Treatment of Acute Heart Failure in the ED
Treatment of Acute Heart Failure in the ED
Despite considerable heterogeneity in patient characteristics, on presentation to the ED, the sensation of breathlessness, or dyspnea, predominates. This symptom reflects underlying pulmonary congestion caused by elevated left ventricular end-diastolic pressure with or without a reduced cardiac output. Classic signs and symptoms of HF, including peripheral edema, elevated jugular venous pressure, an S3 gallop, rales, dyspnea and orthopnea are frequently evident at the time of presentation. Although the clinical picture can vary substantially, the vast majority of patients fit a 'warm and wet' profile on arrival. Thus, our initial approach to management largely focuses on this group (Table 1).
HF is a clinical diagnosis, but accuracy of physical examination is limited, particularly for ruling-out the condition. Although initial clinical impression has decent specificity (0.86), it has limited sensitivity (0.61). A third heart sound, hepatojugular reflux and elevated jugular venous pulse have even greater specificity (0.92–0.99), but their sensitivity is poor, ranging from 0.13 to 0.39.
Chest x-ray findings such as pulmonary venous congestion and interstitial edema suffer from the same limited sensitivity as the physical exam, though they are more specific. Overall, when present, both physical exam and radiographic findings are very helpful; however, absence of these does not exclude a diagnosis of AHF. Natriuretic peptide (NP) testing greatly enhances diagnostic accuracy and is recommended in the evaluation of suspected AHF. NP testing (BNP or NT-proBNP) has a Class I, Level A (best evidence) guideline recommendation to support diagnostic decision-making, especially when diagnosis is in question. This is based on the seminal paper by Maisel et al. and then confirmed in multiple other studies demonstrating both the diagnostic and prognostic value of NP testing. However, because diagnostic uncertainty and misdiagnosis remain prevalent in clinical practice, we recommend that NP levels be obtained routinely when AHF is suspected but not confirmable clinically.
Once the diagnosis has been established, we recommend division of patients into clinical profiles or phenotypes. Several classification schema have been proposed. We advocate the use of the Gheorghiade/Braunwald initial assessment together with the SBP phenotyping suggested by Collins et al. (Figure 1). Of note, neither has been prospectively tested in regards to their impact on outcomes.
(Enlarge Image)
Figure 1.
Initial approach to acute heart failure management algorithm.
Modified and reproduced with permission from [24].
The six-axis model proposed by Gheorghiade and Braunwald highlight the essential considerations in early AHF management: clinical severity; initial blood pressure; heart rate and rhythm; precipitants; de novo versus chronic HF; and comorbidities. At first glance, the absence of EF may be seen unusual; however, at the present time, with the exception of cardiogenic shock or severe advanced HF (which is subsumed under the category of clinical severity), knowledge of the EF does not significantly impact initial management, which focuses on congestion relief and hemodynamic improvement. Knowledge of EF remains critical, however, for pre- and post-discharge management.
As shown in Table 1, stabilization and treatment of life-threatening conditions in AHF is the first priority. Assessment of the airway-breathing-circulation are paramount. Upto 5% of AHF patients will require mechanical ventilation. Overall, meta-analysis supports the benefits of noninvasive ventilation (NIV) in cardiogenic pulmonary edema, showing that early use can prevent the need for endotracheal intubation. However, in the Cardiogenic Pulmonary Oedema trial (3CPO), other than an improvement in dyspnea, no differences were seen in the rates of death or intubation between patients treated with facemask oxygen, continuous positive airway pressure only or bi-level positive airway pressure (i.e., expiratory and inspiratory). Although 3CPO was the largest randomized trial to investigate different methods of oxygen delivery in AHF patients, we continue to recommend the early use of NIV for appropriate patients who present with even moderate severity, given its potential to alleviate symptoms, its noninvasive nature and the intended short-term duration of use.
Once stabilized, treatment of AHF patients is based on three primary BP phenotypes: hypertensive (SBP >140 mmHg), normotensive (90–140 mmHg) and hypotensive <90 mmHg). Note that these cutpoints are intended as guidelines and are not absolute thresholds.
For the hypertensive AHF patient (Figure 2), emphasis is on vasodilation rather than fluid removal, as these patients symptoms may be due more to volume re-distribution, as opposed to total volume overload. This concept of 'vascular' failure is best exemplified by the 'flash' pulmonary edema patient; sudden onset as opposed to days or even weeks of slowly progressive signs and symptoms. Although intravenous (iv.) loop diuretics are still recommended, we emphasize the use of vasodilators for initial treatment.
(Enlarge Image)
Figure 2.
Algorithm for the initial treatment of hypertensive (systolic blood pressure >140 mmHg) acute heart failure.
iv.: Intravenous; NIV: Noninvasive ventilation; SBP: Systolic blood pressure.Modified and reproduced with permission from [24].
