOU management. Other relative contraindications to placing patients in the EDOU include acute renal insufficiency, pregnant patients, patients with an unknown time of AF onset, and patients with an expected length of stay of greater than 36 hours due to other comorbidities or socioeconomic considerations complicating their care. Additionally, patients who have poor health literacy, lack close follow-up, or who are unable to adhere to their treatment plan are not appropriate EDOU patients because they will require more intensive hospital and case management resources.

Patient management

The management of patients with unstable AF is not appropriate for the EDOU and is not reviewed here. After initial ED evaluation, the next step is to decide whether to control the heart rate or convert patients to a normal sinus rhythm. In patients 65 years of age and older, there is no difference in 5-year outcomes of stroke or death with rate versus rhythm control. Therefore, in this age group, rate control and anticoagulation, if indicated, should generally be the EDOU goal.

Younger patients (aged <65 years) have been shown to benefit from rhythm control and should be considered for cardioversion. The risk factors for immediate stroke following cardioversion include an unknown time from onset, a history of transient ischemic attack (TIA) or stroke in the last 6 months, and valve disease. The focus in these patients should be on rate rather than rhythm control. Also, patients with electrolyte abnormalities, such as hypokalemia, or with digoxin toxicity are at higher risk of poor outcomes with immediate rhythm control and should be managed with rate control. If patients are deemed low risk for stroke, with an AF onset of 48 hours ago or less or if patients have been on anticoagulation for 3 weeks or more, then cardioversion can be considered. This low-risk group of patients can be ideally managed in the EDOU.

When controlling rate, the two most commonly used pharmacologic agents are β-blockers (commonly metoprolol) and nondihydropyridine calcium channel blockers (commonly diltiazem). There is no solid evidence of one agent being superior to the other. However, the authors do suggest providers avoid crossing classes; if patients are already taking an oral β-blocker, for example, a parenteral β-blocker should be considered for first-line therapy as opposed to initiating a calcium channel blocker. Following administration of the intravenous agent, patients should be started on the equivalent oral dose of the medication to maintain adequate rate control. The goal heart rate, before EDOU placement, should be less than 100 bpm while at rest and less than 120 bpm when ambulatory. These acceptable rates vary, and some guidelines suggest adequate rate control is achieved at less than 110 bpm while at rest and less than 130 bpm while ambulatory.

The availability of procedural sedation in the EDOU may dictate whether cardioversion occurs in the ED before observation placement. If planning a synchronized electrical cardioversion, it is recommended to use an initial 150 to 200 J of energy with anterior-posterior pad electrical placement using a biphasic waveform device. This method has been shown to successfully convert patients 90% of the time. For patients who are poor sedation candidates, chemical cardioversion may be preferred. The most common pharmacologic agents include intravenous procainamide (Vaughan Williams class Ia), oral flecainide (class Ic), or intravenous ibutilide (class III). Procainamide is not the preferred method in the United States and is not included in the American Heart Association’s (AHA) guidelines. If anticoagulation is going to be used, it is generally advised to start as soon as possible.

In some systems, rapid transesophageal echocardiography (TEE) may be available for patients with unclear or prolonged duration of their AF. These patients, if otherwise appropriate for EDOU management, could receive rate control, anticoagulation, and TEE to evaluate for cardiac thrombi. Cardioversion immediately after TEE could occur during the same procedural sedation. Local resources will dictate the feasibility of that particular strategy.

