Journal of Cardiac Failure Vol. 7 No. 2 2001
Clinical Investigations
Hemodynamic Profiles of Advanced Heart Failure: Association With Clinical Characteristics and Long-term Outcomes MONICA R. SHAH, MD,* VIC HASSELBLAD, PhD,* SANDRA S. STINNETT, DrPH,* MIHAI GHEORGHIADE, MD, FACC,† KARL SWEDBERG, MD, FACC,‡ ROBERT M. CALIFF, MD, FACC,* CHRISTOPHER M. O’CONNOR, MD, FACC* Durham, North Carolina; Chicago, Illinois; Göteborg, Sweden
ABSTRACT Background: Classifying patients with advanced congestive heart failure (CHF) by baseline measures of congestion and perfusion has been used to estimate hemodynamic status and to select and titrate therapy. We describe clinical characteristics of 4 hemodynamic profiles— wet/cold, wet/warm, dry/cold, and dry/warm—in patients with advanced CHF and assess relations between symptoms, physical signs, and outcomes with each profile. Methods and Results: We retrospectively assessed baseline symptoms, physicalexamination variables, and 1-year outcomes of 440 patients in a randomized trial. With univariable and multivariable logistic regression, we examined relations of physicalexamination variables to hemodynamic profiles. We also assessed the rates of death and death or readmission by profile. Severity of CHF symptoms did not predict the wet-versus-dry profile or cold-versus-warm status, despite significant differences in hemodynamics among groups. Of the physical-examination variables, only a lower proportional pulse pressure was a significant multivariable predictor of the wet category. Among wet patients (n ⫽ 348), this same variable was the only significant multivariable predictor of the cold category. For dry patients (n ⫽ 92), the cold category was predicted in multivariable analysis by supine heart rate and hepatomegaly. Survival was similar among profiles: wet/cold, 54.2% (n ⫽ 91); wet/warm, 58.3% (n ⫽ 105); dry/cold, 78.9% (n ⫽ 15); and dry/warm, 67.1%, P ⫽ .13 (n ⫽ 49). Event-free survival also was similar among profiles: wet/cold, 22.0% (n ⫽ 37); wet/warm, 29.4% (n ⫽ 53); dry/cold, 42.1% (n ⫽ 8); and dry/warm, 31.5%, P ⫽ .44 (n ⫽ 23). Conclusions: The patient’s history and physical examination alone may lead to inaccurate estimation of hemodynamic status and thus suboptimal management for patients with advanced CHF. Key words: congestive heart failure, hemodynamics, prognosis, outcomes.
From the *Duke Clinical Research Institute, Durham, North Carolina; †Northwestern Medical Center, Chicago, Illinois; and ‡Sahlgrenska Hospital, Göteborg, Sweden. Supported in part by Glaxo-Wellcome, Research Triangle Park, North Carolina. Manuscript received November 17, 2000; revised manuscript received February 9, 2001; revised manuscript accepted February 12, 2001. Reprint requests: Monica R. Shah, MD, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715. Copyright © 2001 by Churchill Livingstone威 1071-9164/01/0702-0002$35.00/0 doi:10.1054/jcaf.2001.24131
105
106
Journal of Cardiac Failure Vol. 7 No. 2 June 2001
The patient’s history and physical examination are classic tools in the diagnosis and management of patients with heart disease. Patients with advanced congestive heart failure (CHF) are a heterogenous group and may present with different combinations of symptoms, physical signs, and hemodynamic derangements. These patients often have varying degrees of dyspnea, orthopnea, and fatigue; a range of elevations in filling pressures; and decreases in cardiac performance. Classifying patients with advanced CHF by evidence of congestion and adequate perfusion has been hypothesized to be of value in estimating hemodynamic status and in selecting and titrating therapy. The hemodynamic profiles that may assist in identifying true hemodynamic status include the following: 1) hypervolemia and hypoperfusion (wet/cold), 2) hypervolemia and adequate perfusion (wet/warm), 3) euvolemia and hypoperfusion (dry/cold), and 4) euvolemia and adequate perfusion (dry/warm) (1–3). The 4 hemodynamic profiles were developed on the basis of clinical observations and information from the patient’s history and physical examination. The first branch point of the advanced CHF-management algorithm classifies patients as either wet or dry based on evidence of volume overload (2). After fluid status has been ascertained, findings then are used to classify patients as either warm or cold. Treatment approaches for each category of patients have been recommended (3) based on the assumption that clinical assessment can provide an accurate estimate of hemodynamics. However, few data have described how well the hemodynamic profiles correlate with actual hemodynamic measurements and whether the categories carry independent prognostic information. The objectives of this analysis are to describe the baseline features associated with each hemodynamic profile and to assess the relations between symptoms and physical signs for each category. An additional objective is to determine the outcomes (rates of death and death or readmission) associated with each hemodynamic profile.