Nitroglycerin, nitroprusside, nesiritide, hydralazine, angiotensin convering enzyme-inhibitor (ACE-I) and calcium channel blockers have all been used for initial vasodilation. Although experience is greatest with nitrates and multiple studies support their use, nesiritide is the only FDA-approved vasodilator that has been studied in a large, definitive randomized controlled trial. Only one other agent has recently completed Phase III studies with positive findings. In the RELAX-AHF trial, a Phase III study of serelaxin versus placebo, both in addition to standard of care in 1161 patients, the primary end point of dyspnea improvement was achieved. This improvement in dyspnea was observed only by visual analog scale, not by Likert. A statistically significant reduction in the secondary end point of 180-day cardiovascular (CV) mortality was also seen in the serelaxin group; however, there was no improvement in the composite of 60-day CV death or readmission for HF/renal failure. At present, serelaxin lacks FDA approval. Further studies are currently being planned.
Despite the absence of large-scale trials, nitroglycerin is the most widely used vasodilator. Due to its familiarity, rapid onset and clearance, and inexpensive cost, we advocate for its use as first-line therapy for hypertensive AHF (see Table 2 for doses). The sublingual route is recommended initially followed by iv. delivery for those who continue to have symptoms. High doses, much higher than traditionally used, have been shown to be both safe and effective. Cotter et al. demonstrated lower rates of intubation and myocardial infarction as well as improvements in heart rate, respiratory rate and oxygen saturation with high-dose nitrates compared with lower doses. Levy et al. showed similar benefits. Although we prefer the iv. route in the ED setting, sublingual and topical nitrate application are a valid alternative and have also been shown to be safe and effective at reducing NP levels more rapidly within 48 h outside of the ICU setting. Concerns regarding nitrate tolerance in the ED setting are unfounded as the goals of therapy are short term (i.e., dyspnea relief through afterload reduction), and there is no outcome benefit with protracted infusion. However, a small proportion of patients may be nitroglycerin unresponsive as noted by frequent up titration without clinical response. Nesiritide may also be considered first-line treatment, though we use it clinically as second- or third-line therapy. Initial concerns regarding its safety via meta-analysis have been laid to rest by prospective study (ASCEND-HF trial). Its efficacy, however, remains debated, as only a small, but statistically significant improvement in dyspnea was observed.
Other vasodilators such as nitroprusside, hydralazine, ACEI and dihydropyridine calcium channel blockers have also been used in the ED setting. Given the limited evidence base for traditional therapies, it is difficult to state that these other vasodilators are contraindicated in AHF. Despite the effectiveness of nitroprusside at vasodilation, its use in the ED is often limited by requirements for central line delivery and invasive blood pressure monitoring. Small studies suggest the potential efficacy of both ACEI and dihydropyridine calcium channel blockers. With ACEI, there is the potential risk of protracted, first-dose hypotension, which may lead to adverse events, either immediately or downstream. Although the absence of robust safety data does not rule out utilization of alternative vasodilatory therapies, at the present time, nitrates remain our preferential first-line treatment for this AHF phenotype. The normotensive patient (Figure 3) is the prototypical patient with known chronic HF and reduced EF who decompensates. The clinical presentation is usually less dramatic compared to hypertensive AHF, with historical features of progressive worsening signs and symptoms, including fatigue, weight gain, dyspnea on exertion and peripheral edema. Often, escalation of outpatient oral diuretic therapy has failed. For these patients, emphasis is on decongestion through fluid removal. In the ED setting, iv. loop diuretic therapy is the mainstay of fluid overload management.
(Enlarge Image)
Figure 3.
Algorithm for the initial treatment of normotensive acute heart failure (systolic blood pressure of 100–140 mmHg).
iv.: Intravenous; NIV: Noninvasive ventilation; SBP: Systolic blood pressure.
Modified and reproduced with permission from [24].
Despite their widespread use, the ideal dose and method of delivery (bolus versus continuous infusion) continues to be debated. Loop diuretics have been associated with transient worsening of pump function, brief initial elevation of filling pressures, worsening neurohormonal profile, electrolyte abnormalities, renal injury and by secondary analysis, worsening mortality. Importantly, however, there is no definitive proof that iv. loop diuretics given to AHF patients lead to worse outcomes.
The recent DOSE-AHF trial tested high- versus low-dose iv. loop diuretic as well as continuous versus bolus strategies in AHF patients. The study randomized 308 AHF patients within 24 h of presentation. They found that high-dose loop diuretic, defined as 2.5-times the total oral dose divided throughout the day led to greater diuresis and dyspnea improvement compared with the low-dose strategy, defined as the total outpatient oral diuretic dose. The study was not powered for mortality and no differences between groups were identified, but the high-dose group did have transient worsening of renal function. In addition, there were no significant differences between continuous versus bolus iv. diuretic treatment. Thus, we recommend high-dose iv. bolus loop diuretic therapy (i.e., 2.5-times the total outpatient oral loop diuretic dose, divided based on twice a day or 3-times a day dose scheduling) at the time of ED presentation.
