Preventing Readmission in Heart Failure

With reference to recent studies, the authors discuss making the most of current diagnostics and therapeutics to help patients with heart failure fare better and reduce growing cost pressures from a disease of which recidivism is a hallmark.

By W. F. Peacock IV, MD, and Charles L. Emerman, MD

Heart failure is a health care problem of epidemic proportions in the United States. Almost 5 million Americans are living with heart failure, and nearly 500,000 new cases are diagnosed annually. Roughly 75% of men and 85% of women hospitalized with heart failure are more than 65 years of age. Despite treatment advances, heart failure has a significant mortality rate. The number of deaths from heart failure has increased by nearly 150% since 1979. Nearly 80% of men and 70% of women under age 65 die within eight years of diagnosis. After a single heart failure hospitalization, the four-year mortality rate is 61%; the in-hospital mortality rate for acute pulmonary edema is 17%. Recidivism is a hallmark of heart failure. The overall six-month readmission rates are as high as 30%.

Because heart failure is the number one national health care expense, it has a significant impact on hospital economics. Many hospitals are unable to effectively manage length-of-stay and revisit issues, which explains why the average U.S. hospital loses about $1300 for each patient admitted with heart failure. Effective therapies that avoid admissions, decrease length of stay, lower hospital costs, and prevent 30-day readmissions can be used to control financial pressures on hospital systems.
 

CLINICAL SPECTRUM OF ILLNESS

Heart failure is a syndrome that represents a clinical spectrum of illness (see table). Class A patients have no clinical evidence of disease but do have risk factors for overt illness. Since early treatment can delay the onset of symptoms, identifying and treating these patients ensures the best outcome. Class B patients are at the lowest stages of clinically apparent disease. With class B and higher, there is increasing dysfunction and disability.

The New York Heart Association (NYHA) system is used to classify symptomatic patients. Patients who are in the lower categories, NYHA class I or II, may have some symptoms of exercise intolerance but do not have significant pulmonary congestion. Class III patients have dyspnea with minimal exertion. At the highest end of the spectrum, NYHA class IV, patients have dyspnea at rest, and may show signs of poor perfusion related to low cardiac output.

Although the term "congestive heart failure" is commonly used, not all heart failure patients have fluid overload. Compliant end-stage heart failure patients may present without congestion, in a euvolemic state, but with complaints of weakness and fatigue. Heart failure is the disease; congestion is a clinical state. Congestion describes a fluid-overloaded condition in a patient with underlying heart failure.

Systolic heart failure occurs with a depressed ejection fraction of less than 40%; it can be defined as an inability to eject a normal volume of blood from the ventricle. Diastolic heart failure occurs with a preserved ejection fraction of more than 40%. However, because of decreased ventricular compliance, an abnormal pressure-volume relationship develops on filling. Patients with heart failure and preserved systolic function, which used to be referred to as diastolic dysfunction, generally are unable to accept a normal volume of blood in the ventricle. While these distinctions are important in the euvolemic patient, they are of little consequence in the congested patient because treatment is similar regardless of the underlying pathophysiologic condition.

The patient's vital signs, mental status, and skin perfusion can help guide therapy. It is important that the assessment of blood pressure include the determination of whether or not the patient is symptomatic. Heart failure patients are best served by keeping their blood pressure as low as possible, as long as they can mentate, ambulate, and urinate normally. In symptomatic hypotension, inotropic support or fluid therapy may be indicated.
 

PATHOPHYSIOLOGY AND DIAGNOSIS OF HEART FAILURE

Heart failure usually results from a direct myocardial injury, hemodynamic stress, or arrhythmia. Systemic diseases such as hyperthyroidism or sepsis can also result in abnormal cardiac workload requirements that lead to heart failure. Most patients with heart failure have hypertension or coronary artery disease, although many cases are idiopathic.

After an initial myocardial injury and cardiac output reduction, endogenous neurohormonal pathways are stimulated to maintain circulatory hemodynamics and systemic blood pressure. This activates the renin-aldosterone-angiotensin and adrenergic systems. This is counterbalanced by the natriuretic peptic system. B-type natriuretic peptide (BNP) and atrial natriuretic peptide are secreted from the ventricles and atria, respectively, in response to a stretch or volume stimulus. This has the beneficial effect of vasodilation, diuresis, and a decrease in angiotensin and norepinephrine release.

