Utility of History, Physical Examination, Electrocardiogram, and ...

More documents

Recommendations

Info

Table 2. Sensitivity, Specificity, and Positive and Negative Predictive Values of Clinical Findings for Normal Systolic Function (Ejection Fraction 45%) hypertrophy and left atrial abnormality were significantly more common in those with decreased systolic function. Radiographic cardiomegaly and flow cephalization were significantly more common in patients with decreased function, whereas rates of pleural effusion and pulmonary edema were similar. Some clinical parameters had a high sensitivity or a high specificity for normal systolic function, but none had both a high sensitivity and a high specificity (Table 2). No clinical parameters had a high positive or negative Clinical Findings in Heart Failure/Thomas et al Sensitivity Specificity Positive Predictive Value Negative Predictive Value Percentage Historical factors Age 60 years 52 61 53 60 Female sex 56 65 58 63 Hypertension 78 26 48 57 Diabetes mellitus 40 66 51 56 Hyperlipidemia 6 94 46 53 Smoking 39 55 43 51 History of coronary disease 22 70 39 51 Chronic renal failure 15 91 59 55 Alcohol use Symptoms 20 63 32 48 Dyspnea at rest 39 56 42 52 Dyspnea on exertion 98 2 46 50 Orthopnea 86 14 47 55 Paroxysmal nocturnal dyspnea 74 23 45 50 Angina 4 89 24 52 Nonanginal chest pain Physical examination 37 74 54 57 Body mass index 30 kg/m 2 62 52 54 60 Heart rate 100 beats per minute 31 50 34 45 Systolic blood pressure 160 mm Hg 50 73 62 63 Diastolic blood pressure 110 mm Hg 25 87 63 57 Pulse pressure 60 mm Hg 61 55 54 62 Jugular venous distention 72 26 46 53 Pedal edema 83 25 48 63 Rales 80 11 43 38 S3 gallop 28 59 37 49 S4 gallop Electrocardiogram 11 93 56 55 Atrial fibrillation 19 90 61 57 Abnormal Q waves 18 79 42 53 Left ventricular hypertrophy 22 58 31 47 Left atrial abnormality Chest radiograph 22 48 25 45 Cardiomegaly 86 7 45 36 Cephalization 79 9 43 34 Pleural effusion 26 81 55 56 Pulmonary edema 14 85 45 53 440 April 15, 2002 THE AMERICAN JOURNAL OF MEDICINE Volume 112 predictive value for differentiating normal from decreased systolic function. In a multivariate analysis, several clinical findings were independent predictors of normal systolic function, including female sex (odds ratio [OR] 2.3; 95% confidence interval [CI]: 1.3 to 4.2; P 0.03), heart rate 100 beats per minute (OR 1.9; 95% CI: 1.1 to 3.9; P 0.005), use of ACE inhibitors (OR 0.34; 95% CI: 0.17 to 0.67; P 0.001), and use of digoxin (OR 0.35; 95% CI: 0.20 to 0.66; P 0.03). The area under the receiver oper-
ating characteristic curves for the clinical features were 0.62 for sex and 0.63 for heart rate. Of the patients with decreased systolic function, 44 (36%) were male and had tachycardia, compared with 11 (11%) of those with normal function, yielding a sensitivity of 36%, specificity of 89%, positive predictive value of 80%, and negative predictive value of 55% for the combination of these two clinical predictors. When medications were excluded from the multivariate analysis, variables that were significantly more common in patients with normal systolic function included female sex (OR 2.4; 95% CI: 1.2 to 4.6; P 0.002), heart rate 100 beats per minute (OR 3.5; 95% CI: 1.7 to 7.1; P 0.001), and systolic blood pressure 160 mm Hg (OR 3.7; 95% CI: 1.8 to 7.4; P 0.03), while left atrial abnormality on ECG was significantly less common in these patients (OR 0.24; 95% CI: 0.12 to 0.50; P 0.04). There were greater differences between the 89 patients with markedly decreased systolic function (ejection fraction 30%) and the 136 patients with ejection fractions 30% (Table 3). However, the sensitivity, specificity, and positive and negative predictive values for these clinical features were again low (Table 4). DISCUSSION This prospective study of patients hospitalized with heart failure observed that some clinical findings were more common in patients with systolic or diastolic heart failure. However, none of these findings could differentiate reliably between patients with normal and decreased systolic function. Several previous studies have compared clinical findings in patients with systolic or diastolic heart failure, with inconsistent results. Some investigators have found that no clinical variables correlated