It is worth highlighting that many AHF studies do not evaluate patients in the ED, but usually within 24 h or longer from presentation. Whether results may have differed or been affected depending on time of enrollment remains debated. For the hypotensive AHF patient (SBP <90 mmHg), Figure 4 outlines the initial approach. Importantly, this is an uncommon phenotype when considered in the context of the one million hospitalizations for AHF that occur every year. Less than 5% of patients present with SBP <90 mmHg, though the frequency of hypotensive AHF presentation varies depending on hospital type, with a higher reported frequency among some hospitals, especially those that perform cardiac transplantation. Patients in cardiogenic shock without a previous history of HF will present with a more severe clinical manifestations and often a life-threatening precipitant (i.e., acute MI) is the underlying cause. It is the patient with a history of advanced or end-stage HF, defined as persistent HF signs and symptoms despite maximal medical therapy, where careful assessment of volume status is critical. Many of these patients may have chronically low SBP due to the severity of their left ventricle (LV) systolic dysfunction and the effects of concurrent HF therapy. Although worsening HF as the cause of their acute presentation is likely, these patients may also be intravascularly depleted secondary to recent up-titration of outpatient diuretic therapy. Furthermore, as these patients often have multiple comorbidities, other conditions (i.e., infection) may be the primary precipitant with worsening HF a result.
(Enlarge Image)
Figure 4.
Algorithm for initial treatment of hypotensive acute heart failure (systolic blood pressure <100 mmHg).
HF: Heart failure; IABP: Intra-aortic balloon pump; iv.: Intravenous; SBP: Systolic blood pressure.Modified and reproduced with permission from [24].
Perhaps, the most critical point to emphasize in the management of hypotensive AHF is the purpose of treatment, which is to improve hypoperfusion, not simply raise blood pressure. In the past, inotropes were used to optimize the hemodynamic profile of AHF patients despite a clinical presentation inconsistent with shock or advanced HF. Although it is true that inotropes such as dobutamine and milrinone can help the AHF patient in the short term, they are associated with worsening morbidity and mortality. Thus, current guidelines recommend judicious use of inotropic therapy and only in those patients who truly require it. Although dobutamine and milrinone are the most common inotropes used for treatment of hypotensive AHF, they are not the only means of providing hemodynamic support for those who need it. At moderate doses (5–10 mcg/kg/min), dopamine does provide reasonable inotropic and chronotropic effects but at higher doses (>10–20 mcg/kg/min) vasoconstrictive activity predominates. Although the latter will help improve blood pressure, significant increases in afterload may be undesirable in some patients, particularly those with poor systolic function who cannot generate sufficient cardiac work to overcome added impedance to forward blood flow. Alternatives to increase myocardial contractility such as digoxin and levosimendan (available in Europe only) should be considered when patients are on chronic β-blocker therapy as this can diminish the effectiveness of adrenergic receptor agonists. Cardiac myosin activators, a promising new class of inotropes currently under study, work by directly stimulating the contractile apparatus. They may ultimately prove useful for this group of patients and hypotensive AHF in general, as they have none of the deleterious, mechanistic effects of existing agents. For the rare patient with refractory cardiogenic shock (most often ischemic in origin), aortic balloon counterpulsion or emergent intervention with a percutaneous left ventricular assist device may be needed.
Sinus tachycardia should improve with treatment of AHF. Failure to improve suggests an underlying precipitant or etiology that has yet to be identified.
Atrial fibrillation is the most common dysrhythmia complicating AHF management. Similar to the lack of evidence to guide AHF management, there is a paucity of robust data to guide clinicians regarding AF with rapid ventricular response and AHF. Guidelines recommend the use of amiodarone and digoxin for AHF patients with reduced LV ejection function without an accessory pathway, though there is a risk of cardioversion with amiodarone, which should be carefully considered. Caution is also warranted regarding more traditionally used agents to control HR, such as nondihydropyridine calcium channel blockers and β-blockers, given their longer duration of action and negative inotropic effects. Despite these recommendations, anecdotally, nondihydropyridine calcium channel blockers are commonly used. If used, we recommend caution and to start with low doses.
AHF patients who present in shock or extremis secondary to AF require immediate cardioversion. For the vast majority of cases, management is similar to AF management principles without AHF: rate control; consideration of rhythm management; and stroke prevention. Of note, risk of stroke is even higher in HF patients.
It can be difficult to determine if AF is the cause or result of AHF. For patients with de novo HF and AF with rapid ventricular response, AF is the more likely culprit. However, for the majority of patients who have chronic HF and chronic AF, rapid ventricular response often results from HF decompensation. Treatment of AHF will often decrease the HR. It is also important to note that in the setting of AHF, increased HR (even if in AF) at rates of 90–110 b.p.m. may be a compensatory response to increase cardiac output since stroke volume is reduced and can be 'fixed' (i.e., inability to increase stroke volume in response to stress) in AHF patients. Knowledge of LV ejection fraction can help guide initial therapeutic management of AF and AHF. Patients with AHF and a preserved EF are particularly vulnerable to tachycardias due to decreased filling time in conjunction with a noncompliant ventricle. Negative inotropic effects of β-blockers or nondihydropyridine calcium channel blockers are less of a concern; as such either class of drug can be useful for rate control in the patient with AHF, preserved EF and tachycardia. Of note, in more advanced stages of HFpEF and in those with restrictive cardiomyopathy, the LV may become so stiff that reducing HR below a certain threshold may result in a reduction in cardiac output, as stroke volume is reduced and fixed. In patients with reduced LV ejection fraction, the greater the severity of LV dysfunction, the greater the caution in using agents other than digoxin and amiodarone. Treatment of AHF may improve the HR substantially and thus rate control may appropriately be a secondary goal.
