TY - JOUR
AU - Ayas, Najib T.
AB - ContextDyspnea is a common complaint in the emergency department where physicians must accurately make a rapid diagnosis.ObjectiveTo assess the usefulness of history, symptoms, and signs along with routine diagnostic studies (chest radiograph, electrocardiogram, and serum B-type natriuretic peptide [BNP]) that differentiate heart failure from other causes of dyspnea in the emergency department.Data SourcesWe searched MEDLINE (1966-July 2005) and the reference lists from retrieved articles, previous reviews, and physical examination textbooks.Study SelectionWe retained 22 studies of various findings for diagnosing heart failure in adult patients presenting with dyspnea to the emergency department.Data ExtractionTwo authors independently abstracted data (sensitivity, specificity, and likelihood ratios [LRs]) and assessed methodological quality.Data SynthesisMany features increased the probability of heart failure, with the best feature for each category being the presence of (1) past history of heart failure (positive LR = 5.8; 95% confidence interval [CI], 4.1-8.0); (2) the symptom of paroxysmal nocturnal dyspnea (positive LR = 2.6; 95% CI, 1.5-4.5); (3) the sign of the third heart sound (S3) gallop (positive LR = 11; 95% CI, 4.9-25.0); (4) the chest radiograph showing pulmonary venous congestion (positive LR = 12.0; 95% CI, 6.8-21.0); and (5) electrocardiogram showing atrial fibrillation (positive LR = 3.8; 95% CI, 1.7-8.8). The features that best decreased the probability of heart failure were the absence of (1) past history of heart failure (negative LR = 0.45; 95% CI, 0.38-0.53); (2) the symptom of dyspnea on exertion (negative LR = 0.48; 95% CI, 0.35-0.67); (3) rales (negative LR = 0.51; 95% CI, 0.37-0.70); (4) the chest radiograph showing cardiomegaly (negative LR = 0.33; 95% CI, 0.23-0.48); and (5) any electrocardiogram abnormality (negative LR = 0.64; 95% CI, 0.47-0.88). A low serum BNP proved to be the most useful test (serum B-type natriuretic peptide <100 pg/mL; negative LR = 0.11; 95% CI, 0.07-0.16).ConclusionsFor dyspneic adult emergency department patients, a directed history, physical examination, chest radiograph, and electrocardiography should be performed. If the suspicion of heart failure remains, obtaining a serum BNP level may be helpful, especially for excluding heart failure.CLINICAL SCENARIOSCase 1A 70-year-old woman with a history of a previous myocardial infarction and heart failure presents to the emergency department (ED) with a 2-day history of dyspnea at rest, orthopnea, and paroxysmal nocturnal dyspnea. Physical examination reveals an elevated jugular venous pressure, a third heart sound (ventricular filling gallop), bibasilar rales and wheezing, and bilateral lower extremity edema. The chest radiograph reveals cardiomegaly. An electrocardiogram (ECG) shows atrial fibrillation.Case 2A 65-year-old previously healthy man with a 30 pack-year smoking history presents to the ED with a 3-week history of dyspnea on exertion and at rest, associated with productive cough and sputum. Physical examination reveals bilateral rales and wheezing. The chest radiograph reveals pulmonary venous congestion and a pattern of interstitial edema. An ECG shows lateral ST-segment depression.Case 3A 60-year-old man with a history of chronic obstructive pulmonary disease (COPD) and previous myocardial infarction presents to the ED with a 2-week history of worsening dyspnea on exertion and cough. Physical examination reveals an elevated jugular venous pressure, bilateral wheezing, and bilateral lower extremity edema. The chest radiograph shows normal results. An ECG shows Q waves inferiorly.WHY IS THIS QUESTION IMPORTANT?Heart failure is a major public health concern. A heart failure epidemic affects more than 15 million people in North America and Europe, and an additional 1.5 million new cases are diagnosed every year.It is the most costly cardiovascular disorder in western countries, accounting for an estimated total direct annual expenditure of more than $24 billion in the United States in 2001.Failure to diagnose heart failure increases mortality, delays hospital discharge, and increases treatment costs.Dyspnea, an uncomfortable sensation of breathingor an awareness of respiratory distress,is the cause for more than 2.5 million clinician visits per year in the United States.A number of disorders cause dyspnea including congestive heart failure, COPD, asthma, deconditioning, metabolic acidosis, anxiety, upper airway obstruction, and neuromuscular weakness. Identifying patients with heart failure among the other causes allows early institution of appropriate symptomatic and evidence-based therapies.It is not always possible (nor feasible) to promptly evaluate every patient with dyspnea with tests of cardiac function (echocardiography, nuclear scans, or cardiac catheterization). This challenges physicians who must identify heart failure based on history, physical examination, and rapidly available investigations (eg, chest radiograph, electrocardiogram ECG, B-type natriuretic peptide [BNP]). Therefore, the purpose of this review was to identify the most useful symptoms, signs, and tests in diagnosing the clinical syndrome of heart failure in dyspneic patients presenting to the ED. By the syndrome of heart failure, we mean an overall clinical diagnosis of heart failure as the cause of dyspnea (irrespective of etiology, systolic, or diastolic dysfunction), using information from many sources including history, physical examination, chest radiograph, ECG, serum chemistries, and one or more confirmatory tests of cardiac function.Pathophysiology of Dyspnea in Heart FailureMultiple pathophysiological mechanisms have been hypothesized to modulate the sensation of dyspnea in patients with symptomatic heart failure (Table 1).Table 1.Physiological Categories and Mechanisms Causing Dyspnea in Heart Failure*CategoryMechanisms†Increased respiratory driveIncreased left ventricular end diastolic pressure → pulmonary venous congestion → stimulation of pulmonary J receptors (transmitted by vagal afferents to brain)Pulmonary venous congestion → ventilation/perfusion mismatch, shunt‡ → hypoxemia → stimulation of central and peripheral chemoreceptorsIncreased work of breathingPulmonary venous congestion → reduced lung compliance → increased airways resistance → increased elastic and resistive work of breathing → mismatch between afferent information from upper airway, lower airway, chest wall mechanoreceptors, and efferent signals to respiratory musclesWeakness of respiratory pump musclesActivation of catabolic factors → myopathy (structural, biochemical, functional abnormalities of skeletal respiratory muscles) → reduced respiratory muscle efficiency and endurance → mismatch between afferent mechanoreceptors and efferent signals to respiratory musclesPsychologicalAnxiety, depression → altered central perception*Adapted from Murray and Nadel Textbook of Respiratory Medicine,Harrison’s Principles of Internal Medicine,American Thoracic Society Dyspnea Consensus Statement,Heart Disease,and Manning and Schwartzstein.†Arrows denote one mechanism leading to another.