Methods Study Procedures The Flolan International Randomized Survival Trial (FIRST) population consisted of 471 patients with New York Heart Association (NYHA) class IIIb or IV heart failure, an ejection fraction less than 25%, and severe symptoms despite maximum medical therapy. The details of the trial have been reported (4). Briefly, patients were required to be on loop diuretics, digitalis glycosides, and angiotensin-converting enzyme inhibitors. Patients who were candidates for cardiac transplant were not eligible for enrollment. After providing informed
consent, patients who met the clinical criteria underwent a battery of noninvasive tests, including a detailed physical examination, evaluation of heart failure symptoms, assessment of the Yale Dyspnea-Fatigue Index (YDFI), determination of NYHA class, and a 6-minute walk test. Patients then underwent pulmonary artery catheterization within 8 hours. To be eligible for the trial, patients were required to have a cardiac index 2.2 L/min/m2 or less and a pulmonary capillary wedge pressure (PCWP) 15 mm Hg or greater. Patients who met hemodynamic criteria were randomly assigned to receive epoprostenol or placebo, with other standard therapy. The evaluation of CHF symptoms included assessments of dyspnea, orthopnea, and fatigue. These symptoms were rated as 0 (absent) or 1 (present). The composite CHF symptom score ranged from 0 to 3. Determination of NYHA class was made using standard definitions (5). Heart failure symptoms also were evaluated with the YDFI (6), which evaluates the level of functional impairment and the magnitude and pace of tasks that cause dyspnea and fatigue. The composite YDFI score ranges from 0 (severely limited) to 12 (no limitation). The 6-minute walk test was conducted as described by Guyatt et al (7). Physical examination included standard evaluation of supine and standing blood pressures and heart rates. Proportional pulse pressure was calculated by using supine blood pressure measurements and the formula Systolic Blood Pressure – Diastolic Blood Pressure/ Systolic Blood Pressure. Jugular venous distension was measured as less than or equal to 6 or greater than 6 mm Hg as patients reclined at 30° to 45°. Rales, third heart sounds, hepatomegaly, and edema were noted as either present or absent on physical examination. Baseline clinical data were recorded on standard forms for all patients. Hemodynamic Data Collection Heart rate, mean arterial pressure, pulmonary artery systolic pressure, pulmonary artery diastolic pressure, mean pulmonary artery pressure, right atrial pressure, cardiac index and output, and PCWP were measured and recorded at baseline and at the end of the 24-hour drug titration period in patients randomized to epoprostenol. Only baseline measures were obtained from patients randomized to conventional therapy; the catheter was removed after the first hemodynamic readings. Hemodynamic Profiles The 4 hemodynamic profiles defined for acute myocardial infarction and chronic CHF were adapted to the FIRST population (1,8). Earlier studies used a PCWP of greater than 18 mm Hg to define congestion, the wet
Clinical Characteristics Versus Hemodynamic Status
category, and a cardiac index of 2.2 L/min/m2 or less to identify inadequate perfusion, the cold category. We modified this classification to obtain meaningful numbers of patients in each category. The 4 hemodynamic profiles were defined as follows: 1) wet/cold, a PCWP 20 mm Hg or greater and a cardiac index less than 1.8 L/min/m2; 2) wet/warm, a PCWP 20 mm Hg or greater and a cardiac index 1.8 L/min/m2 or greater; 3) dry/cold, a PCWP less than 20 mm Hg and a cardiac index less than 1.8 L/min/m2; and 4) dry/warm, a PCWP less than 20 mm Hg and a cardiac index 1.8 L/min/m2 or greater. Statistical Methods Of the 471 patients randomized, we excluded 31 patients from the analysis because they lacked data for baseline symptoms, physical-examination findings, or hemodynamic measures. Thus, descriptive statistics of these variables reflect 440 patients. Medians with interquartile ranges were used to describe continuous variables. Frequencies and percentages were used to characterize