Concurrent with initial treatment, identification of the precipitant of HF decompensation is a primary goal of early management. In other words, why did the patient decompensate? Dietary indiscretion, medication noncompliance, pneumonia/respiratory process, ischemia and arrhythmias are common precipitants. Determining the precipitant targets, the cause of decompensation and not just the outward manifestations. Although an exhaustive search may not be feasible in the ED, timely identification of precipitants such as infection, ischemia, arrhythmia or pulmonary embolism may prevent or attenuate in-hospital morbidity.
After initial stabilization and management, disposition decisions must be made. As a general rule, patients who present with HF for the first time should be hospitalized. These individuals have not undergone previous work-up for potential etiology and/or precipitant and likely have not had recent evaluation of cardiac structure and function. Although it is possible that initiation of evidence-based therapies could begin as an outpatient, hospitalization allows for more efficient and in-depth evaluation and education. Culprit pathology amenable to rapid intervention can also occur during hospitalization. Decompensated chronic HF patients who present with AHF have, presumably, already undergone in-depth interrogation for potential reversible causes as well as optimization of all available evidence-based therapies. Prior in part driven by how recent will impact hospital management. Figure 5 below highlights potential targets or risk features to be considered during or soon after hospitalization.
(Enlarge Image)
Figure 5.
Comprehensive assessment and cardiac reconstruction.
Viable but dysfunctional myocardium.
Select patients.
Investigational agents.
ACE-I: Angiotensin converting enzyme inhibitor; AF: Atrial fibrillation; AHF: Acute heart failure; ARB: Angiotensin receptor blocker; CABG:
Coronary artery bypass grafting; CAD: Coronary artery disease; CRT: Chronic resynchronization therapy; Hydral: Hydralazine; ICD: Implantable cardiac defibrillator; ISDN: Isosorbide dinitrate; JVP: Jugular venous pulse; LV: Left ventricle.
Modified and reproduced with permission from [22].
As previously mentioned, both CV and non-CV conditions are prevalent in AHF patients. Consideration of comorbidities as a potential precipitant is therefore important and may impact treatment and prevention of readmission. In the ED, unless the principal precipitant (e.g., worsening COPD contributing to an acute decompensation of HF) can be diagnosed and managed acutely, management of comorbidities remains largely a focus of in-hospital or post-discharge management.
In the USA, EDs see over 136,000,000 patients annually. As such, patient throughput is an additional consideration when treating all types of patient, to avoid delays in the care of patients waiting to be seen. Thus, efficient risk stratification and disposition of patients are critical.
In terms of AHF, over 80% of patients evaluated in the ED are subsequently admitted to the hospital, yet past studies have suggested that up to 50% could either be discharged or held in observation status. Multiple markers of risk have been identified that could potentially aid in disposition decision making (Table 3).
At the present time, for most of these prognostic markers, little can be done in the ED to alter the trajectory of these patients. Whether treatment to alter these markers of risk improves outcomes remains an area of active, ongoing research. Although concern for litigation is understandable, especially given the high post-discharge morbidity and mortality associated with AHF, we would argue that this should not be the principal reason behind decision making in the ED. Rather, it is the lack of robust risk stratification instruments combined with a population that is older with multiple comorbidities that drives the high admission and re-admission rate. Risk stratification traditionally focuses on the high-risk patient, assuming that intervention will lower risk. However, the opposite approach is needed in the ED. Who is low risk? There is often little debate regarding the admission of AHF patients with high-risk features, especially when changed from baseline. Rather, it is the patient without high-risk features that often generates debate.
Absence of high risk does not equal low risk. Even the lowest risk group stratified by SBP had a 90-day mortality rate of 5.4% and re-hospitalization rate of 27.6%. Risk stratification instruments have largely been performed on hospitalized patients, limiting their applicability to the ED setting where the impact of hospitalization will not be realized.
Although some risk stratification tools have been created for ED purposes, they either have not been well validated or are cumbersome to use. In addition, outcomes of discharged AHF patients have not been well studied, highlighting our limited understanding of this population. Analysis from a large Canadian administrative database, however, suggests that clinician impression is a poor discriminator, as discharged patients had worse outcomes than those admitted. Although there is insufficient evidence to support strong recommendations regarding ED disposition decision making, there is little evidence to support current strategies focused on reducing 30-day re-admission. Thus, we recommend hospitals or health care systems develop structures and processes to determine various levels of risk to aid in disposition based on local expert consensus, emphasizing safety. A draft proposal of a potential algorithm is proposed below (Figure 6). Importantly, this algorithm has not been validated and requires further study prior to use. It does, however, provide a starting point for discussion. As mentioned, the evidence supporting current plans to decrease 30-day re-admissions are also lacking in evidence.
(Enlarge Image)
Figure 6.
Proposed algorithm for risk stratification in the emergency room.
BNP: B-type natriuretic peptide; BUN: Blood urea nitrogen; DM: Diabetes mellitus; ED: Emergency department; ICU: Intensive care unit; OU: Observation unit; SBP: Systolic blood pressure.
Modified and reproduced with permission from [76].