Because heart failure is a disease of the elderly, who have frequent comorbidities and confounding conditions, it is sometimes difficult to distinguish it from other causes of dyspnea. This leads to the reported overall misdiagnosis rate in emergency departments of 12%. Even when physicians are confident in the diagnosis, rating the probability of heart failure as high, their clinical accuracy is only 74%.

The physical examination has limited sensitivity for detecting heart failure. The most specific predictor is auscultation of a third heart sound or the presence of jugular venous distension. Although these signs have a specificity in the 90% range for predicting heart failure, their sensitivity is in the 20% range.

An ECG is neither sensitive nor specific in the diagnosis of heart failure. However, in the acute setting of possible decompensated heart failure, an ECG can exclude ST-segment elevation acute myocardial infarction (MI). Non-ST segment elevation MI must still be considered. In a retrospective review of 151 decompensated heart failure patients in the emergency department, 21 (14%) had elevated cardiac markers. Chest pain occurred in only 30% of these patients, and the ECG was diagnostic for acute cardiac ischemia in fewer than 1%.

Additionally, the ECG can suggest underlying structural heart disease such as left ventricular hypertrophy, atrial enlargement, and prior MI. It can also detect arrhythmias, electrolyte abnormality (such as hyperkalemia), and drug toxicity (from digoxin, for example), and can help determine the long-term prognosis. Abnormal Q waves, a QRS duration of more than 0.12 ms, or left bundle branch block predicts an increased five-year mortality in heart failure.

Chest x-rays can be helpful in suspected heart failure. Vascular cephalization is an independent predictor of heart failure, but its absence does not exclude an elevated pulmonary capillary wedge pressure (PCWP). Even when PCWP is markedly elevated, pulmonary congestion may be absent in 39% of patients. Also, radiographic changes can lag behind acute clinical changes by as much as six hours.

back to top

ELECTROLYTE AND BNP LEVELS

Measurement of serum electrolyte levels may be useful, particularly if patients have been on diuretics and are going to be receiving repeated doses of diuretics in the hospital. About 5% to 10% of patients with decompensated heart failure will have elevated cardiac markers, which may indicate ischemia as a precipitant of decompensated heart failure. Patients with heart failure, however, may have elevated serum troponin levels in the absence of ischemia.

Levels of BNP correlate with left ventricular dysfunction, the severity of the heart failure, and elevated ventricular pressures. These levels can help differentiate between heart failure and other causes of dyspnea. Using the BNP assay can reduce the misdiagnosis rate.

The Breathing Not Properly Multinational Study involved 1,586 patients presenting to the emergency department with acute dyspnea. A BNP cutoff value of 100 pg/ml was the strongest predictor of heart failure and was superior to clinical evaluation. Patients with levels below 100 pg/ml are unlikely to have heart failure, although there are occasional exceptions. Levels above 100 pg/ml indicate ventricular stress or strain, which may occur from heart failure. Severe chronic obstructive pulmonary disease or large pulmonary emboli would be other diagnostic considerations. Not everyone with dyspnea needs a BNP measurement, but it can be useful when there is some uncertainty about the diagnosis or the patient has complicating medical conditions. The BNP level can be elevated in other conditions that affect ventricular stress, such as acute MI (see table).

TREATMENT STRATEGIES

The goals of therapy for a patient with decompensated heart failure are to improve respiratory and circulatory status. Patients presenting with acute decompensation range from those with sudden onset of "flash" pulmonary edema to those who have gradual onset of decompensation with fluid retention. Fortunately, most patients fall into this latter category.

Management of the patient in cardiogenic shock or with marked pulmonary edema requires prompt evaluation and initiation of therapy. Treatment begins with the administration of supplemental oxygen, which can be accomplished by nasal cannula or nonrebreather mask in most cases, using pulse oximetry to guide therapy. The patient with concomitant severe chronic obstructive pulmonary disease may require controlled administration of oxygen using a Venturi mask because uncontrolled administration can lead to hypercarbia and respiratory acidosis.