Following cardioversion or rate control, the provider must determine the need for long-term anticoagulation. The most commonly used and validated clinical tool for stroke and thromboembolism risk stratification is the CHA ₂ DS ₂ -VASc (congestive HF, hypertension, age, diabetes mellitus, prior stroke or TIA or thromboembolism, vascular disease, age, sex) score ( Table 1 ). The AHA recommends oral anticoagulation for a score of 2 or greater, whereas European guidelines recommend oral anticoagulation for scores of one or greater. The Canadian guidelines suggest giving aspirin to patients with a score of zero but with known coronary artery disease. Providers must also assess patients’ bleeding risk. One validated scheme is the HAS-BLED score. This scoring system applies one point for each of the following risk factors: H ypertension, A bnormal liver or kidney dysfunction, S troke, B leeding predisposition, L abile international normalized ratio (INR), E lderly (older than 65 years), and D rugs (use of antiplatelet medications, such as clopidogrel (Plavix) or aspirin, or alcohol consumption >8 drinks per week). A score of 3 or greater indicates a high risk of bleeding. If patients are at high risk of bleeding and have a high stroke risk, then the decision to prescribe anticoagulation should be made with patients weighing the risk and benefits of anticoagulation and using shared decision-making to come to a conclusion.

Historically, warfarin was the most commonly used anticoagulation agent; however, it requires blood levels to be drawn and interacts with many foods and medications. The non–vitamin K direct oral anticoagulants (NOACs, formerly novel oral anticoagulants, sometimes called target-specific oral anticoagulants or direct oral anticoagulants [DOACs], the term that is used through the remainder of this work) have become increasingly popular in recent years and are generally preferred over warfarin in patients with nonvalvular AF. These medications are more convenient for patients and have proven to be equally safe and effective. Additionally, there is no requirement to bridge with unfractionated heparin (UFH) or low-molecular-weight heparin (LMWH) with rivaroxaban or apixaban, as there is with warfarin. DOACs are also preferred in situations whereby patients cannot reliably get their INR blood levels checked or adhere to the diet restrictions required when taking warfarin. If patients have renal failure or new acute renal insufficiency or hepatic disease, then consultation or discussion with cardiology regarding appropriate anticoagulation is recommended.

If at any point patients become unstable or the EDOU treatment protocol fails, patients should be formally admitted to the hospital for further management and evaluation.

Education

Admission to the EDOU provides a unique opportunity to educate patients on their diagnosis of AF. The discussion should focus on determining patients’ AF symptoms and identifying their specific triggers. If patients are started on anticoagulation, they should be educated about bleeding risks and potential bleeding complications. They should also be instructed to carry a medication card in their wallet or wear a medical alert bracelet that informs emergency medical services and other providers that they are on an anticoagulation medication. Additionally, if patients are being discharged on rate or rhythm control medications, they should be educated about the common side effects of the medications and the importance of taking the medications as prescribed. The most important aspect of patient education is emphasizing the importance of early outpatient follow-up.

Discharge and follow-up

In order for patients to be discharged from the EDOU, they must demonstrate clinical stability with either their AF rate controlled or their rhythm converted to normal sinus for at least 1 to 2 hours. Most EDOU protocols also require negative serial cardiac biomarker tests before discharge if ACS is suspected as a contributory factor. Outpatient follow-up in 3 to 5 days following discharge should be established before patients leave the EDOU. Additionally, a 30-day prescription for the selected method of anticoagulation, if indicated, should be given before discharge. If patients were not cardioverted and are being discharged on rate control medications, they should be prescribed the lowest long-acting oral dose equivalent to what they received in the EDOU. Finally, patients and their families should be provided with strict return precautions so that they understand when to return to the ED or hospital for care. Following these guidelines will allow for a successful EDOU visit and early discharge, thus, minimizing the burden of an AF diagnosis on patients’ daily life.

Acute heart failure

Patient selection

Obviously, not all patients with acute HF are appropriate for observation care. Patients at high risk for inpatient morbidity or mortality, those who will require inpatient procedures or critical care acuity, or those with socioeconomic challenges that would preclude safe disposition within 48 and, ideally, 24 hours will not be good candidates for observation management. Multiple risk stratification schemes have been published that identify high-risk patients. Therefore, it is cognitively simple to identify those patients who are completely inappropriate for observation care. It is important to realize, however, that the absence of a high risk for mortality does not mean that patients are considered low risk and, therefore, appropriate for rapid disposition from an EDOU.