Initial Approach to Management in the ED
Despite considerable heterogeneity in patient characteristics, on presentation to the ED, the sensation of breathlessness, or dyspnea, predominates. This symptom reflects underlying pulmonary congestion caused by elevated left ventricular end-diastolic pressure with or without a reduced cardiac output. Classic signs and symptoms of HF, including peripheral edema, elevated jugular venous pressure, an S3 gallop, rales, dyspnea and orthopnea are frequently evident at the time of presentation. Although the clinical picture can vary substantially, the vast majority of patients fit a 'warm and wet' profile on arrival. Thus, our initial approach to management largely focuses on this group (Table 1).
Diagnosis
HF is a clinical diagnosis, but accuracy of physical examination is limited, particularly for ruling-out the condition. Although initial clinical impression has decent specificity (0.86), it has limited sensitivity (0.61). A third heart sound, hepatojugular reflux and elevated jugular venous pulse have even greater specificity (0.92–0.99), but their sensitivity is poor, ranging from 0.13 to 0.39.
Chest x-ray findings such as pulmonary venous congestion and interstitial edema suffer from the same limited sensitivity as the physical exam, though they are more specific. Overall, when present, both physical exam and radiographic findings are very helpful; however, absence of these does not exclude a diagnosis of AHF. Natriuretic peptide (NP) testing greatly enhances diagnostic accuracy and is recommended in the evaluation of suspected AHF. NP testing (BNP or NT-proBNP) has a Class I, Level A (best evidence) guideline recommendation to support diagnostic decision-making, especially when diagnosis is in question. This is based on the seminal paper by Maisel et al. and then confirmed in multiple other studies demonstrating both the diagnostic and prognostic value of NP testing. However, because diagnostic uncertainty and misdiagnosis remain prevalent in clinical practice, we recommend that NP levels be obtained routinely when AHF is suspected but not confirmable clinically.
Initial Classification
Once the diagnosis has been established, we recommend division of patients into clinical profiles or phenotypes. Several classification schema have been proposed. We advocate the use of the Gheorghiade/Braunwald initial assessment together with the SBP phenotyping suggested by Collins et al. (Figure 1). Of note, neither has been prospectively tested in regards to their impact on outcomes.
(Enlarge Image)
Figure 1.
Initial approach to acute heart failure management algorithm.
Modified and reproduced with permission from [24].
The six-axis model proposed by Gheorghiade and Braunwald highlight the essential considerations in early AHF management: clinical severity; initial blood pressure; heart rate and rhythm; precipitants; de novo versus chronic HF; and comorbidities. At first glance, the absence of EF may be seen unusual; however, at the present time, with the exception of cardiogenic shock or severe advanced HF (which is subsumed under the category of clinical severity), knowledge of the EF does not significantly impact initial management, which focuses on congestion relief and hemodynamic improvement. Knowledge of EF remains critical, however, for pre- and post-discharge management.
Clinical Severity
As shown in Table 1, stabilization and treatment of life-threatening conditions in AHF is the first priority. Assessment of the airway-breathing-circulation are paramount. Upto 5% of AHF patients will require mechanical ventilation. Overall, meta-analysis supports the benefits of noninvasive ventilation (NIV) in cardiogenic pulmonary edema, showing that early use can prevent the need for endotracheal intubation. However, in the Cardiogenic Pulmonary Oedema trial (3CPO), other than an improvement in dyspnea, no differences were seen in the rates of death or intubation between patients treated with facemask oxygen, continuous positive airway pressure only or bi-level positive airway pressure (i.e., expiratory and inspiratory). Although 3CPO was the largest randomized trial to investigate different methods of oxygen delivery in AHF patients, we continue to recommend the early use of NIV for appropriate patients who present with even moderate severity, given its potential to alleviate symptoms, its noninvasive nature and the intended short-term duration of use.
Initial Blood Pressure
Once stabilized, treatment of AHF patients is based on three primary BP phenotypes: hypertensive (SBP >140 mmHg), normotensive (90–140 mmHg) and hypotensive <90 mmHg). Note that these cutpoints are intended as guidelines and are not absolute thresholds.
For the hypertensive AHF patient (Figure 2), emphasis is on vasodilation rather than fluid removal, as these patients symptoms may be due more to volume re-distribution, as opposed to total volume overload. This concept of 'vascular' failure is best exemplified by the 'flash' pulmonary edema patient; sudden onset as opposed to days or even weeks of slowly progressive signs and symptoms. Although intravenous (iv.) loop diuretics are still recommended, we emphasize the use of vasodilators for initial treatment.
(Enlarge Image)
Figure 2.
Algorithm for the initial treatment of hypertensive (systolic blood pressure >140 mmHg) acute heart failure.
iv.: Intravenous; NIV: Noninvasive ventilation; SBP: Systolic blood pressure.Modified and reproduced with permission from [24].
