Author + information
- Received August 6, 2013
- Revision received October 1, 2013
- Accepted October 3, 2013
- Published online February 1, 2014.
- Guillem Caldentey, MD∗,
- Paul Khairy, MD, PhD∗,†,
- Denis Roy, MD∗,
- Hugues Leduc, MSc†,
- Mario Talajic, MD∗,
- Normand Racine, MD∗,
- Michel White, MD∗,
- Eileen O'Meara, MD∗,
- Marie-Claude Guertin, PhD†,
- Jean L. Rouleau, MD∗ and
- Anique Ducharme, MD, MSc∗∗ ()
- ∗Montreal Heart Institute, Université de Montréal, Montreal, Quebec, Canada
- †Montreal Heart Institute Coordinating Center, Montreal, Quebec, Canada
- ↵∗Reprint requests and correspondence:
Dr. Anique Ducharme, Montreal Heart Institute Research Center (Room S-2700), 5000, Belanger East, Montreal (Quebec) H1T 1C8. Canada.
Objectives This study sought to assess the prognostic value of physical examination in a modern treated heart failure population.
Background The physical examination is the cornerstone of the evaluation and monitoring of patients with heart failure. Yet, the prognostic value of congestive signs (i.e., peripheral edema, jugular venous distension, a third heart sound, and pulmonary rales) has not been assessed in the current era.
Methods A post-hoc analysis was conducted on all 1,376 patients, 81% male, mean age 67 ± 11 years, with symptomatic left ventricular systolic dysfunction enrolled in the AF-CHF (Atrial Fibrillation and Congestive Heart Failure) trial. The prognostic value of baseline physical examination findings was assessed in univariate and multivariate Cox regression analyses.
Results Peripheral edema was observed in 425 (30.9%), jugular venous distension in 297 (21.6%), a third heart sound in 207 (15.0%), and pulmonary rales in 178 (12.9%) patients. Death from cardiovascular causes occurred in 357 (25.9%) patients over a mean follow-up of 37 ± 19 months. All 4 physical examination findings were associated with cardiovascular mortality in univariate analyses (all p values <0.01). In multivariate analyses, taking all 4 signs as potential covariates, only rales (hazard ratio 1.41; 95% confidence interval: 1.07 to 1.86; p = 0.013) and peripheral edema (hazard ratio: 1.25; 95% confidence interval: 1.00 to 1.57; p = 0.048) were associated with cardiovascular mortality, independent of other variables.
Conclusions In the modern era, congestive signs on the physical examination (i.e., peripheral edema, jugular venous distension, a third heart sound, and pulmonary rales) continue to provide important prognostic information in patients with congestive heart failure.
Heart failure is considered a major epidemic of the modern era. It is associated with substantial morbidity, mortality, and healthcare resource utilization. Demographic trends suggest that the prevalence of heart failure will continue to rise with the aging population. In patients with chronic heart failure (CHF) due to systolic dysfunction, several factors associated with a poorer prognosis have been identified, including functional parameters (1–3), echocardiographic (4–6) and electrocardiographic (7) indices, and serum biomarkers (8–12).
The value of a careful and thorough physical examination has come into question in an age when detailed imaging studies and ancillary testing are readily available (13,14). Prior publications from studies performed some decades ago, when pharmacological management reflected knowledge of the time, supported the prognostic role of congestive signs. It remains to be demonstrated whether such findings may be generalized to patients with CHF on optimal medical therapy according to current standards. Moreover, the continued relevance of physical signs in a population of patients with atrial fibrillation has not previously been addressed. We, therefore, assessed the prognostic value of 4 physical examination findings (i.e., peripheral edema, jugular venous distension, a third heart sound, and pulmonary rales) in a contemporary cohort of patients with left ventricular systolic dysfunction and a history of nonpermanent atrial fibrillation.
A post-hoc analysis was conducted on all patients enrolled in the AF-CHF (Atrial Fibrillation and Congestive Heart Failure) trial. The study protocol has been previously described (15). Briefly, the AF-CHF trial was an international multicenter trial sponsored by the Canadian Institutes of Health Research that enrolled 1,376 patients between May 2001 and June 2005 with CHF and nonpermanent atrial fibrillation. Patients were required to have a left ventricular ejection fraction (LVEF) ≤35%, and New York Heart Association (NYHA) functional class II to IV symptoms within 6 months of randomization, or functional class I symptoms if the LVEF was ≤25% or if the patient was hospitalized for CHF in the previous 6 months. Patients were randomized 1:1 to rhythm- or rate-control treatment strategies for atrial fibrillation. No treatment differences were observed with respect to the primary outcome (i.e., cardiovascular mortality) and main secondary outcomes (e.g., all-cause mortality, heart failure–related hospitalization, stroke, quality of life). The study protocol was approved by each participating center's institutional review board and all patients provided written informed consent.