Some patients may arrive in such extreme condition that they require immediate intubation. If possible, these patients should be preoxygenated while preparations are made for intubation. Rapid-sequence techniques may minimize airway trauma and other complications of intubation. Various techniques are available but generally involve administration of intravenous (IV) lidocaine to provide some airway anesthesia, sedation with benzodiazepines or short-acting barbiturates, followed by administration of paralytic agents. Once intubated, patients with pulmonary edema may benefit from small amounts of positive end-expiratory pressure, with monitoring for hypotension.

Patients who are alert and cooperative may benefit from noninvasive ventilation rather than intubation. Two techniques have been employed for patients with pulmonary edema—continuous positive airway pressure or bilevel positive airway pressure. Both can reduce ventricular pressure, but they require dedicated attention from a nurse, respiratory therapist, or physician during the initial phase of management. Intubation can be avoided in 75% of patients managed with noninvasive ventilation. These patients have a lower incidence of respiratory complications and shorter stays in the intensive care unit.

Cardiogenic shock can occur with right ventricular infarction. Patients who are not in fluid pulmonary edema may benefit from a trial of a small fluid challenge of 250 ml or so. Patients who do not respond to this fluid challenge or who develop pulmonary edema as a result of it may require inotropic agents. In cardiogenic shock, dopamine has the advantage of being both a vasopressor and an inotrope. Typically, the drug is started at 5 mcg/kg/minute and titrated to a level that improves mentation and provides adequate hemodynamic response.

Milrinone, another inotropic agent, also has vasodilator properties. The blood pressure response with this drug may be variable, although milrinone may be useful in patients who have both depressed cardiac output and a high systemic vascular resistance. Milrinone may also be advantageous in patients on chronic beta-blocker therapy. Some patients may benefit from combination therapy with dopamine, used for its vasopressor effect, and dobutamine, which has primarily an inotropic effect. Management of these patients can be complex; early consideration of invasive monitoring with consultation by an intensivist may be useful.

Most heart failure patients will not present in cardiogenic shock or in such severe respiratory distress that they require immediate mechanical ventilation. These patients should respond to attempts to decrease PCWP, allowing for decongestion with improved respiratory status. Various agents are available that can help in managing the patient with decompensated heart failure. Newer agents such as nesiritide offer significant advantages over previous pharmacologic therapy, while older therapies such as milrinone have been shown recently not to be helpful in most circumstances.

DIURETIC THERAPY

Many physicians have considered diuretics to be the first line of therapy for patients with decompensated heart failure. Typically, furosemide is initiated early. The dosing is controversial, although many clinicians start with intravenous administration of the total outpatient daily dose for patients who are already on furosemide. Patients who have not been on diuretics in the past typically receive 20 to 40 mg. Those with renal insufficiency may require a higher dose of a diuretic.

Patients who are receiving diuretics should be monitored for adequate response; a brisk diuresis should be expected within 30 to 45 minutes. Those with inadequate diuresis may require repeat boluses with higher doses. Other diuretics that may be useful include bumetanide, typically given in doses of 0.5 to 1 mg intravenously. Torsemide is generally initiated at a dose of 10 to 20 mg intravenously, with doses increased up to 200 mg/day if there is an inadequate response. Patients who do not respond to loop diuretics may benefit from a thiazide diuretic, such as metolazone at a dose of 5 to 20 mg orally.

The disadvantage of aggressive use of diuretics is that they activate the neurohormonal cascade, which leads to increased release of norepinephrine and angiotensin. In addition, aggressive diuresis can lead to electrolyte abnormalities, such as hypokalemia and hypomagnesemia. The effects of diuretics may be potentiated with concomitant use of other agents, such as nesiritide, in which case high doses may not be necessary.

Sublingual or topical nitroglycerin may be useful in the prehospital management of patients with decompensated heart failure or in the initial management of patients in the emergency department. Sublingual nitroglycerin is relatively fast-acting. Topical nitroglycerin also has a quick onset of action and an antianginal effect, and it can be quickly removed if the patient develops significant hypotension.

back to top

USE OF NESIRITIDE IN HEART FAILURE

Nesiritide was approved two years ago for use in decompensated heart failure. This drug belongs to a class of agents generally referred to as natriuretic peptides and has a number of actions that are beneficial in heart failure. Studies have demonstrated that nesiritide increases diuresis, lowers PCWP, decreases angiotensin and neuroepinephrine levels, lowers blood pressure, and improves dyspnea.