Nitroglycerin, nitroprusside, nesiritide, hydralazine, angiotensin convering enzyme-inhibitor (ACE-I) and calcium channel blockers have all been used for initial vasodilation. Although experience is greatest with nitrates and multiple studies support their use, nesiritide is the only FDA-approved vasodilator that has been studied in a large, definitive randomized controlled trial. Only one other agent has recently completed Phase III studies with positive findings. In the RELAX-AHF trial, a Phase III study of serelaxin versus placebo, both in addition to standard of care in 1161 patients, the primary end point of dyspnea improvement was achieved. This improvement in dyspnea was observed only by visual analog scale, not by Likert. A statistically significant reduction in the secondary end point of 180-day cardiovascular (CV) mortality was also seen in the serelaxin group; however, there was no improvement in the composite of 60-day CV death or readmission for HF/renal failure. At present, serelaxin lacks FDA approval. Further studies are currently being planned.
Despite the absence of large-scale trials, nitroglycerin is the most widely used vasodilator. Due to its familiarity, rapid onset and clearance, and inexpensive cost, we advocate for its use as first-line therapy for hypertensive AHF (see Table 2 for doses). The sublingual route is recommended initially followed by iv. delivery for those who continue to have symptoms. High doses, much higher than traditionally used, have been shown to be both safe and effective. Cotter et al. demonstrated lower rates of intubation and myocardial infarction as well as improvements in heart rate, respiratory rate and oxygen saturation with high-dose nitrates compared with lower doses. Levy et al. showed similar benefits. Although we prefer the iv. route in the ED setting, sublingual and topical nitrate application are a valid alternative and have also been shown to be safe and effective at reducing NP levels more rapidly within 48 h outside of the ICU setting. Concerns regarding nitrate tolerance in the ED setting are unfounded as the goals of therapy are short term (i.e., dyspnea relief through afterload reduction), and there is no outcome benefit with protracted infusion. However, a small proportion of patients may be nitroglycerin unresponsive as noted by frequent up titration without clinical response. Nesiritide may also be considered first-line treatment, though we use it clinically as second- or third-line therapy. Initial concerns regarding its safety via meta-analysis have been laid to rest by prospective study (ASCEND-HF trial). Its efficacy, however, remains debated, as only a small, but statistically significant improvement in dyspnea was observed.
Other vasodilators such as nitroprusside, hydralazine, ACEI and dihydropyridine calcium channel blockers have also been used in the ED setting. Given the limited evidence base for traditional therapies, it is difficult to state that these other vasodilators are contraindicated in AHF. Despite the effectiveness of nitroprusside at vasodilation, its use in the ED is often limited by requirements for central line delivery and invasive blood pressure monitoring. Small studies suggest the potential efficacy of both ACEI and dihydropyridine calcium channel blockers. With ACEI, there is the potential risk of protracted, first-dose hypotension, which may lead to adverse events, either immediately or downstream. Although the absence of robust safety data does not rule out utilization of alternative vasodilatory therapies, at the present time, nitrates remain our preferential first-line treatment for this AHF phenotype. The normotensive patient (Figure 3) is the prototypical patient with known chronic HF and reduced EF who decompensates. The clinical presentation is usually less dramatic compared to hypertensive AHF, with historical features of progressive worsening signs and symptoms, including fatigue, weight gain, dyspnea on exertion and peripheral edema. Often, escalation of outpatient oral diuretic therapy has failed. For these patients, emphasis is on decongestion through fluid removal. In the ED setting, iv. loop diuretic therapy is the mainstay of fluid overload management.
(Enlarge Image)
Figure 3.
Algorithm for the initial treatment of normotensive acute heart failure (systolic blood pressure of 100–140 mmHg).
iv.: Intravenous; NIV: Noninvasive ventilation; SBP: Systolic blood pressure.
Modified and reproduced with permission from [24].
Despite their widespread use, the ideal dose and method of delivery (bolus versus continuous infusion) continues to be debated. Loop diuretics have been associated with transient worsening of pump function, brief initial elevation of filling pressures, worsening neurohormonal profile, electrolyte abnormalities, renal injury and by secondary analysis, worsening mortality. Importantly, however, there is no definitive proof that iv. loop diuretics given to AHF patients lead to worse outcomes.
The recent DOSE-AHF trial tested high- versus low-dose iv. loop diuretic as well as continuous versus bolus strategies in AHF patients. The study randomized 308 AHF patients within 24 h of presentation. They found that high-dose loop diuretic, defined as 2.5-times the total oral dose divided throughout the day led to greater diuresis and dyspnea improvement compared with the low-dose strategy, defined as the total outpatient oral diuretic dose. The study was not powered for mortality and no differences between groups were identified, but the high-dose group did have transient worsening of renal function. In addition, there were no significant differences between continuous versus bolus iv. diuretic treatment. Thus, we recommend high-dose iv. bolus loop diuretic therapy (i.e., 2.5-times the total outpatient oral loop diuretic dose, divided based on twice a day or 3-times a day dose scheduling) at the time of ED presentation.