categorical variables. The differences in median hemodynamic values among groups were assessed with the Kruskal-Wallis test. The baseline CHF symptom score and baseline NYHA class were evaluated as categorical variables. The baseline YDFI and 6-minute walk test were assessed as continuous variables. Physical-examination findings that were assessed as continuous variables included heart rate and systolic and diastolic blood pressures; all other such findings were assessed as categorical variables. To evaluate the relations between CHF symptom score, YDFI, NYHA class, and the 4 hemodynamic profiles, we first performed univariable regression analysis of the relations between symptoms and the hemodynamic categories of wet versus dry. We then performed
O Shah et al 107
univariable regression analyses to evaluate the ability of symptoms in each of the wet and dry populations to predict the warm and cold categories. Significant physical examination signs then were identified as univariable predictors of the wet category; from these, we created a multivariable model to identify the best predictors of the wet category. Univariable regression analyses were repeated in the wet population to identify the physical-examination findings that predicted the warm or cold categories. A multivariable model was then produced. A similar set of regression of analyses was performed in the dry population. The Kaplan-Meier life-table method was used to calculate survival and readmission-free survival for patients in the 4 hemodynamic profiles. The survival curves were compared using the log-rank test.
Results The most common hemodynamic profile in the FIRST study population was wet/warm (n ⫽ 180) followed by wet/cold (n ⫽ 168, 38.2%), and dry/warm (n ⫽ 73, 16.6%). Only a small percentage of patients met the criteria for the dry/cold profile (n ⫽ 19, 4.3%). There were no significant differences in age, sex, or race among the hemodynamic profiles (Table 1) or in the origin of heart failure, ejection fraction, or NYHA classification. Patients with the dry/warm clinical profile had the same median CHF symptom score and YDFI as patients in the more compromised hemodynamic categories. The median distance achieved in a 6-minute walk was actually lower among patients in the dry/warm category compared with the other groups. Baseline hemodynamic measurements of the 4 groups are displayed in Table 2. The median PCWP ranged from
Table 1. Baseline Characteristics and Symptoms by Hemodynamic Profile Wet/Cold (n ⫽ 168) Age (yr) Male sex Race White Other Origin of heart failure Ischemic Nonischemic Ejection fraction (%) Symptoms NYHA class IV heart failure CHF symptom score Yale Dyspnea-Fatigue Index Meters in 6-minute walk NYHA, New York Heart Association.
Wet/Warm (n ⫽ 180)
Dry/Cold (n ⫽ 19)
Dry/Warm (n ⫽ 73)
64 (58–71) 132 (79%)
65 (60–71) 136 (76%)
66 (58–74) 16 (84%)
67 (60–72) 49 (67%)
134 (80%) 30 (20%)
153 (85%) 25 (15%)
15 (79%) 4 (21%)
59 (81%) 14 (19.2%)
117 (70%) 51 (30%) 16.9 (13–20)
125 (69%) 55 (31%) 18.5 (15–22)
10 (53%) 9 (47%) 17 (14–21)
46 (64%) 26 (36%) 20 (16–23)
103 (62%) 3 (3–3) 2 (0–3) 183 (85–266)
107 (60%) 3 (3–3) 2 (0–3) 186 (58–270)
11 (58%) 3 (3–3) 2 (1–3) 150 (50–289)
42 (58%) 3 (3–3) 2 (0–3) 137 (28–256)
108 Journal of Cardiac Failure Vol. 7 No. 2 June 2001 Table 2. Baseline Hemodynamic Measures by Hemodynamic Profile Wet/Cold (n ⫽ 168) Right atrial pressure (mm Hg) Mean PAP (mm Hg) PCWP (mm Hg) Systemic vascular resistance (mm Hg/L/min) Cardiac index (L/min/m2) Cardiac output (L/min) Heart rate (supine)
Wet/Warm (n ⫽ 180)
Dry/Cold (n ⫽ 19)
Dry/Warm (n ⫽ 73)
P Value
13 (10–19) 41.5 (35–46.5) 29 (24–34.5) 1986 (1557–2509)
10 (7–15) 40 (35–44) 27 (22–30) 1449 (1230–1702)
10.5 (7–14) 28 (23–32) 17 (16–18) 2163 (1869–2545)
7 (4–10) 27.5 (24–33) 17 (16–18) 1429 (1305–1729)
.0001 .0001 NC .0001
1.5 (1.3–1.7) 2.7 (2.3–3.0) 88 (76–99)
2.1 (1.9–2.2) 3.8 (3.5–4.2) 86 (76–95)
1.6 (1.4–1.7) 3.0 (2.6–3.3) 74 (68–92)
2.1 (2.0–2.3) 3.9 (3.5–4.3) 86 (76–100)
NC NC .04
NC, not calculated (because definitions of the hemodynamic profiles were based on these variables); PAP, pulmonary arterial pressure; PCWP, pulmonary capillary wedge pressure.