The maximum tolerated dose of beta-blockers (i.e., carvedilol, bisoprolol, or metoprolol) and angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARB) was recommended in all patients. In addition, spironolactone was recommended in patients on loop diuretics with NYHA functional class III or IV symptoms. Patients randomized to rhythm-control received amiodarone as the initial antiarrhythmic drug. Targeted heart rates for rate-control were <80 beats/min at rest and <110 beats/min during 6-min walk tests. Anticoagulation was recommended in both treatment arms.
Baseline variables and physical examination
A baseline physical examination was performed by local site investigators and included an assessment for peripheral edema (none, mild [+1] or marked [≥+2]; dichotomized as present [≥+1] or absent), jugular venous distension (present or absent), a third heart sound (present or absent), and rales on pulmonary auscultation (present or absent). Additional baseline variables collected included patient demographics (e.g., age at randomization, ethnic background, sex, body mass index), type of heart disease, comorbidities (e.g., coronary artery disease, hypertension, diabetes, stroke or transient ischemic attacks), pharmacological therapy (e.g., ACE inhibitors, ARBs, beta-blockers, diuretics, lipid lowering drugs, digoxin, calcium channel blockers, and oral anticoagulants), echocardiographic parameters (e.g., left atrial size, LVEF, mitral regurgitation grade), electrocardiographic indices (e.g., heart rate, QRS duration, QTc), serum markers (e.g., electrolytes, creatinine, complete blood count), NYHA functional class (I to II vs. III to IV), baseline rhythm, and randomized treatment arm (rate vs. rhythm control).
Patients were followed at 3 weeks, 4 months, every 4 months until 4 years, and then biannually until 6 years. As in the main trial, the primary outcome for the current study was cardiovascular mortality. Secondary outcomes included all-cause mortality, heart failure–related mortality, sudden cardiac death, and heart failure–related hospitalization. All events were classified by an independent adjudicating committee blinded to the treatment assignment.
Continuous variables are summarized by mean ± SD. Categorical variables are represented by frequencies and percentages. Two-group baseline comparisons were performed by Student t or chi-square tests where appropriate. Event-free survival for mortality and hospitalization outcomes was plotted using the Kaplan-Meier method, with comparisons by log-rank statistics. Time 0 was defined as the time of physical examination. Univariate and multivariate Cox regression models were created for each outcome variable (i.e., cardiovascular mortality, all-cause mortality, heart failure–related death, sudden cardiac death, and heart failure–related hospitalization). The covariates considered in the multivariate regression models are listed in Table 1.
Two approaches to multivariate modeling were used. The first consisted of building separate multivariate models for each congestive sign, with the congestive sign of interest forced in the model. The second approach considered all 4 congestive signs as potential covariates within a single multivariate model. With both approaches, randomization to rate- versus rhythm-control therapy was forced into the model. Other baseline covariates that were significant at the 0.2 level in univariate analyses were included in stepwise multivariate Cox regression models. To account for missing data, a multiple imputation analysis was performed as the main approach, with a secondary complete-case approach as a sensitivity analysis. Given that the 2 approaches yielded similar results, data from the main approach are presented. Imputed data sets were generated using IVEware version 0.1, a user-contributed implementation of sequential generalized regression techniques in SAS (SAS Institute, Cary, North Carolina).
Two-tailed p values <0.05 were considered statistically significant. Statistical testing was performed using SAS software Version 9.2 (SAS Institute).
A total of 1,376 patients, mean age 67 ± 11 years, 81% male, were enrolled. Data on peripheral edema was available in 1,374 (99.9%), jugular venous distension in 1,369 (99.5%), a third heart sound in 1,373 (99.8%), and pulmonary rales in 1,375 (99.9%) patients. Complete data on all 4 congestive signs were available in 1,369 (99.5%) patients. The incidence of peripheral edema was 30.9%, jugular venous distension 21.7%, a third heart sound 15.0%, and pulmonary rales 13.0%, with no differences between rhythm- versus rate-control treatment strategies. Baseline characteristics in all patients and according to whether they had positive findings on physical examination are summarized in Table 1. In general, patients with congestive signs were older, had higher serum creatinine levels, faster heart rates, and were more likely to have ischemic heart disease and markers of advanced heart failure such as NYHA functional class III or IV symptoms and diuretic therapy. Atrial fibrillation at baseline, hypertension, and digoxin use were more common among patients with peripheral edema and/or rales. Patients with rales were less likely to be treated with beta-blockers.