It is worth highlighting that many AHF studies do not evaluate patients in the ED, but usually within 24 h or longer from presentation. Whether results may have differed or been affected depending on time of enrollment remains debated. For the hypotensive AHF patient (SBP <90 mmHg), Figure 4 outlines the initial approach. Importantly, this is an uncommon phenotype when considered in the context of the one million hospitalizations for AHF that occur every year. Less than 5% of patients present with SBP <90 mmHg, though the frequency of hypotensive AHF presentation varies depending on hospital type, with a higher reported frequency among some hospitals, especially those that perform cardiac transplantation. Patients in cardiogenic shock without a previous history of HF will present with a more severe clinical manifestations and often a life-threatening precipitant (i.e., acute MI) is the underlying cause. It is the patient with a history of advanced or end-stage HF, defined as persistent HF signs and symptoms despite maximal medical therapy, where careful assessment of volume status is critical. Many of these patients may have chronically low SBP due to the severity of their left ventricle (LV) systolic dysfunction and the effects of concurrent HF therapy. Although worsening HF as the cause of their acute presentation is likely, these patients may also be intravascularly depleted secondary to recent up-titration of outpatient diuretic therapy. Furthermore, as these patients often have multiple comorbidities, other conditions (i.e., infection) may be the primary precipitant with worsening HF a result.
(Enlarge Image)
Figure 4.
Algorithm for initial treatment of hypotensive acute heart failure (systolic blood pressure <100 mmHg).
HF: Heart failure; IABP: Intra-aortic balloon pump; iv.: Intravenous; SBP: Systolic blood pressure.Modified and reproduced with permission from [24].
Perhaps, the most critical point to emphasize in the management of hypotensive AHF is the purpose of treatment, which is to improve hypoperfusion, not simply raise blood pressure. In the past, inotropes were used to optimize the hemodynamic profile of AHF patients despite a clinical presentation inconsistent with shock or advanced HF. Although it is true that inotropes such as dobutamine and milrinone can help the AHF patient in the short term, they are associated with worsening morbidity and mortality. Thus, current guidelines recommend judicious use of inotropic therapy and only in those patients who truly require it. Although dobutamine and milrinone are the most common inotropes used for treatment of hypotensive AHF, they are not the only means of providing hemodynamic support for those who need it. At moderate doses (5–10 mcg/kg/min), dopamine does provide reasonable inotropic and chronotropic effects but at higher doses (>10–20 mcg/kg/min) vasoconstrictive activity predominates. Although the latter will help improve blood pressure, significant increases in afterload may be undesirable in some patients, particularly those with poor systolic function who cannot generate sufficient cardiac work to overcome added impedance to forward blood flow. Alternatives to increase myocardial contractility such as digoxin and levosimendan (available in Europe only) should be considered when patients are on chronic β-blocker therapy as this can diminish the effectiveness of adrenergic receptor agonists. Cardiac myosin activators, a promising new class of inotropes currently under study, work by directly stimulating the contractile apparatus. They may ultimately prove useful for this group of patients and hypotensive AHF in general, as they have none of the deleterious, mechanistic effects of existing agents. For the rare patient with refractory cardiogenic shock (most often ischemic in origin), aortic balloon counterpulsion or emergent intervention with a percutaneous left ventricular assist device may be needed.
Heart Rate & Rhythm
Sinus tachycardia should improve with treatment of AHF. Failure to improve suggests an underlying precipitant or etiology that has yet to be identified.
Atrial fibrillation is the most common dysrhythmia complicating AHF management. Similar to the lack of evidence to guide AHF management, there is a paucity of robust data to guide clinicians regarding AF with rapid ventricular response and AHF. Guidelines recommend the use of amiodarone and digoxin for AHF patients with reduced LV ejection function without an accessory pathway, though there is a risk of cardioversion with amiodarone, which should be carefully considered. Caution is also warranted regarding more traditionally used agents to control HR, such as nondihydropyridine calcium channel blockers and β-blockers, given their longer duration of action and negative inotropic effects. Despite these recommendations, anecdotally, nondihydropyridine calcium channel blockers are commonly used. If used, we recommend caution and to start with low doses.
AHF patients who present in shock or extremis secondary to AF require immediate cardioversion. For the vast majority of cases, management is similar to AF management principles without AHF: rate control; consideration of rhythm management; and stroke prevention. Of note, risk of stroke is even higher in HF patients.
It can be difficult to determine if AF is the cause or result of AHF. For patients with de novo HF and AF with rapid ventricular response, AF is the more likely culprit. However, for the majority of patients who have chronic HF and chronic AF, rapid ventricular response often results from HF decompensation. Treatment of AHF will often decrease the HR. It is also important to note that in the setting of AHF, increased HR (even if in AF) at rates of 90–110 b.p.m. may be a compensatory response to increase cardiac output since stroke volume is reduced and can be 'fixed' (i.e., inability to increase stroke volume in response to stress) in AHF patients. Knowledge of LV ejection fraction can help guide initial therapeutic management of AF and AHF. Patients with AHF and a preserved EF are particularly vulnerable to tachycardias due to decreased filling time in conjunction with a noncompliant ventricle. Negative inotropic effects of β-blockers or nondihydropyridine calcium channel blockers are less of a concern; as such either class of drug can be useful for rate control in the patient with AHF, preserved EF and tachycardia. Of note, in more advanced stages of HFpEF and in those with restrictive cardiomyopathy, the LV may become so stiff that reducing HR below a certain threshold may result in a reduction in cardiac output, as stroke volume is reduced and fixed. In patients with reduced LV ejection fraction, the greater the severity of LV dysfunction, the greater the caution in using agents other than digoxin and amiodarone. Treatment of AHF may improve the HR substantially and thus rate control may appropriately be a secondary goal.