‡Occurs when blood moves through the lung without coming into contact with oxygenated air.A previous review in The Rational Clinical Examination series assessed the usefulness of the clinical examination in predicting decreased left ventricular ejection fraction or increased filling pressure.Our current review extends the previous report by focusing on the prediction of the clinical syndrome of heart failure in dyspneic patients. This clinical focus is useful because not every patient with left ventricular dysfunction and/or high filling pressures on objective cardiac testing will be subjectively dyspneic; furthermore, patients with a reduced ejection fraction may be dyspneic from causes other than heart failure.Therefore, the use of the syndrome of heart failure takes into account a patient’s subjective sensation and findings on routine investigations, in addition to objective cardiac testing. One previous literature review has reported on the use of the clinical examination for discriminating causes of dyspnea; however, it was not restricted specifically to the syndrome of heart failure, and summary measures of sensitivity, specificity, and likelihood ratios (LRs) were not reported.We included serum BNP testing in this review because recent evidence suggests that it is useful in diagnosing heart failure.BNP is a neurohormone that is secreted almost exclusively from the ventricles in response to pressure and volume overload that produces natriuresis, diuresis, and smooth muscle relaxation.There is also emerging evidence that BNP is useful in prognosticating cardiovascular mortality in both acute and chronic heart failure.Studies are currently ongoing regarding the use of serial BNP levels as an indicator of treatment response and for titrating therapy.How to Elicit Symptoms and SignsAppropriate history taking and physical examination of the cardiopulmonary system has been described in detail in previous The Rational Clinical Examination articles,with the exception of the Valsalva maneuver. The Valsalva maneuver is performed by inflating and locking a blood pressure cuff to 15 mm Hg above the resting supine systolic pressure (Korotkoff sounds should not be audible at this point), at which point the patient performs a sustained Valsalva (exhalation against a closed glottis) for at least 10 seconds. In a normal response, systolic blood pressure immediately rises 30 to 40 mm Hg above baseline for 1 to 3 seconds (phase 1, appearance of Korotkoff sounds). As venous return decreases, systolic blood pressure drops sharply below baseline (phase 2, disappearance of Korotkoff sounds). When the Valsalva is released, there is a further drop of systolic blood pressure below baseline (phase 3, continued absence of Korotkoff sounds). Between 3 to 15 seconds after release, systolic blood pressure rises 15 mm Hg or more above the baseline level (phase 4, reappearance of Korotkoff sounds).Two abnormal responses have been described in heart failure. In the absent overshoot response, phases 1 to 3 are normal, but Korotkoff sounds do not reappear in phase 4. In the squarewave response, phase 1 is normal, but Korotkoff sounds are present in phases 2 and 3, followed by disappearance in phase 4.METHODSSearch StrategyWe conducted a computerized search of MEDLINE from 1966 to July 2005 concerning the precision and diagnostic accuracy of components of the clinical examination and simple investigations in diagnosing patients with dyspnea. Our strategy was deliberately broad to minimize the possibility of overlooking relevant articles. Multiple searches were performed with the first search using a similar strategy developed for The Rational Clinical Examination series.This strategy combined 4 exploded Medical Subject Headings (physical examination, medical history taking, professional competence, routine diagnostic tests) with 8 keyword categories (physical exam, medical history taking, professional competence, sensitivity and specificity, reproducibility of results, observer variation, decision support techniques, Bayes theorem) and 1 textword category (sensitivityand specificity); and intersected with 1 exploded Medical Subject Heading (dyspnea). The search was limited to studies published in English about humans. Further MEDLINE searches were conducted combining the following Medical Subject Headings textword and keyword searches: brain natriuretic peptide, natriuretic peptide, BNP, Valsalva, hepatojugular, abdominojugular, and breathlessness. These were intersected with the exploded medical subject heading dyspneaand the textword dyspnoea.The computerized search was supplemented with a manual search of reference lists of retrieved studies, review articles, and standard physical examination textbooks to identify additional articles not captured through the computerized search strategy.Study SelectionOne author (C.S.W.) screened the titles and abstracts of the computerized search to identify all potentially relevant articles. All retrieved articles were independently reviewed by 2 authors (C.S.W. and N.T.A.) for eligibility, assessment of methodological quality, and data abstraction. Only studies that evaluated the diagnostic accuracy of some element of the medical history, physical examination, or readily available diagnostic tests in adult patients with undifferentiated dyspnea presenting to the ED, regardless of whether the patients had known cardiac or pulmonary diseases, were included. Data had to be presented so that 2 × 2 contingency tables could be extracted. Because there currently is no widely accepted criterion standard for diagnosing heart failure, and because the focus of this review was a syndrome of heart failure, we accepted as a reasonable reference standard for heart failure that was a diagnosis agreed upon by a panel of physicians after evaluating for appropriate symptoms and signs of heart failure and an appropriate measure of cardiac dysfunction.We included studies that evaluated common and rapidly available tests (chest radiograph, ECG, and serum BNP) since clinicians rely on these basic investigations in conjunction with their history and physical examination in bedside decision making.There are currently multiple BNP assays approved by the US Food and Drug Administration for clinical use. To date, the largest published randomized clinical trials have been funded by industry and have reported using the BNP assay of one single manufacturer.An a priori decision was made to exclude studies that investigated other cardiac neurohormones such as A-type natriuretic peptide or other forms of BNP (eg, NT-proBNP). It was thought at the time of this review that there would be insufficient published data on these other neurohormones to draw significant conclusions. We also excluded studies that (1) were review articles with no original data, (2) had no clinical examination performed or reported, (3) used only echocardiography, computed tomography scans, or invasive hemodynamic monitoring alone as the reference standard for heart failure without clinical correlation because the results from these tests serve as part of the reference standard for a clinical diagnosis, (4) were population based, (5) enrolled patients younger than 18 years, and (6) did not specifically include patients reporting dyspnea. We resolved disagreements between reviewers on study selection, assessment of quality, and abstraction of data by consensus.Assessment of Study QualityStudy quality was assigned based on the grading scheme developed by Sackett et aland previously used for this series.Level 1 studies were primary prospective studies of the accuracy or precision of the clinical examination that involved comparisons of clinical findings (symptom or sign) with a reference standard of diagnosis among a large number (sufficient to have narrow confidence limits on the resulting sensitivity, specificity, or LRs) of consecutive or random patients with dyspnea. For precision studies, this required 2 or more independent blinded raters of symptoms or signs in a large number of patients. Level 2 studies were similar to level 1 but with smaller numbers of patients. Level 3 studies were comparisons of clinical findings with a reference standard of diagnosis among nonconsecutive or nonrandom patients with dyspnea. Studies of a retrospective nature were included as level 3. Level 4 studies were comparisons of clinical findings with a reference standard of diagnosis among convenience samples of patients who obviously have the target condition. Finally, level 5 studies were comparisons of clinical findings with a reference standard of unknown or uncertain validity among convenience samples of patients and perhaps, healthy patients.Statistical MethodsTwo authors (C.S.W. and N.T.A.) independently extracted data for analysis. Published raw data were used to construct 2 × 2 contingency tables for each clinical variable. Where data for the same variable was available from 2 or more sources, meta-analytical techniques were applied to combine results across studies. When multiple articles from the same group were found, the studies were carefully reviewed to ensure no data were analyzed in duplicate. Summary positive and negative LRs and 95% confidence intervals (CIs) were calculated using random-effects models based on the delta method.We display only the CIs of the LRs in the data