Prevalence of cardiovascular outcomes
Over a mean follow-up of 37 ± 19 months, 445 (32.3%) patients died and 347 (25.2%) required at least 1 hospitalization for heart failure. Cardiovascular mortality accounted for 357 (80.2%) deaths, which were subclassified as sudden cardiac death of presumed arrhythmic etiology in 159 and heart failure–related in 130 patients. The prevalence of each cardiovascular outcome is summarized in Table 2 according to whether or not congestive signs were present on physical examination. Freedom from cardiovascular mortality is plotted in Figure 1 according to the presence or absence of jugular venous distension (Fig. 1A), peripheral edema (Fig. 1B), pulmonary rales (Fig. 1C), and a third heart sound (Fig. 1D). Cardiovascular outcomes according to whether patients had at least 1 positive congestive sign are plotted in Figure 2.
Cardiovascular outcomes associated with each physical sign separately
In univariate analyses, peripheral edema (hazard ratio [HR]: 1.648, 95% confidence interval [CI]: 1.331 to 2.039), p <0.0001), jugular venous distension (HR: 1.518, 95% CI: 1.201 to 1.920), p = 0.0005), a third heart sound (HR: 1.473, 95% CI: 1.134 to 1.913), p = 0.0037), and pulmonary rales (HR: 1.923, 95% CI: 1.480 to 2.497), p <0.0001) were all associated with increased risk of cardiovascular mortality. Table 3 summarizes the univariate and multivariate hazard ratios for all cardiovascular outcomes associated with each congestive sign considered separately. In addition to cardiovascular mortality, peripheral edema was independently associated with all-cause mortality and heart failure–related death. Jugular venous distension was independently associated with heart failure–related mortality but not the other outcomes. No cardiovascular outcome was independently predicted by the presence of a third heart sound. In contrast, pulmonary rales was independently associated with all mortality outcomes except sudden death. In addition, pulmonary rales was the only congestive sign independently associated with heart failure–related hospitalization.
Multivariate models with all 4 physical signs
In multivariate analyses that included all 4 congestive signs as potential covariates, peripheral edema (HR: 1.25; 95% CI: 1.00 to 1.57; p = 0.0482) and pulmonary rales (HR: 1.41; 95% CI: 1.08 to 1.86; p = 0.0133) remained independent predictors of cardiovascular mortality. Other factors independently associated with cardiovascular mortality were older age (HR: 1.019 per year; 95% CI: 1.008 to 1.031; p = 0.0007), lower systolic blood pressure (HR: 0.992 per mm Hg; 95% CI: 0.986 to 0.998; p = 0.0105), higher serum creatinine level (HR: 1.003 per mmol/l; 95% CI: 1.001 to 1.005; p = 0.0003), NYHA functional class III or IV symptoms (HR: 1.30; 95% CI: 1.04 to 1.62; p = 0.0234), ischemic heart disease (HR: 1.75; 95% CI: 1.39 to 2.21; p < 0.0001), diuretics (HR: 1.49; 95% CI: 1.06 to 2.10; p = 0.0212), and aldosterone antagonists (HR: 1.39; 95% CI: 1.11 to 1.72; p = 0.0034).
Peripheral edema was independently associated with all-cause mortality and heart failure–related death. Pulmonary rales was independently associated with heart failure–related death and hospitalization. No cardiovascular outcome was independently predicted by jugular venous distension or presence of a third heart sound.
Sophisticated laboratory techniques provide the contemporary clinician with a wealth of diagnostic and prognostic information that calls into question the residual value of a thoughtful physical examination. Our analyses in this cohort of patients with CHF and history of atrial fibrillation suggest that each of the 4 congestive signs on physical examination continue to provide relevant prognostic information. Peripheral edema, jugular venous distension, a third heart sound, and pulmonary edema were each associated with increased cardiovascular mortality, all-cause mortality, and heart failure–related death. In addition, peripheral edema was significantly associated with sudden cardiac death, and jugular venous distension, a third heart sound, and pulmonary edema with heart failure–related hospitalizations. In multivariate analyses that included standard clinical, electrocardiographic, and echocardiographic parameters: 1) peripheral edema and pulmonary rales remained independent predictors of all-cause and cardiovascular mortality, associated with a 2-fold increased risk of heart failure–related death; 2) pulmonary rales independently predicted hospitalizations for heart failure; and 3) jugular venous distension was associated with a 48% increased risk of heart failure–related death.