17 mm Hg in both of the dry profiles to 29 mm Hg in the wet/cold category. Most patients classified as wet had a PCWP of greater than 25 mm Hg. The median cardiac index ranged from 1.5 L/min/m2 in the wet/cold profile to 2.1 L/min/m2 in the warm categories. Most patients classified as warm had a cardiac index greater than 2.0 L/min/m2. Right atrial pressure differed significantly across the 4 groups (P ⫽ .0001), primarily driven by variations in this value between the wet/cold and dry/warm profiles. There also was a significant difference in the mean pulmonary artery pressure among the 4 groups (P ⫽ .0001). Patients with greater filling pressures, identified as either wet/cold or wet/warm, had median pulmonary artery pressures greater than 10 mm Hg higher than the other 2 profiles. In addition, there was a significant difference in systemic vascular resistance among the profiles (P ⫽ .0001). This measure was markedly higher in patients with less adequate perfusion, either wet/cold or dry/cold. The severity of CHF symptoms did not predict the wet-versus-dry profile, despite the significant differences in hemodynamic variables among the groups (Table 3). Furthermore, among patients identified as either wet or dry, the CHF symptom score, YDFI, 6-minute walk, and NYHA class did not predict cold-versus-warm status (Table 4). Only 2 physical-examination variables had a significant univariable association with the wet category. These included lower proportional pulse pressure (odds ratio
[OR], 1.42 for each 10% decrease; 95% confidence interval (CI), 1.09–1.87; P ⫽ .01) and the presence of hepatomegaly (OR, 1.69; 95% CI, 1.05–2.71; P ⫽ .03) (Table 5). Other classic physical signs of congestion, such as jugular venous distention, rales, and edema, did not predict the wet profile. In a multivariable model, decreasing proportional pulse pressure remained the only significant predictor of the wet category (OR, 1.39 for each 10% decrease; 95% CI, 1.06–1.83; P ⫽ .02). Among patients who were categorized as wet (n ⫽ 348), there were only 2 significant univariable predictors of the cold category (Table 6). These included lower supine systolic blood pressure (OR, 1.21 for each 10-mm Hg decrease; 95% CI, 1.06–1.38; P ⫽ .006) and lower proportional pulse pressure (OR, 1.61 for each 10% decrease; 95% CI, 1.24–2.09; P < .001). In multivariable analysis, proportional pulse pressure remained a significant predictor of the cold profile (OR, 1.58 for each 10% decrease; 95% CI, 1.22–2.06; P ⫽ .0005). For patients classified as dry (n ⫽ 92), the cold category was predicted by hepatomegaly (OR, 3.97; 95% CI, 1.38–11.45; P ⫽ .01) and a slower supine heart rate (OR, 0.67 for each 10-bpm increase; 95% CI, 0.48–0.95; P ⫽ .02) (Table 7). These 2 variables remained significant predictors in multivariable analysis. Although hemodynamic variables differed significantly among the 4 profiles, their survival was similar: wet/cold, 54.2% (n ⫽ 91); wet/warm, 58.3% (n ⫽ 105); dry/cold, 78.9% (n ⫽ 15); and dry/warm, 67.1% (n ⫽ 49,
Table 3. Univariable Relations of Baseline Symptoms to Wet Hemodynamic Status Odds Ratio (95% CI) CHF symptom score Meters achieved in 6-minute walk* NYHA class Yale Dyspnea-Fatigue Index*
1.83 1.00 1.14 1.03
(0.86–3.89) (0.99–1.00) (0.71–1.81) (0.90–1.16)
Wald 2 2.48 1.44 0.29 0.15
CI, confidence interval; CHF, congestive heart failure; NYHA, New York Heart Association. *Per 1-unit increase.