The Society of Chest Pain Centers (now the Society for Cardiovascular Patient Care) published recommendations on risk stratification as well as EDOU management of acute HF syndrome, which were later validated in an external data set. However, the strength of evidence for most of these recommendations was based on expert consensus. At that time, a definitive evidence base for patient risk stratification from the ED for acute HF was lacking. This evidence deficit continues to the current era. The literature base on ED risk stratification and the specific identification of low-risk, short-stay-appropriate patients continues to evolve ; but further work is needed to define a broadly applicable, validated decision tool that adds to physician decision-making. Therefore, the statements that follow are based on the best available evidence but should be considered recommendations, as opposed to rigid guidelines, in most cases.

Most patients with newly diagnosed HF presenting with acute symptoms should be admitted to an inpatient unit. The initial workup and management required will generally surpass the recommended 24-hour observation time. Mild cases of new-onset HF could conceivably be referred for outpatient management, but this may not be typical of patients’ acuity presenting to the ED. Likewise, access to appropriate outpatient follow-up in a timely fashion from the ED may be difficult for many patients.

Any new ischemic changes noted in the electrocardiogram (EKG) should be exclusion criteria for EDOU management, as should an unstable cardiac arrhythmia. Hypotension is less clear. Patients presenting with a SBP less than 120 mm Hg have been shown to have increased inpatient mortality. However, given the adverse effects of increased afterload in chronic HF, target blood pressures for outpatient management are frequently lower than this. Therefore, it is more useful to recommend that poor perfusion, or symptomatic hypotension, serve as the criterion for EDOU exclusion as opposed to an absolute number.

Dyspnea is one of the most common presenting symptoms of acute HF. After initial treatment, respiratory rate and oxygen requirements should be taken into account when deciding on EDOU versus inpatient admission. Although noninvasive ventilatory support (NIV) may reduce respiratory distress and improve physiologic parameters, patients who are unable to be weaned in the ED should be admitted to the hospital and are not appropriate for EDOU management. However, the patient response to initial management should be considered as well. Flash pulmonary edema may initially require aggressive management with afterload reduction and NIV. Frequently, though, these patients quickly respond and no longer require intensive care unit–level interventions within hours of treatment initiation. If patients are rapidly improving and have successfully been weaned from NIV and parenteral vasodilators, EDOU management may be considered if respiratory distress has resolved with initial treatment. If patients require more than minimal oxygen supplementation, are not improving with ED treatment, or have substantial persistent respiratory distress, they should be admitted to the inpatient setting.

Troponin elevation is associated with inpatient mortality, and patients with normal troponin levels on presentation have a greater likelihood of successful discharge from the EDOU within 24 hours and lower adverse event rates at 30 days. However, many patients with stable chronic HF have detectable or elevated troponin levels at baseline, although these patients do have poorer long-term survival. If the physician has access to previous troponin measurements that show chronic mild elevation, a trial of observation with serial biomarker measurement may be reasonable if the presenting troponin level is not elevated beyond the baseline. However, if these previous data are not available or if patients present with symptoms of angina concomitant with acute HF, troponin elevation should prompt an inpatient admission.

Multiple studies have evaluated the prognostic value of natriuretic peptides (NP), most commonly B-type natriuretic peptide (BNP) or its biological precursor N-terminal pro-BNP (NT-proBNP). The authors acknowledge that there are many nuances that apply to the interpretation of NP testing, including the effect of age, adiposity, renal clearance, and preserved versus reduced ejection fraction. In general, though, the higher the NP level, the more specific the diagnosis of acute HF and the higher the risk of mortality. There is insufficient evidence to define an explicit NP level that delineates admission from observation with certainty, although a BNP greater than 1000 pg/mL or NT-proBNP greater than 5000 pg/mL has been suggested as the threshold for admission. When available, outpatient or discharge dry-weight NP levels may aid in the interpretation of the test obtained in the ED.