The few studies that addressed the prognostic value of congestive signs in the setting of acute heart failure have yielded consistent results (16,17). The presence of an S3 has been associated with mortality in patients listed for heart transplantation (18). In patients recently hospitalized with decompensated heart failure and NYHA functional class IV symptoms, the absence of any congestive sign (defined by jugular venous distension, edema, orthopnea, weight gain, or need for increased diuretic dose) was associated with superior 2-year survival (19). Conversely, rales or peripheral edema 1 week after hospital discharge predicted adverse cardiovascular outcomes at 1 year of follow-up (20). A post-hoc analysis of the EVEREST (Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study With Tolvaptan) found that patients with a higher composite congestion score experienced high mortality and readmission rates (21).
The prognostic value of the physical examination in the outpatient setting, as performed in the current study, is more controversial. A retrospective analysis of the SOLVD (Studies Of Left Ventricular Dysfunction) reported that jugular venous distension and a third heart sound were associated with increased cardiovascular mortality (22); the presence of peripheral edema and rales was not assessed. Moreover, this analysis was based on a population enrolled over 25 years ago in whom less than 10% received beta-blockers, ACE inhibitors were randomly allocated, and 9% were treated with potassium sparing diuretics. In contrast, a substudy from the Digitalis Investigator Group created a global congestion score that combined the 4 physical examination signs assessed in our study along with symptoms and radiological findings (23). This score independently predicted mortality at 36 months in patients with systolic dysfunction on ACE inhibitors. Because the individual contribution of physical signs was not independently analyzed, the prognostic value of each cannot be surmised. Moreover, concomitant use of beta-blockers and aldosterone antagonists was not reported.
A community-based study reported that a combination of right- and left-sided congestive signs predicted cardiovascular mortality in patients with CHF and preserved or depressed left ventricular function (24). Analyses were not specifically conducted in the subgroup of patients (45%) with left ventricular systolic dysfunction. Compared to our study population, patients had less advanced heart failure, as reflected by a higher LVEF (34 ± 6%) and lower prevalence of congestive signs (e.g., rales 4%). Only 38% received an ACE inhibitor or ARB and 45% a beta-blocker. More recently, a retrospective analysis of 2,647 patients with NYHA functional class III to IV symptoms enrolled in the CIBIS-II (Cardiac Insufficiency Bisoprolol Study-II) trial found that jugular venous distension and ascites were independently associated with renal dysfunction and all-cause mortality (25). Peripheral edema and jugular venous distension were also associated with heart failure deaths and cardiovascular hospitalizations. Importantly, the prognostic value of pulmonary rales was not assessed. As in other studies, beta-blockers and aldosterone antagonists were underprescribed according to current standards (50% and 10% of the population, respectively) (25).
In our more contemporary patient population, 86% received ACE inhibitors, 79% beta-blockers, and 45% aldosterone antagonists. All 4 congestive signs (i.e., peripheral edema, jugular venous distension, the presence of a third heart sound, and pulmonary rales) were associated with increased cardiovascular mortality. Interestingly, physical examination findings associated with edema (i.e., peripheral edema and pulmonary rales) remained more powerful independent predictors of cardiovascular outcomes than jugular venous distension and a third heart sound. Increased hydrostatic pressure in right heart cavities is transmitted to systemic veins and capillaries, leading to venous congestion (including jugular distension), peripheral edema, and/or ascites. While increased central venous pressures by catheterization have been associated with worsening renal function and prognosis in patients with CHF (26–28), jugular venous distension on physical examination frequently underestimates right atrial pressures (29). In contrast, peripheral edema may be considered a more advanced marker of heart failure in that it represents the inability of the cardiovascular system to tolerate the increase in intravascular pressure once venous compliance has been exceeded.