with left ventricular systolic function (10,13,14,16,17), while others found— as we did—that older age (8,11,12), hypertension (7,9,15), female sex (8,9), absence of an S 3 gallop (11,12), and obesity (9) were associated with normal systolic function. Other studies have reported that normal systolic function was associated with atrial fibrillation (8), absence of jugular venous distention (9), peripheral edema (15), or the presence (15) or absence (12) of paroxysmal nocturnal dyspnea. One reason for the inconsistency between studies may be differences in study design. Most studies have been retrospective, and some were reviews of patients who underwent echocardiography or other tests to evaluate left ventricular function. In contrast, we performed prospective measurements in consecutive patients who had presented with heart failure. In addition, different definitions of “normal” systolic function have been used. We Clinical Findings in Heart Failure/Thomas et al defined normal function as an ejection fraction 45%, as suggested by the European Study Group on Diastolic Heart Failure (20). Other studies have also used this definition (3,6,7,14,21,22), whereas some have used an ejection fraction 50% (2,4,8,11,13,23–27), 55% (12,28), or other values (9,10,29). When an ejection fraction cutoff value of 30% was used, significant differences in clinical features were more evident in our study, but the sensitivity, specificity, and predictive values of these clinical findings for identifying patients with severe left ventricular systolic dysfunction remained poor. Given the inconsistent and even contradictory findings of previous studies, it is not surprising that we did not find any clinical parameters that could distinguish patients who had normal function from those with decreased systolic function. Increased filling pressures are characteristic of both diastolic and systolic heart failure; all patients in our study were hospitalized, perhaps explaining why the clinical findings of congestion were similar in the two groups. Other studies have compared physical findings and invasive hemodynamic measurements in patients with chronic heart failure, and have shown that the physical examination is a relatively insensitive technique for identifying elevated filling pressures (30). There were several differences in medication use between the patients with normal and decreased systolic function. It is possible that medications may have affected the clinical findings; for example, the higher prevalence of ACE inhibitor use may have reduced the blood pressure and the frequency of an S 3 gallop in patients with decreased systolic function. Similarly, lower heart rates in the patients with normal function may have been due in part to use of beta-blockers or rate-lowering calcium antagonists. One puzzling issue is why the prevalence of an S 3 gallop did not differ significantly between the groups; the borderline P values (univariate P 0.07, multivariate P 0.09) suggest that this may have been due to an insufficient sample size. However, an S 3 gallop was present in 41% of patients with decreased systolic function and in 28% of those with normal function, suggesting that it would not be a clinically useful discriminator even if the difference was statistically significant. The mechanism of an S 3 gallop in patients with normal function, and the reason for the relatively low prevalence of an S 3 gallop in those with decreased function, are not clear. Prior studies have also noted that an S 3 gallop is common in patients with diastolic heart failure (7,11,24). We found an independent association between left atrial abnormality on ECG and systolic dysfunction. Left atrial abnormality correlates more with left atrial volume than with left atrial pressure (19), and heart failure patients with low ejection fractions commonly have an enlarged left atrium. Left ventricular hypertrophy on ECG April 15, 2002 THE AMERICAN JOURNAL OF MEDICINE Volume 112 441
Page 1 and 2: CLINICAL STUDIES Utility of History
Page 3: Table 1. Characteristics of Patient
Page 7 and 8: Table 4. Sensitivity, Specificity,
Page 9: 26. Kitzman DW, Higginbotham MB, Co

Utility of History, Physical Examination, Electrocardiogram, and ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?