Nesiritide has been studied with other agents in several studies. The PRECEDENT trial compared two doses of nesiritide against dobutamine and found that the patients treated with dobutamine had a higher incidence of ventricular ectopy and cardiac arrest. Six-month survival was lower for patients receiving dobutamine than for those receiving nesiritide.

The VMAC trial compared nesiritide to nitroglycerin in patients monitored with or without a Swan-Ganz catheter. It found that nesiritide led to rapid and sustained decreases in PCWP and early improvement in dyspnea. Nitroglycerin, on the other hand, resulted in a delayed decrease in PCWP that occurred on average about two hours after the infusion was started. Additionally, the VMAC trial demonstrated that patients rapidly developed a tolerance to nitroglycerin with increases in PCWP that called for higher doses of the drug.

Nesiritide is started at a dose of 2 mcg/kg, based on actual body weight, administered as a bolus, followed by an infusion at a rate of 0.01 mcg/kg/minute. The dose can be increased every three hours by administering an initial bolus of 1 mcg/kg, followed by increases in the infusion rate of 0.005 mcg/kg/minute. Symptomatic hypotension occurs in a small number of patients given nesiritide. These patients may be managed by observation or by decreasing the dose rate. Occasionally, patients with symptomatic hypotension may require small fluid boluses. It is common for patients to have systolic blood pressures in the 90 to 100 mm Hg range after nesiritide is started. Blood pressures in this range are readily tolerated by most patients. As is the case with nitroglycerin, nesiritide has been shown in clinical and animal studies to have coronary vasodilator effects. Its use in patients with acute coronary syndrome has not been well studied, although preliminary data indicate that it is safe.

Intravenous nitroglycerin has been used in the management of congestive heart failure with little clinical evidence to guide its use. Patients are frequently started at a relatively low dose, such as 10 to 20 mcg/minute, with rapid titration based on blood pressure measurements, unless a Swan-Ganz catheter has been inserted. Typically, high doses of nitroglycerin are required, usually in the range of 120 mcg/minute. As demonstrated in the VMAC trial, patients will rapidly develop tolerance to nitroglycerin and clinicians should be prepared to continue to increase the dose rate.

Nitroprusside is infrequently used in the management of congestive heart failure. It is a very potent arterial vasodilator and its use, particularly in patients with renal insufficiency, may be complicated by thiocyanate toxicity.
 

INOTROPIC AGENTS

Inotropic agents are also used in the management of decompensated pulmonary edema, aside from their use in cardiogenic shock, but without much evidence to support such use. Dopamine, as noted above, has inotropic properties and, at low doses, is thought to dilate renal, cardiac, and splenic vessels. The notion that dopamine can be used at so-called renal doses to enhance diuresis has recently become controversial. The use of dopamine is complicated by the occurrence of cardiac arrhythmias, peripheral vasoconstriction, and tachycardia, leading to increased myocardial oxygen demand and possibly ischemia.

Dobutamine is used somewhat more commonly in managing decompensated heart failure. It increases cardiac output and leads to peripheral vasodilation. Generally, dobutamine will reduce PCWP without having much effect on systemic blood pressure. Typically, it is started at low doses in the range of 1 mcg/kg/minute and then titrated up, based on PCWP for those patients who have a Swan-Ganz catheter or a clinical assessment of PCWP in patients without invasive monitoring. The PRECEDENT trial, however, demonstrated that the use of dobutamine leads to increased ventricular ectopy with occasional ventricular tachycardia and cardiac arrest. Use of dobutamine has been shown to increase mortality at six months compared with nesiritide.