Patients who experience acute kidney injury and decreased renal function have been shown to have increased in-hospital mortality and postdischarge mortality. Fonarow and colleagues showed that both serum urea nitrogen (BUN) greater than 43 mg/dL and serum creatinine greater than 2.75 mg/dL were markers of increased inpatient mortality. Other studies have shown that elevated creatinine greater than 3.0 mg/dL and BUN greater than 40 mg/dL are strongly correlated to mortality. In addition to being a poor prognostic factor, impaired renal function complicates rapid, aggressive diuresis that may be required to successfully discharge patients from the EDOU within 24 hours.

Patient management

The goal of therapy in the EDOU should be to reduce symptoms, continue risk stratification, avoid unnecessary admission to the hospital, provide patient education and link follow-up after discharge.

Patients should have a recorded weight and oxygen requirement on arrival to the unit. Strict input and output measurement should be initiated at the start of their EDOU stay. Low urine output may indicate a need for higher loop diuretic doses, the addition of synergistic thiazide diuretics, or worsening renal failure necessitating admission. Given the high rate of malignant atrial and ventricular arrhythmias in patients with HF, cardiac monitoring is required.

Electrolytes should be measured to ensure that hypokalemia or hypomagnesemia is not induced with aggressive diuresis. The authors recommend serial troponin measurements for all patients being treated in an EDOU for HF, regardless of cause. Elevated troponin is a marker of worsening prognosis in both ischemic and nonischemic cardiomyopathy. In patients with ischemic cardiomyopathy, increasing troponins may indicate a need for admission and revascularization. The practice of following serial NP levels to assess response to therapy and readiness for discharge has not been validated in the literature.

Patients will need fluid restriction (<2000 mL/d) and to be placed on a 1500-mg sodium diet. Home medications should be restarted in the absence of intolerance or contraindication. Ideally, this will consist of an angiotensin-converting enzyme inhibitor (ACE) or angiotensin receptor blocker as well as a β-blocker. Contraindications to ACE include previous angioedema, pregnancy, hypotension, renal artery stenosis, creatinine levels greater than 3 mg/dL, and hyperkalemia (serum levels >5.0 mEq/L). To date, there are no clinical trials evaluating the initiation of an angiotensin receptor-neprilysin inhibitor in the setting of acute HF in the ED or EDOU.

Most patients with acute HF present acutely and typically have an SBP greater than140 mm Hg. These patients often present with greater initial symptoms but the potential for overall shorter length of stay and lower mortality. These patients are often in afterload crisis, with a fluid distribution problem to the lungs but not total body fluid overload. There may be benefit from a strategy of emphatic afterload reduction and modest diuresis.

Those patients who are normotensive tend to experience a gradual accumulation of total body fluid and present with milder dyspnea and greater peripheral edema compared with hypertensive patients. Because total body fluid is often the cause of their presentation, these patients can be more aggressively diuresed. Patients who are compliant with their diuretic medication but still presenting with volume overload should receive an initial intravenous dose that is at least equal to or exceeds their usual oral daily dose, followed by initiation or escalation of their outpatient regimen. Felker and colleagues found no benefit to administering diuretics in continuous infusion as opposed to bolus dosing in the DOSE (Diuretic Optimization Strategies Evaluation) trial. Escalation of vasodilators should be used with caution as these patients can become hypotensive quickly and must be reassessed after initial diuresis. Initiation of a synergistic thiazide diuretic should be considered if there is inadequate response to the initial intravenous diuretic dose.

Education

EDOUs provide clinicians more time to not only treat symptoms but to take a more in depth look at what lead to the HF exacerbation. The EDOU also allows a better environment for patient education.

Many hospitalizations could be avoided with adherence to medication and diet and by monitoring developing symptoms. As the evidence base for HF therapy grows, polypharmacy is a frequent consequence. Providers should look for common precipitants of medication noncompliance, including financial barriers, depression, forgetfulness, real or perceived side effects, and understanding the importance of the medication. For those who are unable to afford medications, many pharmaceutical companies offer assistance programs; consultation with the social work or case management teams may provide an important contribution to the patients’ care.