tables, since these values are most useful to clinicians and include the sensitivity and specificity in the calculation. The choice of random-effects measures lowers the risk of CIs that are too optimistically narrow.Sensitivity is defined as the proportion of patients with heart failure who have a particular finding; specificity is the proportion of patients without heart failure who do not have the particular finding. The positive LR is the change in the odds of having heart failure when a particular finding is present, whereas the negative LR is the change in the odds of having heart failure when the particular finding is absent.RESULTSSearch ResultsA total of 815 citations were identified in our literature search. Of these, 682 were excluded after review of their titles and abstracts, with 133 studies remaining. These studies were reviewed in detail and we identified a total of 22 studies that evaluated the role of the clinical examination or basic routine investigation (chest radiograph, ECG, serum BNP) in patients with undifferentiated dyspnea and that also met our inclusion criteria.Study CharacteristicsOnly studies of sufficient quality (levels 1-3) were considered for the quantitative analysis. Of the 22 studies meeting inclusion, 18 were included in the meta-analysis (Table 2); while the remaining 4 studieswere levels 4 or 5 and were not included in the evidence tables.Table 2.Summary of Studies in Emergency Department PatientsSourceStudy Quality*Study DesignStudy CriteriaTotal Men, No. (%)Mean Age, yIncidence of Heart Failure, %Criteria Standard; Objective MeasureMueller et al,20051ProspectiveInclusion: ED with dyspnea Exclusion: acute myocardial infarction, trauma251 (93)72.855Retrospective review by 1 physician; echocardiographyLainchbury et al,20031ProspectiveInclusion: ED with dyspnea, able to give blood within 8 h of arrival Exclusion: n/a205 (49)70.034Retrospective review by 2 independent cardiologists; echocardiography, RVGLogeart et al,20021ProspectiveInclusion: ED with acute severe dyspnea Exclusion: acute myocardial infarction, chest injury, recent surgery, therapy instituted >2 h prior to arrival in ED, emergency echocardiography not feasible163 (67)67.471Retrospective review by 2 independent cardiologists and 1 pulmonologist; echocardiography, CC, RVG, PFTKnudsen et al,20042ProspectiveInclusion: ED with dyspnea Exclusion: chest pain, dyspnea clearly not secondary to heart failure155 (45)NA†48Retrospective review by 2 independent cardiologists; echocardiography, CC, RVG, PFTBayes-Genis et al,20042ProspectiveInclusion: ED with dyspnea, aged 40-88 y Exclusion: NYHA classes I and II, dyspnea secondary to chest trauma or cardiac tamponade, acute coronary syndromes without dyspnea, severe renal insufficiency, liver cirrhosis89 (60)70.783Retrospective review by 2 independent cardiologists; echocardiography, PFTVillacorta et al,20022ProspectiveInclusion: ED with dyspnea Exclusion: obvious diagnosis of dyspnea, acute coronary syndromes without dyspnea70 (47)72.051Retrospective review by 1 cardiologist; echocardiographyDavis et al,19942ProspectiveInclusion: ED with dyspnea requiring admission Exclusion: obvious cause of dyspnea, severe renal failure, acute chest pain52 (40)74.061Retrospective review by committee of physicians and a radiologist; echocardiography, PFTMarantz et al,19902ProspectiveInclusion: ED with dyspnea, aged ≥40 y, English speaking, able to consent, presented during study hours Exclusion: clinically unstable, non–English speaking, disoriented or unable to cooperate, refusal to consent, left against medical advice51 (39)64.045Retrospective review by 1 physician; echocardiographyAlibay et al,20053Convenience sampleInclusion: ED with dyspnea Exclusion: n/a160 (48)80.137.5Retrospective review by 2 independent cardiologists; echocardiographyRay et al,20043Convenience sampleInclusion: ED with dyspnea <2 wk, aged ≥65 y, respiratory rate >25/min or PaO2<70 mm Hg or PaCO2>45 mm Hg or SpO2<92% Exclusion: none308 (49)80.354.2Retrospective review by 2 independent experts; echocardiography, high-resolution computed tomography scan, PFTSpringfield et al,20043Convenience sampleInclusion: ED with dyspnea or respiratory rate >20/min or PaO2<90 mm Hg on room air Exclusion: pregnancy, aged ≤18 y, trauma patients, unconscious or unable to speak, <3 ′ 11’’ or >7 ’ 8’’ tall, <66 lb or >341 lb38 (42)67.232Retrospective review by 1 physician; echocardiographyMorrison et al,20023Convenience sampleInclusion: ED with dyspnea Exclusion: dyspnea clearly not secondary to heart failure, unstable angina/myocardial infarction without dyspnea321 (95)NA42Retrospective review by 2 independent cardiologists; echocardiography, CC, RVG, PFTMaisel et al,20023ProspectiveInclusion: ED with dyspnea as prominent symptom Exclusion: aged ≤18 y, dyspnea clearly not secondary to heart failure, acute myocardial infarction, unstable angina without dyspnea, renal failure on dialysis or creatinine clearance <0.25 mL/s1586 (56)64.047Retrospective review by 2 independent cardiologists; echocardiography, CC, RVG, PFTMcCullough et al,20023See Maisel et alSubgroup of Maisel et alwith information recorded for ED physician assessment of probability of heart failure1538 (56)64.047See Maisel et alDao et al,20013Convenience sampleInclusion: ED with dyspnea Exclusion: dyspnea clearly not secondary to heart failure, acute coronary syndromes without dyspnea250 (94)63.039Retrospective review by 2 independent cardiologists; echocardiography, CC, RCG, PFTAbbreviations: CC, cardiac catheterization; ED, emergency department; NYHA, New York Heart Association (classification of heart disease); PaCO2, arterial pressure of carbon dioxide; PaO2, arterial partial pressure of oxygen; PFT, pulmonary function test; RVG, radionuclide ventriculography; SpO2, peripheral oxygen saturation.*Study quality was assigned based on the grading scheme developed by Sackett et aland previously used for this series.See also “Assessment of Study Quality” in the “Methods” section for more details.†NA denotes that the mean age was not published in the source article.Precision of Clinical Examination and InvestigationsPrecision refers to the degree of variation between observers (interobserver variation) or within observers (intraobserver variation) for a particular finding. No study has specifically addressed the interobserver or intraobserver variability in the recording of findings in dyspneic patients ultimately diagnosed with the clinical syndrome of heart failure. However, analogous work has been done in other diagnoses including pulmonary diseases, acute coronary syndromes; and in comparison with echocardiography, nuclear imaging, and cardiac catheterization.In general, there is much variability in the precision of clinical findings associated with heart failure, reflecting the potentially subtle nature of findings, and variable examination skills of the clinician.Accuracy of the Clinical ExaminationThirteen studies examined the accuracy of the clinical examination for predicting the presence of heart failure in dyspneic patients assessed in the ED. The sensitivities, specificities, and corresponding positive and negative LRs for the findings are shown in Table 3.Table 3.Summary of Diagnostic Accuracy of Findings on History and Physical Examination in Emergency Department PatientsFindingPooledSummary LR (95% CI)*SensitivitySpecificityPositiveNegativeInitial clinical judgment0.610.864.4 (1.8-10.0)0.45 (0.28-0.73)HistoryHeart failure0.600.905.8 (4.1-8.0)0.45 (0.38-0.53)Myocardial infarction0.400.873.1 (2.0-4.9)0.69 (0.58-0.82)Coronary artery disease0.520.701.8 (1.1-2.8)0.68 (0.48-0.96)Dyslipidemia0.230.871.7 (0.43-6.9)0.89 (0.69-1.1)Diabetes mellitus0.280.831.7 (1.0-2.7)0.86 (0.73-1.0)Hypertension0.600.561.4 (1.1-1.7)0.71 (0.55-0.93)Smoker0.620.270.84 (0.58-1.2)1.4 (0.58-3.6)Chronic obstructive pulmonary disease0.340.570.81 (0.60-1.1)1.1 (0.95-1.4)SymptomsParoxysmal nocturnal dyspnea0.410.842.6 (1.5-4.5)0.70 (0.54-0.91)Orthopnea0.500.772.2 (1.2-3.9)0.65 (0.45-0.92)Edema0.510.762.1 (0.92-5.0)0.64 (0.39-1.1)Dyspnea on exertion0.840.341.3 (1.2-1.4)0.48 (0.35-0.67)Fatigue and weight gain0.310.701.0 (0.74-1.4)0.99 (0.85-1.1)Cough0.360.610.93 (0.70-1.2)1.0 (0.87-1.3)Physical examinationThird heart sound (ventricular filling gallop)0.130.9911 (4.9-25.0)0.88 (0.83-0.94)Abdominojugular reflux0.240.966.4 (0.81-51.0)0.79 (0.62-1.0)Jugular venous distension0.390.925.1 (3.2-7.9)0.66 (0.57-0.77)Rales0.600.782.8 (1.9-4.1)0.51 (0.37-0.70)Any murmur0.270.902.6 (1.7-4.1)0.81 (0.73-0.90)Lower extremity edema0.500.782.3 (1.5-3.7)0.64 (0.47-0.87)Valsalva maneuver0.730.652.1 (1.0-4.2)0.41 (0.17-1.0)Systolic blood pressure <100 mm Hg0.060.972.0 (0.60-6.6)0.97 (0.91-1.0)Fourth heart sound (atrial gallop)0.050.971.6 (0.47-5.5)0.98 (0.93-1.0)Systolic blood pressure ≥150 mm Hg0.280.731.0 (0.69-1.6)0.99 (0.84-1.2)Wheezing0.220.580.52 (0.38-0.71)1.3 (1.1-1.7)Ascites0.010.970.33 (0.04-2.9)1.0 (0.99-1.1)Abbreviations: CI, confidence interval; LR, likelihood ratio.