Background & Objectives: Heart failure (HF) is a pathophysiologic state in which cardiac dysfunction leads to insufficient pumpage of blood for metabolic needs of body. One of the important problems in primary approach to patients with HF, is estimation of HF severity. The present research is an attempt to investigate the accuracy of CXR in diagnosis of HF severity especially in emergency wards.

  Methods: This sensitivity analysis study was conducted on 59 patients admitted to emergency ward of Buali Hospital during March–September 2003, with primary diagnosis of HF. One randiologist used to observe and report radiologic findings without knowing the results of each cardiography.Then cardiologists used to check on the reports and selected patients whose EF was under 50% and eventually sensitivity and specificity of radiologic findings in diagnosing severity of HF (EF ≤ 35%) were calculated.

  Results: 59 patients (53% female and the rest male) were investigated whose age mean was 65.6±10.9 EF mean was 33.4±9.09. In the diagnosis of pulmonary arterial hypertension, CXR had 72% sensivity and 18% specifity, considering the reliance on the CXR in diagnosis of severe HF (EF ≤ 35%) findings were: cardiomeglay (94%), pulmonary artery congestion (75%), interstitial edema (78%) and alveolar edema (76%) with high specifity the reliance on the combination of pulmonary cephalization, interstitial edema and pleural effusion was to have 92% sensivity and 10% specifity.

  Conclusion: According to high sensitivity of cardiomegaly, pulmonary arterial congestion, interstitial edema and pleural effusion, lack of there findings in CXR reduce probability of severe HF. Alveolar edema because of its high specificity increases probability of severe HF. CXR also has high sensivity in diagnosis and R/O of PAH.