P Value .12 .23 .59 .70
Clinical Characteristics Versus Hemodynamic Status
O Shah et al 109
Table 4. Univariable Relations of Baseline Symptoms to Cold Hemodynamic Status Odds Ratio (95% CI) Wet patients CHF symptom score NYHA class Yale Dyspnea-Fatigue Index* Meters achieved in 6-minute walk* Dry patients Yale Dyspnea-Fatigue Index* CHF symptom score Meters achieved in 6-minute walk* NYHA class
Wald 2
P Value
2.4 (0.97–5.94) 1.1 (0.70–1.67) 0.9 (0.84–1.06) 1.00 (0.99–1.00)
3.9 0.29 0.13 0.001
.06 .72 .30 .97
1.2 (0.92–1.56) 0.40 (0.10–1.53) 1.00 (0.99–1.01) 1.1 (0.37–2.82)
1.82 1.79 0.30 0.001
.18 .18 .59 .98
CI, confidence interval; CHF, congestive heart failure; NYHA, New York Heart Association. *Per 1-unit increase.
P ⫽ .13) by log-rank test (Fig. 1). In a secondary analysis, patients in the wet/cold category had a significantly higher mortality rate than did those in the dry/warm category (P ⫽ .05). There was no significant difference in survival between the dry/warm versus the wet/warm profiles or the wet/warm versus the wet/cold profiles. Analysis of the rate of death or readmission showed no significant differences among the 4 groups: wet/cold, 22.0% (n⫽37); wet/warm, 29.4% (n⫽53); dry/cold, 42.1% (n⫽8); and dry/warm, 31.5% (n⫽23, P ⫽ .44) by log-rank test (Fig. 2).
Discussion The patient’s history and physical examination are the classic noninvasive methods used to determine hemodynamic status, and to select and titrate therapy, in patients with NYHA class IV heart failure. We had hypothesized that clinical features could predict hemodynamic status and that a given hemodynamic profile would carry
independent prognostic significance in such patients. The severity of symptoms on presentation did not predict hemodynamic status, however, and more severe dyspnea, orthopnea, or fatigue did not correlate with the wet hemodynamic profile. In addition, among patients who had a PCWP 20 mm Hg or greater (the wet category), symptoms alone could not distinguish those with severe hypoperfusion (the cold category). Physical examination findings also were poor predictors of hemodynamic status. Among the classic signs used to diagnose fluid overload, such as jugular venous distension, rales, and edema, the only variables predictive of a wet hemodynamic profile were decreasing proportional pulse pressure and the presence of hepatomegaly. In addition, among patients identified as wet, the only physical signs significantly associated with severe hypoperfusion were decreasing supine systolic blood pressure and decreasing proportional pulse pressure. Among patients with a dry hemodynamic profile, slower heart rate and the presence of hepatomegaly significantly predicted severe hypoperfusion, or the cold
Table 5. Relations of Baseline Physical-Examination Findings to Wet Hemodynamic Status Odds Ratio (95% CI) Univariable analysis Proportional pulse pressure* Hepatomegaly (v none) Rales (v none) Jugular venous distention (>6 v ⱕ6 cm) Supine diastolic BP (mm Hg)† Supine heart rate (bpm)† S4 (v S3) Supine systolic BP (mm Hg)* Edema (v none) Multivariable analysis Proportional pulse pressure* Hepatomegaly (v none) CI, confidence interval; BP, blood pressure. *Per 10-unit decrease. † Per 10-unit increase.