*LRs are not independent of each other and should not be multiplied in series when multiple findings are considered.Overall Clinical GestaltThe overall clinical gestalt of the initial treating ED physician was associated with a high positive LR (4.4; 95% CI, 1.8-10.0) for a final diagnosis of heart failure. When the emergency physician assessed the dyspneic patient as unlikely to have heart failure, the odds decreased by about half (LR, 0.45; 95% CI, 0.28-0.73).Historical ItemsThe most useful historical features in confirming the presence of heart failure were a past medical history of congestive heart failure (LR, 5.8; 95% CI, 4.1-8.0), myocardial infarction (LR, 3.1; 95% CI, 2.0-4.9), or coronary artery disease (LR, 1.8; 95% CI, 1.1-2.8). Likewise, patients without a history of heart failure (LR, 0.45; 95% CI, 0.38-0.53), coronary artery disease (LR, 0.68; 95% CI, 0.48-0.96), or myocardial infarction (LR, 0.69; 95% CI, 0.58-0.82) were less likely to have their dyspnea explained by current heart failure.The results of other historical findings in Table 3had LR CIs that included 1.SymptomsThe presence of paroxysmal nocturnal dyspnea (LR, 2.6; 95% CI, 1.5-4.5), orthopnea (LR, 2.2; 95% CI, 1.2-3.9), or dyspnea on exertion (LR, 1.3; 95% CI, 1.2-1.4) increased the likelihood of heart failure. Likewise, the absence of dyspnea on exertion (LR, 0.48; 95% CI, 0.35-0.67), orthopnea (LR, 0.65; 95% CI, 0.45-0.92), or paroxysmal nocturnal dyspnea (LR, 0.70; 95% CI, 0.54-0.91) decreased the likelihood of heart failure.The results of other findings in Table 3had CIs that included 1.Physical ExaminationThe presence of a third heart sound (ventricular filling gallop) increased the likelihood of heart failure the most (LR, 11; 95% CI, 4.9-25.0). The presence of several other findings had CIs that excluded 1: jugular venous distension (LR, 5.1; 95% CI, 3.2-7.9), pulmonary rales (LR, 2.8; 95% CI, 1.9-4.1), any cardiac murmur (LR, 2.6; 95% CI, 1.7-4.1), and leg edema (LR, 2.3; 95% CI, 1.5-3.7). The presence of an abnormal abdominojugular reflux response (LR, 6.4; 95% CI, 0.81-51.0) had a high LR, but its evaluation in only 1 study of 51 patients led to broad CIs.An abnormal response to the Valsalva maneuver in the same study had an LR of 2.1 but the lower limit of the 95% CI was 1.0. The presence of the other findings in Table 3did not appear useful for assessing the likelihood of heart failure in dyspneic patients.The absence of pulmonary rales (LR, 0.51; 95% CI, 0.37-0.70), leg edema (LR, 0.64; 95% CI, 0.47-0.87), or jugular venous distension (LR, 0.66; 95% CI, 0.57-0.77) were the most useful findings that lowered the likelihood of heart failure. Wheezing also decreased the likelihood that a dyspneic patient had heart failure (LR, 0.52; 95% CI, 0.38-0.71). The absence of a third heart sound or a murmur lowered the likelihood of heart failure but the point estimate of the LR of these findings approached 1. The absence of the other findings in Table 3did not appear useful as the CI included 1. Diaphoresis as a sign of heart failure was of uncertain validity, having been evaluated in only 2 studies that were each of level 4 quality.Accuracy of Chest RadiographsSeven studies examined the accuracy of various chest radiograph findings in the ED setting (Table 4). The presence of any of these findings (except for any edema) had high positive LRs with CIs exceeding 1 and therefore, increased the likelihood of heart failure in dyspneic patients. The presence of pulmonary venous congestion (distension of pulmonary veins and redistribution to the apices) (n = 4 studies; summary LR, 12.0; 95% CI, 6.8-21.0) and cardiomegaly (n = 6 studies; summary LR, 3.3; 95% CI, 2.4-4.7) increased the likelihood of heart failure and have undergone more extensive evaluation so that the results may be more reliable. The presence of interstitial edema also had a high LR (n = 2 studies; summary LR, 12.0; 95% CI, 5.2-27.0). The presence of pneumonia or hyperinflation lowered the likelihood of heart failure but were only assessed in one study.Table 4.Summary of Diagnostic Accuracy of Findings on Chest Radiograph and Electrocardiogram in Emergency Department PatientsFindingPooledSummary LR (95% CI)*SensitivitySpecificityPositiveNegativeChest radiographPulmonary venous congestion†0.540.9612.0 (6.8-21.0)0.48 (0.28-0.83)Interstitial edema0.340.9712.0 (5.2-27.0)0.68 (0.54-0.85)Alveolar edema0.060.996.0 (2.2-16.0)0.95 (0.93-0.97)Cardiomegaly0.740.783.3 (2.4-4.7)0.33 (0.23-0.48)Pleural effusion(s)0.260.923.2 (2.4-4.3)0.81 (0.77-0.85)Any edema0.700.773.1 (0.60-16.0)0.38 (0.11-1.3)Pneumonia0.040.920.50 (0.29-0.87)1.0 (1.0-1.1)Hyperinflation0.030.920.38 (0.20-0.69)1.1 (1.0-1.1)ElectrocardiogramAtrial fibrillation0.260.933.8 (1.7-8.8)0.79 (0.65-0.96)New T-wave changes0.240.923.0 (1.7-5.3)0.83 (0.74-0.92)Any abnormal finding0.500.782.2 (1.6-3.1)0.64 (0.47-0.88)ST elevation0.050.971.8 (0.80-4.0)0.98 (0.94-1.0)ST depression0.110.941.7 (0.97-2.9)0.95 (0.90-1.0)Abbreviations: CI, confidence interval; LR, likelihood ratio.*LRs are not independent of each other and should not be multiplied in series when multiple findings are considered.†Pulmonary venous congestion, manifest as distension of pulmonary veins and redistribution to the apices.The most extensively evaluated chest radiograph findings (pulmonary venous congestion and cardiomegaly), were also the findings that when absent, had an LR that was appreciably different from 1. The absence of cardiomegaly was particularly useful (LR, 0.33; 95% CI, 0.23-0.48), with narrower CIs than the absence of pulmonary venous congestion (LR, 0.48; 95% CI, 0.28-0.83).Accuracy of ElectrocardiogramSeven studies examined the accuracy of various ECG findings in the ED setting (Table 4). The presence of atrial fibrillation in a dyspneic patient was the most important (LR, 3.8; 95% CI, 1.7-8.8) and evaluated in several studies (n = 5 studies). The presence of new T-wave changes (LR, 3.0; 95% CI, 1.7-5.3) or abnormal ECG findings (LR, 2.2; 95% CI, 1.6-3.1) increased the likelihood of heart failure but were evaluated in fewer studies. A completely normal ECG (LR, 0.64; 95% CI, 0.47-0.88) decreased the likelihood of heart failure and was the only normal finding that had a negative LR with a clinically meaningful difference from 1.Accuracy of BNPEleven studies examined the operating characteristics of various cutoffs of serum BNP in the ED setting (Table 5). Eight of these reported pharmaceutical industry sponsorship, 2 did not disclose funding sources, and only 1 study reported no pharmaceutical relationship.Table 5.Summary of Operating Characteristics of Serum BNP in Emergency Department PatientsPooledSummary LR (95% CI)SensitivitySpecificityPositiveNegativeClinical judgment or BNP ≥100 pg/mL*0.940.703.1 (2.8-3.5)0.09 (0.06-0.11)BNP alone, pg/mL≥2500.890.814.6 (2.6-8.0)0.14 (0.06-0.33)≥2000.920.753.7 (2.6-5.4)0.11 (0.07-0.18)≥1500.890.713.1 (2.1-4.5)0.15 (0.11-0.21)≥1000.930.662.7 (2.0-3.9)0.11 (0.07-0.16)≥800.960.713.3 (1.8-6.3)0.06 (0.03-0.13)≥500.970.441.7 (1.2-2.6)0.06 (0.03-0.12)Abbreviations: BNP, B-type natriuretic peptide; CI, confidence interval; LR, likelihood ratio.