Precipitants
Concurrent with initial treatment, identification of the precipitant of HF decompensation is a primary goal of early management. In other words, why did the patient decompensate? Dietary indiscretion, medication noncompliance, pneumonia/respiratory process, ischemia and arrhythmias are common precipitants. Determining the precipitant targets, the cause of decompensation and not just the outward manifestations. Although an exhaustive search may not be feasible in the ED, timely identification of precipitants such as infection, ischemia, arrhythmia or pulmonary embolism may prevent or attenuate in-hospital morbidity.
De Novo HF Versus Acute Decompensation of Chronic HF
After initial stabilization and management, disposition decisions must be made. As a general rule, patients who present with HF for the first time should be hospitalized. These individuals have not undergone previous work-up for potential etiology and/or precipitant and likely have not had recent evaluation of cardiac structure and function. Although it is possible that initiation of evidence-based therapies could begin as an outpatient, hospitalization allows for more efficient and in-depth evaluation and education. Culprit pathology amenable to rapid intervention can also occur during hospitalization. Decompensated chronic HF patients who present with AHF have, presumably, already undergone in-depth interrogation for potential reversible causes as well as optimization of all available evidence-based therapies. Prior in part driven by how recent will impact hospital management. Figure 5 below highlights potential targets or risk features to be considered during or soon after hospitalization.
(Enlarge Image)
Figure 5.
Comprehensive assessment and cardiac reconstruction.
Viable but dysfunctional myocardium.
Select patients.
Investigational agents.
ACE-I: Angiotensin converting enzyme inhibitor; AF: Atrial fibrillation; AHF: Acute heart failure; ARB: Angiotensin receptor blocker; CABG:
Coronary artery bypass grafting; CAD: Coronary artery disease; CRT: Chronic resynchronization therapy; Hydral: Hydralazine; ICD: Implantable cardiac defibrillator; ISDN: Isosorbide dinitrate; JVP: Jugular venous pulse; LV: Left ventricle.
Modified and reproduced with permission from [22].
Comorbidities
As previously mentioned, both CV and non-CV conditions are prevalent in AHF patients. Consideration of comorbidities as a potential precipitant is therefore important and may impact treatment and prevention of readmission. In the ED, unless the principal precipitant (e.g., worsening COPD contributing to an acute decompensation of HF) can be diagnosed and managed acutely, management of comorbidities remains largely a focus of in-hospital or post-discharge management.
Risk Stratification & Disposition
In the USA, EDs see over 136,000,000 patients annually. As such, patient throughput is an additional consideration when treating all types of patient, to avoid delays in the care of patients waiting to be seen. Thus, efficient risk stratification and disposition of patients are critical.
In terms of AHF, over 80% of patients evaluated in the ED are subsequently admitted to the hospital, yet past studies have suggested that up to 50% could either be discharged or held in observation status. Multiple markers of risk have been identified that could potentially aid in disposition decision making (Table 3).
At the present time, for most of these prognostic markers, little can be done in the ED to alter the trajectory of these patients. Whether treatment to alter these markers of risk improves outcomes remains an area of active, ongoing research. Although concern for litigation is understandable, especially given the high post-discharge morbidity and mortality associated with AHF, we would argue that this should not be the principal reason behind decision making in the ED. Rather, it is the lack of robust risk stratification instruments combined with a population that is older with multiple comorbidities that drives the high admission and re-admission rate. Risk stratification traditionally focuses on the high-risk patient, assuming that intervention will lower risk. However, the opposite approach is needed in the ED. Who is low risk? There is often little debate regarding the admission of AHF patients with high-risk features, especially when changed from baseline. Rather, it is the patient without high-risk features that often generates debate.
Absence of high risk does not equal low risk. Even the lowest risk group stratified by SBP had a 90-day mortality rate of 5.4% and re-hospitalization rate of 27.6%. Risk stratification instruments have largely been performed on hospitalized patients, limiting their applicability to the ED setting where the impact of hospitalization will not be realized.
Although some risk stratification tools have been created for ED purposes, they either have not been well validated or are cumbersome to use. In addition, outcomes of discharged AHF patients have not been well studied, highlighting our limited understanding of this population. Analysis from a large Canadian administrative database, however, suggests that clinician impression is a poor discriminator, as discharged patients had worse outcomes than those admitted. Although there is insufficient evidence to support strong recommendations regarding ED disposition decision making, there is little evidence to support current strategies focused on reducing 30-day re-admission. Thus, we recommend hospitals or health care systems develop structures and processes to determine various levels of risk to aid in disposition based on local expert consensus, emphasizing safety. A draft proposal of a potential algorithm is proposed below (Figure 6). Importantly, this algorithm has not been validated and requires further study prior to use. It does, however, provide a starting point for discussion. As mentioned, the evidence supporting current plans to decrease 30-day re-admissions are also lacking in evidence.
(Enlarge Image)
Figure 6.
Proposed algorithm for risk stratification in the emergency room.
BNP: B-type natriuretic peptide; BUN: Blood urea nitrogen; DM: Diabetes mellitus; ED: Emergency department; ICU: Intensive care unit; OU: Observation unit; SBP: Systolic blood pressure.
Modified and reproduced with permission from [76].
Source...