1.42 1.69 1.45 1.40 1.14 1.08 1.38 1.06 1.21
Wald 2
P Value
(1.09–1.87) (1.05–2.71) (0.92–2.31) (0.88–2.22) (0.92–1.41) (0.94–1.26) (0.77–2.47) (0.93–1.21) (0.76–1.92)
6.57 4.70 2.50 2.00 1.36 1.30 1.19 0.72 0.64
.01 .03 .11 .15 .24 .25 .28 .39 .42
1.39 (1.06–1.83) 1.56 (0.96–2.54)
5.57 3.27
.02 .07
110
Journal of Cardiac Failure Vol. 7 No. 2 June 2001 Table 6. Relations of Baseline Physical Examination Findings to Cold Hemodynamic Status, Wet Patients Odds Ratio (95% CI)
Univariable analysis Proportional pulse pressure* Supine systolic BP (mm Hg)* Hepatomegaly (v none) S3 (v none) Supine heart rate (bpm)† S4 (v S3) Edema (v none) Supine diastolic BP (mm Hg)† Jugular venous distension (>6 v ⱕ6 cm) Rales (v none) Multivariable analysis Proportional pulse pressure* Hepatomegaly (v none)
1.61 1.21 1.48 1.36 1.07 0.27 1.22 1.05 1.12 0.96
Wald 2
P Value
(1.24–2.09) (1.06–1.38) (0.97–2.26) (0.83–2.22) (0.94–1.22) (0.53–1.44) (0.79–1.87) (0.87–1.27) (0.73–1.69) (0.62–1.49)
12.98 7.58 3.30 1.53 0.99 0.88 0.81 0.22 0.22 0.02
.008 .006 .07 .22 .31 .60 .37 .63 .64 .86
1.58 (1.22–2.06) 1.39 (0.90–2.15)
11.99 2.21
.0005 .14
CI, confidence interval; BP, blood pressure. *Per 10-unit decrease. † Per 10-unit increase.
hemodynamic profile. However, there were only 19 patients in the dry/cold subgroup. Previous Studies Earlier work has yielded conflicting information about the reliability of symptoms and physical signs in identifying elevations in ventricular filling pressures and decreases in cardiac output. Various relationships have been reported between physical signs and hemodynamic measurements. In an analysis of 50 patients, Stevenson and Perloff (1) showed that a combination of rales, jugular venous distension, and edema identified only 42% of patients with a PCWP 22 mm Hg or greater. They also showed that a proportional pulse pressure of 25% or less correlated well with cardiac index (r ⫽ 0.82). In another analysis of 52 patients, Butman et al (9) noted that the presence of both jugular venous distension and
hepatojugular reflux was highly sensitive (81%) in the identification of patients with a PCWP 18 mm Hg or greater. In an evaluation of 52 patients referred for cardiac transplantation, however, Chakko et al (10) reported that orthopnea, edema, jugular venous distension, rales, and an S3 gallop did not predict elevated left ventricular filling pressures (10). Finally, Eisenberg et al (11) prospectively assessed the ability of physicians to predict hemodynamics in 97 critically ill patients with a variety of diagnoses. In this analysis, clinicians correctly estimated the PCWP and cardiac output only 30% and 51% of the time, respectively. The data in this report suggest that the physicalexamination findings most predictive of hemodynamic status in patients with advanced CHF are objective measures, such as supine blood pressure and proportional pulse pressure. These signs may be more reliable because they are more quantitative and depend less on skill,
Table 7. Relations of Baseline Physical Examination Findings to Cold Hemodynamic Status, Dry Patients Odds Ratio (95% CI) Univariable analysis Hepatomegaly Supine heart rate (bpm)* Supine diastolic BP (mm Hg)* Jugular venous distension (>6 v ⱕ6 cm) Supine systolic BP (mm Hg)† S3 (v none) Proportional pulse pressure† S4 (v S3) Edema (v none) Rales (v none) Multivariable analysis Hepatomegaly (v none) Supine heart rate (bpm)* CI, confidence interval; BP, blood pressure.