Milrinone has chronotropic, inotropic, and vasodilator effects similar to those of aminophylline. Because it does not act on beta receptors, milrinone may be useful as an inotropic agent in patients who are already on beta-blocker therapy. A recent trial has called into question the routine use of milrinone for patients with decompensated heart failure. The OPTIME study evaluated the addition of milrinone to otherwise standard therapy in patients with decompensated heart failure. Those patients receiving milrinone had a higher adverse event rate compared to those receiving standard therapy. There was no advantage to the addition of milrinone in decreasing patients' length of stay in the hospital, and there was a trend toward an increased mortality rate in patients receiving milrinone. When milrinone is used, it is typically given in a dose of 50 mcg/kg as a bolus over 10 minutes, followed by an infusion of 0.375 to 0.75 mcg/kg/minute. The dose should be reduced in patients with chronic renal insufficiency. The patient should be observed carefully for hypotension during administration of the loading dose.

Morphine has been used in the management of pulmonary edema for years because of its venodilating properties. However, morphine is a weak venodilator, and more effective agents should be used. Typically, morphine is administered in 2- to 5-mg boluses, with titration based on the patient's hemodynamic and mental status. Some studies have shown that morphine use is associated with a higher incidence of intubation. Given that there are more effective vasodilating agents, the use of morphine has fallen out of favor recently.

Angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor-blocking agents have a clear role in the long-term management of heart failure. The use of these agents has been shown to improve long-term survival. Their use in acute decompensated heart failure is not as well established. Studies done years ago showed that sublingual captopril given in repeated doses can lower PCWP and improve hemodynamic status. It is reasonable to start one of these agents during hospitalization in order to monitor for effect. There does not seem to be a compelling reason to start these agents in the emergency department, provided that the patient's blood pressure can be controlled by other means.

Beta blockers have also been shown to improve long-term survival in patients being treated for heart failure. The combined alpha- and beta-blocker agent carvedilol has demonstrated improved survival in the most severe heart failure patients. Generally, however, these agents are not initiated in the emergency department in patients with fluid overload.

The role of digoxin has been controversial for a number of years. It may be useful in patients with atrial fibrillation and a fast ventricular response. Beta blockers, however, may be more effective in slowing heart rate than digoxin. The long-term use of digoxin in patients with heart failure has been shown to decrease hospitalization rates, but there is little evidence that immediate initiation of digoxin will help the patient being hospitalized for heart failure.
 

RECOMMENDED APPROACH

For the majority of patients, therapy is initiated with oxygen, diuretics, and either sublingual or topical nitrates. There have been two further approaches proposed: waiting for the results of diuresis or initiating vasodilator therapy for those patients who are most likely to benefit. The patients most likely to benefit from vasodilator therapy are those with an elevated systolic blood pressure, evidence of fluid overload, chronic renal insufficiency, and systolic dysfunction. Our approach to these patients has been to start nesiritide early in the course of therapy, at the bolus dose and infusion rate noted earlier. The advantage of using nesiritide over nitroglycerin is that it has a more rapid onset of action, does not require frequent titration, does not lead to the development of tolerance, and does not require intensive unit monitoring. There is no safety data to guide the use of simultaneous IV nitroglycerin and nesiritide. It is acceptable to use topical, sublingual, or oral nitrates in conjunction with nesiritide.

We generally withhold ACE inhibitors for a few hours after starting nesiritide in order to observe its effects on blood pressure. As long as patients are not symptomatic from the drop in blood pressure, nesiritide is maintained at the initial dose. Patients who have symptomatic hypotension have either a dose reduction, dose interruption, or, rarely, a small fluid bolus.

Once patients have met therapeutic goals, including improvement in dyspnea, adequate diuresis, and weight loss, the nesiritide is discontinued. Typically, admitted patients are on nesiritide for one to two days prior to discontinuation; observation unit patients are treated for a shorter period of time. Nesiritide does not require weaning on discontinuation because of its 20-minute half-life.

At discharge, patients should have an early referral for reevaluation. Those patients not taking ACE inhibitors or beta blockers should be considered for initiation of such therapy. Many of these patients have risk factors for or have known coronary artery disease, and consideration for low-dose daily aspirin therapy and statin therapy may be warranted. Smoking cessation counseling is a JCAHO quality indicator and should be documented once it is performed.

back to top

Printable version of this article

E-mail this article