Dietary nonadherence can also contribute to acute HF exacerbations. The AHA currently advocates restriction of sodium intake to 1500 mg per day, compared with the daily American average of greater than 3400 mg per day. It is important to teach each patient how to read nutrition labels and to provide common low-sodium options for dining out and cooking at home using spices and herbs. Every attempt should be made to include family and friends in regard to education of diet and medication adherence.

Self-care education has been associated with decreased HF readmission rates. Educating patients about signs and symptoms associated with worsening HF along with their families will provide an opportunity to have multiple people watching for changes in patients’ physical status. Signs and symptoms of increasing dyspnea, increase in weight gain, and increasing fatigue should be stressed to both patients and families.

Disposition and follow-up

There is scant evidence defining when patients are ready for discharge after an EDOU stay for acute HF. Often, patients who are discharged are done so based on resolution dyspnea and peripheral edema. The amount of improvement required to avoid a revisit to the ED has not been quantified. If fluid balance, weight, oxygen requirement, and vital signs have not improved during the observation period or if patients have a decline in cardiac, pulmonary, or renal function, they should be admitted to the hospital for further management.

Before discharge, patients should receive education regarding medication and diet compliance as well as education of signs and symptoms of early HF exacerbation. It is appropriate to collaborate with the primary physician or cardiologist regarding adjustments to the medication regimen at discharge. A recent meta-analysis evaluated home-health visits, HF multidisciplinary clinic referral, and telephone follow-up; all reduced mortality and HF readmissions. Early physician follow-up, regardless of whether with cardiology or primary care, also reduces HF recidivism. Early on in the EDOU stay, case management should be engaged to begin discharge planning to include specialty clinic referral, physician follow-up, and home health services when appropriate. Even if patients do not improve substantially and require admission from the EDOU, they will still require outpatient care; early initiation of discharge planning has been shown to improve postdischarge outcomes.

Syncope

Patient selection

It should be clear that the authors are discussing the management of patients without a clear cause for syncope after the initial ED evaluation. When a clear cause, such as subarachnoid hemorrhage, pulmonary embolism, acute GI bleeding, hypovolemia, or malignant arrhythmia, is diagnosed during the initial ED course, the underlying condition should be addressed. Although certain conditions may lend themselves to observation management (dehydration secondary to gastroenteritis, for example), severe disease will merit inpatient admission.

Multiple risk stratification rules have been published for syncope, but none have demonstrated an improvement in outcome over physician judgment. Published studies have suggested a range of markers of both high and low risk for poor outcomes, resulting in nebulously defined and inconsistent recommendations from various specialty society guidelines and invalidated decision rules. The most commonly cited markers of risk center on cardiac dysfunction and include family history of sudden cardiac death, severe cardiac disease (including dilated cardiomyopathy, aortic stenosis, hypertrophic cardiomyopathy, or left ventricular [LV] outflow disease), EKG abnormalities (such as preexcitation syndrome, Brugada syndrome, QT interval abnormalities, or conduction blocks), or exertional syncope.

Patients who are considered low risk should usually be discharged. If all of the following characteristics apply, admission or observation is not likely to be value added: age less than 50 years, clear vasovagal cause, no history of cardiovascular disease, normal EKG, and a normal cardiovascular examination.

Specialty society guidelines have indicated that the history of cardiovascular disease should be considered a marker of high risk. However, Numeroso and colleagues recently published research suggesting that the simple history of cardiovascular disease (known chronic HF or known coronary artery disease), unless presenting with an acute exacerbation, does not convey elevated risk for adverse outcomes. Likewise, increasing age was associated with more high-risk features but was not independently predictive of adverse outcomes. Although requiring external validation and confirmation, this would suggest that patients with clinical stability but a history of cardiac disease who would otherwise not meet high-risk criteria, or those with advanced age but no other high-risk features, could potentially be managed in an EDOU.