*Either an initial clinical probability of heart failure ≥80% or BNP ≥100 pg/mL was considered a positive result. A negative result was a clinical probability of heart failure <80% and BNP <100 pg/mL.As the BNP cutoff increased, the positive LR generally increased. Thus, the higher the value of BNP, the more suggestive it was of heart failure. However, no BNP threshold indicated the presence of heart failure with certainty. At any BNP threshold up to 250 pg/mL, values below the threshold always made heart failure much less likely (LR 0.06-0.15).BNP levels must be interpreted differently for patients with renal insufficiency. Based upon an analysis of data from the Breathing Not Properly Multinational Study,no adjustment in the 100-pg/mL threshold appears necessary for patients with an estimated glomerular filtration rate of 60 to 89 mL/min/1.73 m2, with an area under the receiver operating characteristic curve of 0.90 (a measure of overall accuracy). The loss of accuracy with worsening renal function can be minimized by using thresholds of 225 and 201 pg/mL, respectively, for patients with estimated glomerular filtration rates of 15to 29 and 30 to 59 mL/min/1.73 m2(areas under receiver operating characteristic curves of 0.86 and 0.81, respectively). The utility of BNP levels in patients with advanced renal insufficiency (estimated glomerular filtration rate <15 mL/min/1.73 m2or on dialysis) is unclear as these patients were not included in that study.Accuracy of Findings in Patients With History of Pulmonary DiseaseOne study (Table 6) examined the accuracy of symptoms, signs, ECG, and serum BNP in diagnosing heart failure in dyspneic ED patients with a prior history of asthma or COPD.This study was a subgroup analysis of the Breathing Not Properly Multinational Study.Table 6.Diagnostic Accuracy of History, Physical Examination, and Tests of Cardiac Function in Emergency Department Patients With History of Asthma or Chronic Obstructive Pulmonary Disease*FindingSensitivitySpecificityPositive LR (95% CI)†Negative LR (95% CI)†Initial clinical judgment0.370.969.9 (5.3-18)0.65 (0.55-0.77)HistoryAtrial fibrillation0.320.924.1 (2.5-6.6)0.74 (0.63-0.85)Coronary artery bypass grafting0.130.952.8 (1.3-5.8)0.92 (0.84-0.99)Myocardial infarction0.250.882.2 (1.4-3.5)0.84 (0.74-0.96)Diabetes mellitus0.260.872.0 (1.3-3.2)0.85 (0.74-0.97)Coronary artery disease0.490.752.0 (1.5-2.6)0.67 (0.54-0.84)Angina0.210.881.7 (1.0-2.8)0.90 (0.80-1.0)Hypertension0.540.551.2 (0.95-1.5)0.84 (0.65-1.1)SymptomsOrthopnea0.700.441.3 (1.1-1.5)0.68 (0.48-0.95)Fatigue0.740.341.1 (0.96-1.3)0.79 (0.54-1.2)Nocturnal cough0.490.470.93 (0.73-1.2)1.1 (0.85-1.4)Physical examinationThird heart sound (ventricular filling gallop)0.171.0057.0 (7.6-425)0.83 (0.75-0.91)Jugular venous distension0.410.904.3 (2.8-6.5)0.65 (0.54-0.78)Lower extremity edema0.690.752.7 (2.2-3.5)0.41 (0.30-0.57)Rales0.710.732.6 (2.1-3.3)0.39 (0.28-0.55)Hepatic congestion0.140.942.4 (1.2-4.7)0.91 (0.84-1.0)Enlarged heart0.030.981.6 (0.43-6.2)0.99 (0.95-1.0)Wheezing0.420.500.85 (0.65-1.1)1.2 (0.94-1.4)Chest radiographEdema0.340.9711.0 (5.8-22.0)0.68 (0.58-0.79)Cardiomegaly0.490.937.1 (4.5-11.0)0.54 (0.44-0.67)Pleural effusion(s)0.260.944.6 (2.6-8.0)0.78 (0.69-0.89)Pneumonia0.080.921.0 (0.46-2.3)1.0 (0.93-1.1)Hyperinflation0.080.850.53 (0.25-1.1)1.1 (1.0-1.2)Normal0.050.570.11 (0.04-0.28)1.7 (1.5-1.8)ElectrocardiogramAtrial fibrillation0.310.956.0 (3.4-10.0)0.73 (0.63-0.84)Ischemic ST-T waves0.210.954.6 (2.4-8.7)0.83 (0.74-0.93)Q waves0.220.933.1 (1.8-5.5)0.84 (0.75-0.94)B-type natriuretic peptide ≥100 pg/mL0.930.774.1 (3.3-5.0)0.09 (0.04-0.19)Abbreviation: LR, likelihood ratio.*Adapted fromMcCullough et al.†LRs are not independent of each other and should not be multiplied in series when multiple findings are considered.Initial Clinical GestaltA high initial clinical suspicion by the emergency physician (≥80% probability) was associated with a high likelihood for a final diagnosis of heart failure (LR, 9.9; 95% CI, 5.3-18.0) while an intermediate (21%-79%) or low (≤20%) initial clinical suspicion decreased the likelihood of heart failure (LR, 0.65; 95% CI, 0.55-0.77) but did not exclude it. In fact, 32% of patients in the intermediate suspicion group and 9% of patients in the low clinical suspicion group were ultimately diagnosed with heart failure. Assigning a lower probability to the low suspicion group (eg, ≤5%) would likely have reduced misclassification in that study.Historical ItemsThe presence of most historical findings in Table 6increased the likelihood of heart failure with CIs excluding 1. A history of prior atrial fibrillation (LR, 4.1; 95% CI, 2.5-6.6) or coronary bypass surgery (LR, 2.8; 95% CI, 1.3-5.8) were the most useful findings that increased the likelihood of heart failure. The absence of relevant historical features did not result in clinically meaningful LRs less than 1 other than perhaps the absence of coronary artery disease (LR, 0.67; 95% CI, 0.54-0.84).SymptomsOnly the absence of orthopnea (LR, 0.68; 95% CI, 0.48-0.95) had an LR that was appreciably different from 1. Thus, symptoms were not particularly useful among dyspneic patients with lung disease in determining who might also have heart failure.Physical ExaminationThe presence of a third heart sound had a very high diagnostic value for heart failure (LR, 57; 95% CI, 7.6-425.0). Other useful physical examination findings, when present, included jugular venous distension (LR, 4.3; 95% CI, 2.8-6.5), lower extremity edema (LR, 2.7; 95% CI, 2.2-3.5), pulmonary rales (LR, 2.6; 95% CI, 2.1-3.3), or hepatic congestion (LR, 2.4; 95% CI, 1.2-4.7). The absence of pulmonary rales (LR, 0.39; 95% CI, 0.28-0.55), lower extremity edema (LR, 0.41; 95% CI, 0.30-0.57), or jugular venous distension (LR, 0.65; 95% CI, 0.54-0.78) decreased the likelihood of heart failure.Chest RadiographThe presence of edema was the most useful radiographic finding for increasing the likelihood of heart failure (LR, 11.0; 95% CI, 5.8-22.0). Other very useful findings were cardiomegaly (LR, 7.1; 95% CI, 4.5-11.0) or pleural effusion(s) (LR, 4.6; 95% CI, 2.6-8.0). A normal chest radiograph (LR, 0.11; 95% CI, 0.04-0.28), absence of cardiomegaly (LR, 0.54; 95% CI, 0.44-0.67), or absence of edema (LR, 0.68; 95% CI, 0.58-0.79) decreased the likelihood of heart failure.ElectrocardiogramThe presence of ECG findings of atrial fibrillation (LR, 6.0; 95% CI, 3.4-10.0), ischemic ST-T wave changes (LR, 4.6; 95% CI. 2.4-8.7), or Q waves (LR, 3.1; 95% CI, 1.8-5.5) were all helpful toward suggesting a diagnosis of heart failure in the dyspneic ED patient with a history of pulmonary disease. No single ECG result had clinically useful outcomes for lowering the likelihood of heart failure.B-type Natriuretic PeptideBNP levels can rise in patients with chronic pulmonary diseases due to right ventricular strain. Nevertheless, BNP appears to still be useful in these patients. Studies have demonstrated that BNP levels are significantly higher in patients with a history of chronic lung disease but acute dyspnea from heart failure, compared with those with a history of heart failure but acute dyspnea from lung disease.Serum BNP for dyspneic patients with a history of asthma or COPD was useful for identifying heart failure (BNP ≥100 pg/mL: LR 4.1; 95% CI, 3.3-5.0). However, it was more powerful for excluding heart failure when low (BNP <100 pg/mL: LR 0.09; 95% CI, 0.04-0.19). However, this was only one study and thus, the optimal cutoff for BNP to diagnose or exclude clinical heart failure in dyspneic patients with chronic lung diseases is unclear.COMMENTIt is both important and difficult to rapidly differentiate among the common causes of dyspnea in ED patients. The syndrome of heart failure requires appropriate symptoms along with objective measures of cardiac dysfunction.Although sophisticated and invasive tests such as Swan-Ganz catheterization can help to distinguish between cardiac and pulmonary causes of dyspnea, they are frequently unavailable in the acute setting and thus, the diagnosis of heart failure and the decision to institute therapy on an emergent basis rests on the bedside clinical assessment (chest radiograph, ECG, and recently, serum BNP). Relying purely on echocardiography to diagnose clinical heart failure is also problematic because it is often not easily accessible, requires specialized training,and may not always truly reflect the current cause of dyspnea.That is, not every patient presenting with heart failure will have a diminished left ventricular ejection fraction; patients with diastolic heart failure for instance, may have elevated filling pressures and dyspnea in the