Wald 2
P Value
3.97 (1.38–11.45) 0.67 (0.48–0.95) 1.47 (0.91–2.36) 1.97 (0.71–5.48) 1.17 (0.88–1.56) 2.01 (0.52–7.66) 1.28 (0.67–2.43) 1.23 (0.35–4.31) 1.13 (0.41–3.15) 0.97 (0.35–2.66)
6.54 5.20 2.5 1.69 1.13 1.05 0.55 0.11 0.05 0.004
.01 .02 .11 .19 .29 .31 .46 .74 .81 .95
3.98 (1.33–11.89) 0.67 (0.47–0.96)
6.11 4.88
.01 .03
Clinical Characteristics Versus Hemodynamic Status
O Shah et al 111
Fig. 1. Kaplan-Meier estimates of freedom from death over 1 year by hemodynamic profile. P ⫽ .13 by log-rank test.
judgment, or the time invested in the examination. Physical findings associated with severe fluid overload, such as hepatomegaly, also predicted hemodynamic status better than signs of hypervolemia, such as jugular venous distension, rales, or edema. In patients identified as wet, lower proportional pulse pressure was strongly associated with the cold hemodynamic status, or lower cardiac index. Although proportional pulse pressure did not predict severe hypoperfusion for the dry hemodynamic profile, as noted earlier, this subgroup included only 19 patients. Prognosis and Hemodynamic Profiles This analysis describes the distribution of hemodynamic profiles in patients with advanced CHF. Most patients in this study fell into the wet/warm or wet/cold categories; very few met the criteria for dry/cold CHF. Overall mortality did not differ significantly among the 4 profiles. Patients who were classified as dry/warm did not have better outcomes than those in the wet/warm or the dry/cold categories. Importantly, there was no significant difference in survival among patients categorized
as wet/cold versus those identified as wet/warm. Patients identified as wet/cold did have significantly lower survival than the dry/warm group. There were no significant differences in the rates of death or readmission among the 4 groups. Hemodynamic measures may assist in predicting mortality after acute myocardial infarction. In the Forrester classification system, patients with a cardiac index less than 2.2 L/min/m2 and a PCWP 18 mm Hg or greater had higher mortality than those with a cardiac index 2.2 L/min/m2 or greater and a PCWP less than 18 mm Hg (8,12). Shell et al (13) likewise showed that patients with a PCWP greater than 18 mm Hg after acute myocardial infarction had significantly higher mortality than those with a PCWP 18 mm Hg or less (13). In contrast, earlier studies of patients with advanced CHF showed that baseline hemodynamic measures did not predict survival. Rather, the degree of reduction in the PCWP and its response to therapy have been shown to correlate with prognosis (14,15). This analysis shows that the 4 hemodynamic profiles may have limited use in estimating prognosis in patients with advanced CHF. Patients in the wet/cold profile did
112
Journal of Cardiac Failure Vol. 7 No. 2 June 2001
Fig. 2. Kaplan-Meier estimates of freedom from death or readmission over 1 year by hemodynamic profile. P ⫽ .44 by log-rank test.
have significantly higher mortality than those in the dry/warm category, but this was a secondary analysis of a subgroup analysis and thus should be interpreted with caution. In addition, there was no significant difference between patients who had hemodynamic measurements consistent with fluid overload and severe hypoperfusion (wet/cold) versus those who had only fluid overload (wet/warm), suggesting that both groups should be treated aggressively with evidence-based therapy.
much attention on this part of the protocol. These limitations actually may make the data more generalizable to standard practice, however, because CHF is diagnosed and managed by a spectrum of health care professionals with limited amounts of time for and experience with the physical examination.
Limitations
Clinicians rely on the history and physical examination to select and titrate therapies for patients with CHF. This analysis shows, however, that routine history and physical examination may lead to inaccurate estimation of hemodynamic status in these patients. In addition, unlike the Forrester classification, the 4 hemodynamic profiles may not provide significant information from which to estimate prognosis in patients with advanced heart failure. Prospective studies of hemodynamic profiles may help clinicians better understand the different presentations of advanced CHF and refine the management of this heterogenous group. Further research about the use of the
Subgroup analyses are subject to error because they are not prospectively specified, which can result in reduced statistical power and less accurate and complete data collection. Our results thus should be considered hypothesis generating, not definitive. This analysis retrospectively evaluated physical examinations performed as part of a clinical trial. Because of the variety of participating investigators, the degree of skill in the examination may have varied. Assessment of baseline physical-examination variables was not a prospective goal of the study. Investigators may not have focused as
Conclusions
Clinical Characteristics Versus Hemodynamic Status
patient’s history and physical examination also is needed, especially in relation to more objective ways to determine hemodynamics, such as pulmonary-artery catheterization, echocardiography, or neurohormonal levels. Identification of their relative merits could allow more judicious use of invasive modalities while optimizing therapy and improving outcomes.