On the left side of the circulation, elevated filling pressures have been associated with adverse outcomes in patients with CHF (18,30,31). The sympathetic nervous system is activated as result of myocardial stretch (32) and increased left atrial pressures are transmitted to the pulmonary venous system (33). The capillary barrier responds to chronic increases in pressure by reducing its permeability, with consequent interstitial edema. In advanced heart failure, compensatory mechanisms are overwhelmed, resulting in pulmonary congestion and rales on physical examination.
In contrast, a third heart sound, which occurs in early diastole during the rapid left ventricular filling phase, reflects increased left ventricular filling pressures and decreased compliance (34,35). It has been associated with mortality in patients listed for heart transplantation (18). However, concerns have been raised over reliability and interobserver agreement (36–38). In addition, of all the physical signs assessed in our study population, it may be the most likely to be confounded by irregular heart rates in atrial fibrillation. Moreover, a poor correlation between a third heart sound and filling pressures by catheterization has been reported (39). The observed lack of independent association between a third heart sound and cardiovascular outcomes in our study may reflect these observations.
The study is retrospective and subject to limitations inherent to observational research. While analyses control for baseline imbalances and potential confounders, they cannot adjust for unknown or unmeasured variables. Outcome variables were subject to standardized definitions and adjudicated by a blinded committee of experts. In contrast, the physical examination was performed without investigator blinding or independent adjudication. As such, findings may be influenced by extraneous factors such as clinical acumen and knowledge of the patient's medical chart. Nevertheless, the physical examination reflects standard clinical practice in an adult population. Finally, while our analyses controlled for underlying rhythm at the time of assessment, all patients had a history of nonpermanent atrial fibrillation. As such, results may not be generalizable to a patient population with a low prevalence of atrial fibrillation.
In patients with CHF and a history of atrial fibrillation, all 4 congestive signs on physical examination (peripheral edema, jugular venous distension, the presence of a third heart sound and pulmonary rales) were associated with an increased risk of cardiovascular mortality. This increased mortality was predominantly driven by an excess in heart failure–related deaths. In addition, congestive signs of edema on physical examination (i.e., peripheral edema and pulmonary rales) independently predicted all-cause and cardiovascular mortality. The association between jugular venous distension and heart failure–related mortality was likewise significant but of lesser magnitude. Taken together, these results suggest that the physical examination in modern times remains relevant in providing important prognostic information above and beyond standard clinical, electrocardiographic, and echocardiographic parameters.
The AF-CHF trial was funded by the Canadian Institutes of Health Research (MCT-41552). Dr. Ducharme holds a senior research grant from “le Fond de Recherche du Québec en Santé” (FRQS). Dr. White holds the Caroline and Richard Renault Chair in Heart Failure of the Montreal Heart Institute. Dr. Khairy is supported by a Canada Research Chair in Electrophysiology and Adult Congenital Heart Disease. Dr. Caldentey holds a fellowship grant from the Spanish Society of Cardiology. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- angiotensin receptor blocker
- chronic heart failure
- confidence interval
- hazard ratio
- left ventricular ejection fraction
- New York Heart Association
- Received August 6, 2013.
- Revision received October 1, 2013.
- Accepted October 3, 2013.
- American College of Cardiology Foundation
- Mancini D.M.,
- Eisen H.,
- Kussmaul W.,
- Mull R.,
- Edmunds L.H. Jr..,
- Wilson J.R.
- Solomon S.D.,
- Anavekar N.,
- Skali H.,
- et al.
- Pinamonti B.,
- Di Lenarda A.,
- Sinagra G.,
- Camerini F.
- de Groote P.,
- Millaire A.,
- Foucher-Hossein C.,
- et al.
- Tabrizi F.,
- Englund A.,
- Rosenqvist M.,
- Wallentin L.,
- Stenestrand U.
- Lee W.H.,
- Packer M.
- Drazner M.H.,
- Hellkamp A.S.,
- Leier C.V.,
- et al.
- Nohria A.,
- Tsang S.W.,
- Fang J.C.,
- et al.
- Ambrosy A.P.,
- Pang P.S.,
- Khan S.,
- et al.
- Damman K.,
- van Deursen V.M.,
- Navis G.,
- et al.
- Mullens W.,
- Abrahams Z.,
- Francis G.S.,
- et al.
- Bock J.S.,
- Gottlieb S.S.
- Aaronson K.D.,
- Schwartz J.S.,
- Chen T.M.,
- Wong K.L.,
- Goin J.E.,
- Mancini D.M.
- Azevedo E.R.,
- Newton G.E.,
- Floras J.S.,
- Parker J.D.