presence of normal ejection fraction. The reverse is also true in that patients with a decreased left ventricular ejection fraction may be dyspneic from noncardiac causes such as COPD, and furthermore, the severity of impairment of ejection fraction does not always correlate with subjective severity of dyspnea.In this systematic review, many features on clinical examination, chest radiograph, ECG, and serum BNP were useful in diagnosing heart failure in adult ED patients presenting with dyspnea in whom heart failure was suspected. Features listed in Box 1were assessed in more than 1 study and were useful when either present or absent. Other findings may prove useful when evaluated further.Box 1. Features Useful in Diagnosing Heart Failure in Adult Emergency Department Patients With DyspneaHistorical FeaturesHeart failureMyocardial infarctionCoronary artery diseaseSymptomsParoxysmal nocturnal dyspneaOrthopneaDyspnea on exertionPhysical ExaminationListening for a third heart sound (ventricular filling gallop)Jugular venous pressure assessmentAuscultating for rales and wheezingAuscultating for a murmurAssessing the legs for edemaChest RadiographPulmonary venous congestionInterstitial edemaCardiomegalyPleural effusion(s)Electrocardiogram FindingsAtrial fibrillationAn abnormal resultB-type Natriuretic PeptideMost useful when <100 pg/mL for decreasing the likelihood of heart failureClinician’s Overall AssessmentOur results are consistent with those of Marcus and colleagues.They recently studied patients undergoing elective left heart catheterization, comparing the test characteristics of third and fourth heart sounds with objective measures of left ventricular dysfunction. While the patient population and reference standard for heart failure were different in our study when compared with theirs (eg, ventricular dysfunction vs a clinical diagnosis of heart failure), both studies found that third and fourth heart sounds had greater specificity than sensitivity, and that a third heart sound had a better specificity than a fourth heart sound for the diagnosis of heart failure.We did not find any studies examining combinations of historical and physical examination findings in making a diagnosis of heart failure. However, our analysis suggests that the initial clinical gestalt of the physician based on available information (history, physical, chest radiograph, ECG) is valuable. Because the overall clinical gestalt had LRs that approximate some of the individual findings, along with a lack of consistent multivariate models, we do not know whether all the symptoms and signs are independently useful. When clinicians are not confident in their clinical gestalt, they should preferentially rely on the results of the few findings that have LR estimates most different from 1.It is notable that a high initial clinical suspicion alone (LR 4.4; 95% CI, 1.8-10.0) (Table 3) had a greater positive LR than a composite of (high clinical suspicion or BNP ≥100 pg/mL or both), which had a combined positive LR of 3.1 (95% CI, 2.8-3.5) (Table 5). This suggests that BNP may not contribute much more in patients for whom the initial clinical suspicion of heart failure was already very high. However, in patients for whom the initial clinical suspicion of heart failure was not very high, BNP at a threshold value of 100 pg/mL was useful, especially for excluding heart failure in this group of patients. To apply these results correctly, it is necessary that clinicians first quantify and acknowledge their clinical suspicion (eg, formulate a pretest probability). If the physician waits until the BNP results are available before establishing clinical suspicion, these tests are no longer independent and the clinical suspicion becomes biased by the BNP. The results of our BNP analysis adds support to recent European guidelines for diagnosing heart failure, which state that BNP may be a clinically useful test to rule out heart failure due to its high negative predictive values.Clinicians should be aware that factors other than heart failure can affect serum BNP levels (Box 2). Algorithms for the use of the BNP test have been proposedbut not extensively validated.Box 2. Factors That Can Affect B-type Natriuretic Peptide (BNP) Levels*Factors (Other Than Heart Failure) That Cause Elevated BNP LevelsAdvanced ageRenal failureAcute coronary syndromesLung disease with cor pulmonaleAcute large pulmonary embolismHigh output cardiac statesFactors That Lower BNP in the Setting of Heart FailureAcute pulmonary edemaStable New York Heart Association class I patients with low ejection fractionAcute mitral regurgitationMitral stenosisAtrial myxoma*Adapted from Maisel et al.LimitationsThe results of our meta-analysis should be interpreted in the context of study limitations. One limitation of this review is the reference standard for heart failure (adjudication by a panel of physicians). Given the subjectivity and potential bias of such a standard, many of the studies had disagreement (up to 10%) among the adjudicators of whether heart failure was the contributing cause of dyspnea. However, in the absence of a true criterion standard for this clinical syndrome, the reference standard, while imperfect, is likely the best available and consistent with the clinical focus of this review. Another limitation that arises from using a clinical reference standard is that the final diagnosis of heart failure may not have been made independently of the individual findings of interest. That is, the panel of physicians may have used some of the clinical findings in deciding whether patients ultimately had heart failure as the cause of their dyspnea. As such, this may overestimate sensitivities and specificities. While this is a valid concern, we believe the effects on each individual finding would be small because the final diagnosis relied on a combination of information from many diverse sources, including any or all of the following: history, physical examination, routine laboratory tests, chest radiograph, ECG, heart failure scores, objective measures of cardiac function (eg, echocardiography, radionuclide ventriculography, radionuclide angiography, and left ventriculography at cardiac catheterization), pulmonary function tests, response to treatment, hospitalization course, and follow-up records.It is also important to note that our data are derived from studies of patients presenting to the ED with dyspnea. Therefore, these results may not generalize to inpatients, outpatients in clinic settings who may have more chronic dyspnea, or patients without dyspnea. The 18 studies included in this meta-analysis represent diverse and heterogeneous populations with various comorbidities. The majority of the studies excluded patients with acute coronary syndromes and in whom an obvious cause of dyspnea (eg, pneumothorax, trauma) was present. All the studies of BNP excluded patients in whom dyspnea was clearly not secondary to heart failure. Therefore, the usefulness of BNP from our analysis can only be applied to patients in whom the diagnosis of heart failure is a consideration. In patients in whom the suspicion of heart failure is very low (after taking a careful history and performing the physical examination, chest radiograph, and ECG), a BNP level is unlikely to affect diagnosis or management (eg, an obvious pulmonary etiology of dyspnea).Other limitations include the inherent subjectivity of clinical findings on history, physical examination, chest radiograph, and ECG. It is impossible to confirm the accuracy of individual findings presented in each study, and no formal definitions were given. For example, we do not have standardized information on the technique used for each chest radiograph performed (eg, anteroposterior, posteroanterior, portable).The Bottom LineThe features evaluated in more than one study with the highest LRs (LR >3.5) for diagnosing heart failure were the following: the overall clinical judgment, history of heart failure, a third heart sound, jugular venous distension, radiographic pulmonary venous congestion or interstitial edema, and electrocardiographic atrial fibrillation.The features evaluated in more than one study with the lowest LRs (LR <0.60) for diagnosing of heart failure were the following: the overall clinical judgment, no prior history of heart failure, no dyspnea on exertion, the absence of rales, and the absence of radiographic pulmonary venous congestion, or cardiomegaly. The single finding that decreased the likelihood of heart failure the most was a BNP <100 pg/mL (for patients with an estimated glomerular filtration rate of 15-60 mL/min/1.73 m2, a threshold of 201 pg/mL can be used). However, the clinician must always remember to first quantify and acknowledge his/her clinical suspicion based on their clinical examination prior to interpreting the BNP result.In the subgroup of ED patients with a prior history of asthma or COPD, the features that strongly suggested a diagnosis of heart failure were the overall clinical assessment, a third heart sound, radiographic edema or cardiomegaly, and electrocardiographic atrial fibrillation. The features that suggested against the diagnosis were normal chest radiograph and low serum BNP (<100 pg/mL). However, these results are from a subgroup analysis of one study and requires further confirmation.While the findings of this study are useful when assessing dyspneic patients suspected of having heart failure, no individual feature is sufficiently powerful in isolation to rule heart failure in or out. Therefore, an overall clinical impression based on all available information is best. If the appropriate constellation of findings with high LRs for heart failure are present, that may be sufficient to warrant empirical treatment without further urgent investigations. Conversely, if the clinical suspicion of heart failure is very low (eg, pulmonary disease), the physician should investigate and treat other causes of dyspnea.SCENARIO RESOLUTIONCase 1The patient has many features that raise the suspicion of heart failure, such as previous myocardial infarction (LR 3.1), previous heart failure (LR 5.8), orthopnea (LR 2.2), paroxysmal nocturnal dyspnea (LR 2.6), elevated jugular venous pressure (LR 5.1), a third heart sound (LR 11.0), rales (LR 2.8), extremity edema (LR 2.3), cardiomegaly (LR 3.3), and atrial fibrillation (LR 3.8), and only the single feature of wheezing (LR 0.52) that lowers the suspicion slightly. The overall constellation of symptoms and signs is so suggestive of heart failure that additional testing is not needed to make the diagnosis.Case 2Both heart failure and obstructive airways disease are considerations. The symptoms of dyspnea on exertion and cough are not helpful in making a diagnosis of heart failure because their LRs are close to 1. Rales (LR 2.8) and ECG showing ST depression (LR 1.7) both increase the likelihood of heart failure but more importantly, the findings of pulmonary venous congestion and interstitial edema are both associated with large LRs (>10.0) that significantly raise the suspicion for heart failure. Wheezing reduces the likelihood somewhat (LR 0.52). Based on the information available, the patient likely has acute heart failure and should be treated without waiting for further tests. An ECG should be ordered nonurgently. Furthermore, the patient may also be having a superimposed COPD or asthma exacerbation. The physician should consider ordering pulmonary function tests to confirm a diagnosis of obstructive airways disease.Case 3There are some features that increase the likelihood of heart failure, such as history of myocardial infarction (LR 2.2), elevated jugular venous pressure (LR 4.3), lower extremity edema (LR 2.7), and Q waves (LR 3.1), while other features decrease the likelihood (wheezing, LR 0.85) and normal chest radiograph (LR 0.11). Based on these LRs, there are insufficient data to make or exclude a diagnosis of heart failure. In this case, a BNP level could be very helpful. If it were less than 100 pg/mL, heart failure would be extremely unlikely (LR 0.09). If it were elevated, the probability of heart failure is higher but not diagnostic. More urgent echocardiogram and pulmonary function studies would be appropriate next steps.Corresponding Author:Najib T. Ayas, MD, MPH, Division of Respiratory Medicine, Vancouver Hospital and Heath Sciences Centre, 2775 Heather St, Vancouver, British Columbia, Canada V5Z 3J5 (najib.ayas@vch.ca).Financial Disclosures: None of the authors of this review have any professional or financial affiliations with manufacturers of BNP assays.Funding/Support:Dr Ayas is a recipient of a New Investigator Award from the Canadian Institute for Health Research and BC Lung Association, a Departmental Scholar Award from the University of British Columbia Department of Medicine, and a Scholar Award from the Michael Smith Foundation for Health Research. Dr FitzGerald is a recipient of a Canadian Institute for Health Research and BC Lung Association Scientist Award, and a Michael Smith Distinguished Scholar Award.Role of the Sponsor:The funding sources had no role in the design, collection, analysis, or interpretation of data; nor in the decision to submit the manuscript for publication.Acknowledgment:The authors thank David Simel, MD, for his valuable guidance during this study. We are also grateful to Robert Badgett, MD, University of Texas Health Science Center at San Antonio; Michael Cuffe, MD, Duke University Health System, and Karen Welty-Wolf, MD, Division of Pulmonary and Critical Care Medicine, Duke University Medical Center, Durham VA Medical Center, Durham, NC; for their expert advice and helpful reviews of earlier versions of the manuscript.REFERENCESEBraunwaldShattuck lecture—cardiovascular medicine at the turn of the millennium: triumphs, concerns, and opportunities.N Engl J Med1997337136013699358131SAHuntDWBakerMHChinAmerican College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1995 Guidelines for the Evaluation and Management of Heart Failure); International Society for Heart and Lung Transplantation; Heart Failure Society of AmericaACC/AHA Guidelines for the Evaluation and Management of Chronic Heart Failure in the Adult: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1995 Guidelines for the Evaluation and Management of Heart Failure.Circulation20011042996300711739319MMRedfieldHeart failure: an epidemic of uncertain proportions.N Engl J Med20023471442144412409548PAMcCulloughEFPhilbinJASpertusSKaatzKRSandbergWDWeaverConfirmation of a heart failure epidemic: findings from the Resource Utilization Among Congestive Heart Failure (REACH) study.J Am Coll Cardiol200239606911755288WJRemmeKSwedbergGuidelines for the diagnosis and treatment of chronic heart failure.Eur Heart J2001221527156011492984CMuellerAScholerKLaule-KilianUse of B-type natriuretic peptide in the evaluation and management of acute dyspnea.N Engl J Med200435064765414960741MWRichRFNeaseCost-effectiveness analysis in clinical practice: the case of heart failure.Arch Intern Med19991591690170010448770ACBalesMJSorrentinoCauses of congestive heart failure: prompt diagnosis may affect prognosis.Postgrad Med199710144469008688RCWuerzSAMeadorEffects of prehospital medications on mortality and length of stay in congestive heart failure.Ann Emerg Med1992216696741590605HLManningRMSchwartzsteinPathophysiology of dyspnea.N Engl J Med1995333154715537477171KWassermanRCasaburiDyspnea: physiological and pathophysiological mechanisms.Annu Rev Med1988395035153285788CLSpringfieldFSebatDJohnsonSLengleCSebatUtility of impedance cardiography to determine cardiac vs. noncardiac cause of dyspnea in the emergency department.Congest Heart Fail200410141615073480JFMurrayedJANadeledTextbook of Respiratory Medicine.3rd ed. Philadelphia, Pa: WB Saunders Co; 2000EBraunwaldHeart failure.In: Braunwald E, Fauci AS, Kasper DL, et al, eds. Harrison's Principles of Internal Medicine.15th ed. New York, NY: McGraw-Hill; 2001:1318-29DyspneaMechanisms, assessment, and management: a consensus statement: American Thoracic Society.Am J Respir Crit Care Med19991593213409872857EBraunwaldPathophysiology of heart failure.In: Heart Disease.Philadelphia, Pa: WB Saunders Co; 1992:393RGBadgettCRLuceyCDMulrowCan the clinical examination diagnose left-sided heart failure in adults?JAMA1997277171217199169900PAPoole-WilsonRelation of pathophysiologic mechanisms to outcome in heart failure.J Am Coll Cardiol19932222A29A8376693AJCoatsSAdamopoulosTEMeyerJConwayPSleightEffects of physical training