C-reactive protein as a predictor of improvement and readmission in heart failure

Objectives: Only recently, new risk factors to explain atherosclerotic disease have been identified. One of the most important clinical manifestations of atherosclerosis is heart failure. Our study was aimed at investigating C‐reactive protein (CRP), a marker of systemic inflammation, in the context of heart failure, and to determine its usefulness in predicting the need for readmission in patients with heart failure and their degree of improvement. Design: We studied patients admitted to our hospital due to heart failure, independent of the cause. CRP levels were measured with a sensitive standard assay on a Nephelometer analyser. Patients were classified on admission and discharge following New York Heart Association (NYHA) functional criteria; left ejection fraction was also determined by transthoracic echocardiography. Patients presenting clear sources of infection or inflammatory disease were excluded. Our control group consisted of patients admitted for syncope. Each patient was followed up through a computer system controlling admissions to and discharge from the hospital, for a period of 18 months after initial admission. End points considered were NYHA functional class on discharge, readmission and death. Results: We studied prospectively 76 patients with a mean age of 73.5±11 [95% confidence interval (CI) 71.2–75.8]; 44 were male (58%) and 32 female (42%). The mean CRP level in patients with heart failure was 3.94±5.87 (95% CI, 1.26–7.60), while in 15 patients with syncope it was 0.84±1.95 (95% CI, 0.96–2.94) (P‐0.0007). The principal causes of heart failure included dilated cardiomyopathy due to coronary arterial disease (30%), valvular disease (28%) and heart failure secondary to hypertension (25%). The mean left ejection fraction adequately measured in 72 (95%) patients was 50.41±9.88 (95% CI, 41.20–59.65). We observed a trend of higher CRP levels in relation to ejection fractions below 35%: 7.50±9.88 vs. 3.75±4.57, (P‐0.09). Our results showed that on discharge CRP levels increased in relation to NYHA class: I: 0.74±0.69; II: 3.78±3.76; III: 7.4±8.65; IV: 12.2±15.27 (P0.9 mg/dl were identified as candidates for earlier hospitalisation than those with levels below 0.9 mg/dl (P‐0.02) RR‐1.43. In logistic‐regression analysis the only group of tested variables pr