8.
9.
References 10. 1. Stevenson LW, Perloff JK: The limited reliability of physical signs for estimating hemodynamics in chronic heart failure. JAMA 1989;261:884–888 2. Stevenson LW: When is heart failure a surgical disease. In: Rose E, ed. Management of end-stage heart disease. Baltimore: Lippincott, Williams & Wilkins, 1998;129–146 3. Stevenson LW, Massie BM, Francis GS: Optimizing therapy for complex or refractory heart failure: a management algorithm. Am Heart J 1998;135:S293–309 4. Califf RM, Adams KF, McKenna WJ, Gheorghiade M, Uretsky BF, McNutty SE, Darius H, Schulman K, Zannad F, Handberg-Thurmond E, Harrell FE, Wheeler W, SolerSoler J, Swedberg K: A randomized controlled trial of epoprostenol therapy for severe congestive heart failure: The Flolan International Randomized Survival Trial (FIRST). Am Heart J 1997;134:44–54 5. Criteria Committee of the American Heart Association, New York City Affiliate: Diseases of the heart and blood vessels—nomenclature and criteria for diagnosis, 9th ed. Little, Brown, Boston, 1994 6. Feinstein AR, Fisher MB, Pigeon JG: Changes in dyspnea-fatigue ratings as indicators of quality of life in the treatment of congestive heart failure. Am J Cardiol 1989;64:50–55 7. Guyatt GH, Sullivan MJ, Thompson PJ, Fallen EL, Pugsley SO, Taylor DW, Berman LB: The 6-minute walk: a
11.
12.
13.
14.
15.
O Shah et al 113
new measure of exercise capacity in patients with chronic heart failure. CMAJ 1985;132:919–923 Forrester JS, Diamond G, Chatterjee K, Swan HJ: Medical therapy of acute myocardial infarction by application of hemodynamic subsets. N Engl J Med 1976;295:1356– 1413 Butman SM, Ewy GA, Standen JR, Kern KB, Hahn E: Bedside cardiovascular examination in patients with severe chronic heart failure: importance of rest or inducible jugular venous distension. J Am Coll Cardiol 1993;22: 968–974 Chakko S, Woska D, Martinez H, de Marchena E, Futterman L, Kessler KM, Myerberg RJ: Clinical, radiographic, and hemodynamic correlations in chronic congestive heart failure: conflicting results may lead to inappropriate care. Am J Med 1991;90:353–359 Eisenberg PR, Jaffe AS, Schuster DP: Clinical evaluation compared to pulmonary artery catheterization in the hemodynamic assessment of critically ill patients. Crit Care Med 1984;12:549–553 Siniorakis E, Arvanitakis S, Voyatzopoulos G, Hatziandreov P, Plataris G, Alexandris A, Bonoris P: Hemodynamic classification in acute myocardial infarction—has anything changed in the last 3 decades? Chest 2000;117: 1286–1290 Shell WE, DeWood MA, Peter T, Mickle D, Prause JA, Forrester JS, Swan HJ: Comparison of clinical signs and hemodynamic state in the early hours of transmural myocardial infarction. Am Heart J 1982;104:521–528 Stevenson LW, Couper G, Natterson B: Target heart failure populations for newer therapies. Circulation 1995; 92:II174–181 Stevenson LW, Tillisch JH, Hamilton M, Luu M, Chelimsky-Fallick C, Moriguchi J: Importance of hemodynamic response to therapy in predicting survival with ejection fraction ⱕ20% secondary to ischemic or nonischemic dilated cardiomyopathy. Am J Cardiol 1990;66: 1348–1354