in chronic heart failure.Lancet199033563661967416JRWilsonDMManciniWBDunkmanExertional fatigue due to skeletal muscle dysfunction in patients with heart failure.Circulation1993874704758425294CDMulrowCRLuceyLEFarnettDiscriminating causes of dyspnea through clinical examination.J Gen Intern Med199383833928410400MRCowiePJourdainAMaiselClinical applications of B-type natriuretic peptide (BNP) testing.Eur Heart J2003241710171814522565KLBaughmanB-type natriuretic peptide: a window to the heart.N Engl J Med200234715815912124403DRHollemanJrDLSimelDoes the clinical examination predict airflow limitation?JAMA19952733133197815660JPMetlayWNKapoorMJFineDoes this patient have community-acquired pneumonia? diagnosing pneumonia by history and physical examination.JAMA1997278144014459356004SDChunilalJWEikelboomJAttiaDoes this patient have pulmonary embolism?JAMA20032902849285814657070RAReevesDoes this patient have hypertension? how to measure blood pressure.JAMA1995273121112187707630DJCookDLSimelDoes this patient have abnormal central venous pressure?JAMA19962756306348594245EEtchellsCBellKRobbDoes this patient have an abnormal systolic murmur?JAMA19972775645719032164NKChoudhryEEEtchellsDoes this patient have aortic regurgitation?JAMA19992812231223810376577PRMarantzMCKaplanMHAldermanClinical diagnosis of congestive heart failure in patients with acute dyspnea.Chest1990977767812182296RWArdRHTwinningEvaluation of Valsalva test in bedside diagnosis of dyspnea.Am J Med Sci195723440341213469801JHKnowlesRGorlinCFStoreyClinical test for pulmonary congestion with use of the Valsalva maneuver.JAMA1956160444813271161JWEngstromJBMartinDisorders of the autonomic nervous system.In: Braunwald E, Fauci AS, Kasper DL, et al, eds. Harrison's Principles and Practice of Internal Medicine.15th ed. New York, NY: McGraw-Hill Publishers; 2001:2416-2421FALederleDLSimelDoes this patient have abdominal aortic aneurysm?JAMA199928177829892455LKMorrisonAHarrisonPKrishnaswamyRKazanegraPCloptonAMaiselUtility of a rapid B-natriuretic peptide assay in differentiating congestive heart failure from lung disease in patients presenting with dyspnea.J Am Coll Cardiol20023920220911788208DLSackettRBHaynesGHGuyattPTugwellClinical Epidemiology: A Basic Science for Clinical Medicine.2nd ed. Boston, Mass: Little Brown; 1991CRRaoLinear Statistical Inference and Its Applications.New York, NY: John Wiley & Sons; 1968ASMaiselPKrishnaswamyRMNowakRapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure.N Engl J Med200234716116712124404PAMcCulloughRMNowakJMcCordB-type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: analysis from Breathing Not Properly (BNP) Multinational Study.Circulation200210641642212135939CWKnudsenTOmlandPCloptonDiagnostic value of B-type natriuretic peptide and chest radiographic findings in patients with acute dyspnea.Am J Med200411636336815006584JGLainchburyECampbellCMFramptonTGYandleMGNichollsAMRichardsBrain natriuretic peptide and n-terminal brain natriuretic peptide in the diagnosis of heart failure in patients with acute shortness of breath.J Am Coll Cardiol20034272873512932611DLogeartCSaudubrayPBeyneComparative value of Doppler echocardiography and B-type natriuretic peptide assay in the etiologic diagnosis of acute dyspnea.J Am Coll Cardiol2002401794180012446063CWKnudsenJSRiisAVFinsenDiagnostic value of a rapid test for B-type natriuretic peptide in patients presenting with acute dyspnoe: effect of age and gender.Eur J Heart Fail20046556215012919ABayes-GenisMSantalo-BelEZapico-MunizN-terminal probrain natriuretic peptide (NT-proBNP) in the emergency diagnosis and in-hospital monitoring of patients with dyspnoea and ventricular dysfunction.Eur J Heart Fail2004630130814987580HVillacortaADuarteNMDuarteThe role of B-type natriuretic peptide in the diagnosis of congestive heart failure in patients presenting to an emergency department with dyspnea.Arq Bras Cardiol20027956957212532240MDavisEEspinerGRichardsPlasma brain natriuretic peptide in assessment of acute dyspnoea.Lancet19943434404447905953QDaoPKrishnaswamyRKazanegraUtility of B-type natriuretic peptide in the diagnosis of congestive heart failure in an urgent-care setting.J Am Coll Cardiol20013737938511216950LHBrownJEGoughRHSeimCan quantitative capnometry differentiate between cardiac and obstructive causes of respiratory distress?Chest19981133233269498946RMMcNamaraDJCionniUtility of the peak expiratory flow rate in the differentiation of acute dyspnea: cardiac vs pulmonary origin.Chest19921011291321729059DFleischerEAEspinerTGYandleRapid assay of plasma brain natriuretic peptide in the assessment of acute dyspnoea.N Z Med J199711071749137299PAMcCulloughJEHollanderRMNowakUncovering heart failure in patients with a history of pulmonary disease: rationale for the early use of B-type natriuretic peptide in the emergency department.Acad Emerg Med20031019820412615582JLJanuzziJrCACamargoSAnwaruddinThe N-terminal Pro-BNP investigation of dyspnea in the emergency department (PRIDE) study.Am J Cardiol20059594895415820160YAlibayABeauchetREl MahmoudPlasma N-terminal pro-brain natriuretic peptide and brain natriuretic peptide in assessment of acute dyspnea.Biomed Pharmacother200559202415740931PRayMArthaudYLefortSBirolleauCBeigelmanBRiouUsefulness of B-type natriuretic peptide in elderly patients with acute dyspnea.Intensive Care Med2004302230223615502931TMuellerAGegenhuberWPoelzMHaltmayerDiagnostic accuracy of B type natriuretic peptide and amino terminal proBNP in the emergency diagnosis of heart failure.Heart20059160661215831643NGadsbollPFHoilund-CarlsenGGNielsenSymptoms and signs of heart failure in patients with myocardial infarction: reproducibility and relationship to chest X-ray, radionuclide ventriculography and right heart catheterization.Eur Heart J198910101710282591393AAIshmailSWingJFergusonTAHutchinsonSMagderKMFlegelInterobserver agreement by auscultation in the presence of a third heart sound in patients with congestive heart failure.Chest1987918708733581934DJCookClinical assessment of central venous pressure in the critically ill.Am J Med Sci19902991751782316561SMButmanGAEwyJRStandenKBKernEHahnBedside cardiovascular examination in patients with severe chronic heart failure: importance of rest or inducible jugular venous distension.J Am Coll Cardiol1993229689748409071RGBadgettCDMulrowPMOttoGRamirezHow well can the chest radiograph diagnose left ventricular dysfunction?J Gen Intern Med1996116256348945695AAPanjuBRHemmelgarnGHGuyattDLSimelIs this patient having a myocardial infarction?JAMA1998280125612639786377DRHollemanJrDLSimelJSGoldbergDiagnosis of obstructive airways disease from the clinical examination.J Gen Intern Med1993863688441077PAMcCulloughPDucTOmlandB-type natriuretic peptide and renal function in the diagnosis of heart failure: an analysis from the Breathing Not Properly Multinational Study.Am J Kidney Dis20034157157912612980LCabanesBRichaud-ThiriezYFullaBrain natriuretic peptide blood levels in the differential diagnosis of dyspnea.Chest20011202047205011742939JBSewardPSDouglasRErbelHand-carried cardiac ultrasound (HCU) device: recommendations regarding new technology: a report from the Echocardiography Task Force on New Technology of the Nomenclature and Standards Committee of the American Society of Echocardiography.J Am Soc Echocardiogr20021536937311944016RBDevereuxPRLiebsonMJHoranRecommendations concerning use of echocardiography in hypertension and general population research.Hypertension19879II97II1042948915NKWengerLeft ventricular dysfunction, exercise capacity and activity recommendations.Eur Heart J19889(suppl F)63662968910GMMarcusILGerberBHMcKeownAssociation between phonocardiographic third and fourth heart sounds and objective measures of left ventricular function.JAMA20052932238224415886379AMaiselB-type natriuretic peptide measurements in diagnosing congestive heart failure in the dyspneic emergency department patient.Rev Cardiovasc Med20023(suppl 4)S10S1712439426
TI - Does This Dyspneic Patient in the Emergency Department Have Congestive Heart Failure?
JF - JAMA
DO - 10.1001/jama.294.15.1944
DA - 2005-10-19
UR - https://www.deepdyve.com/lp/american-medical-association/does-this-dyspneic-patient-in-the-emergency-department-have-congestive-7Ycy4FgR9A
SP - 1944
EP - 1956
VL - 294
IS - 15
DP - DeepDyve
ER -