Development and validation of a clinical history form for the diagnosis of congestive heart failure

Our aim was to develop and validate a new diagnostic tool for congestive heart failure (CHF) based only on clinical examination, medical history and current drug use. In the developmental phase, we enrolled 520 consecutive patients with heart disease of different etiology; the diagnosis of CHF was made by means of Smith's clinical and radiological criteria, and the Boston criteria, with a substantial interscale agreement (kappa = 0.89). The initial version of our Form included 37 items on clinical examination, medical history and drug use information. After an item reduction process, the final version of the Clinical History Form (CH Form) included 15 items, each with a score 1 to 4. The CH Form score showed a progressive, significant increase as NYHA Class increased. With a cut-off of 4 points, sensitivity was 88.6% and specificity 86.8% against Smith's criteria for diagnosis of CHF. Substantial interrater agreement was observed for all the 15 items (kappa > 0.6) on a subsample of 250 patients. In the validation phase, we studied an independent sample of 72 patients with heart disease. The CH Form was significantly correlated with left ventricular ejection fraction (r = 0.42; p < 0.0005) and peak oxygen consumption (r = 0.69; p < 0.0001). In the 64 (89%) patients who underwent non-emergent right-heart catheterization, the CH Form score was significantly correlated with pulmonary capillary wedge pressure (r = 0.84; p < 0.0001). The CH Form may represent a useful instrument for the diagnosis of CHF.