In summary, the chest radiograph has only moderate accuracy in visualizing opacification caused by cardiopulmonary abnormalities and may be quite nonspecific as to etiology, whereas it has high diagnostic accuracy for detecting malpositioning of tubes and lines. While focal parenchymal abnormalities are usually visualized on chest radiographs, identification of concomitant abnormalities when ARDS or PE already exist is more difficult. Atelectasis, aspiration, pneumonia, pulmonary hemorrhage, pulmonary thromboembolism, atypical cardiogenic edema, asymmetric ARDS, and neoplasms may be indistinguishable. Repeat chest radiographs and different views may be helpful, as the progression and time course of various etiologies can be quite different. On the other hand, Winer-Muram et al found that review of prior radiographs and clinical data did not improve the diagnostic accuracy for either ARDS or pneumonia. Pleural effusions may even be difficult to distinguish from parenchymal processes, particularly when the patient is in the supine position. Additional views with the patient in a different position--semi-erect, decubitus, or cross-table lateral--may be of assistance. In most cases, pneumothorax is readily detected. Additional studies such as the decubitus view occasionally may be necessary for further evaluation when there is uncertainty about the findings. Subcutaneous air is readily visualized radiographically. Pneumomediastinum and interstitial pulmonary emphysema may be more difficult to see. It is well known that CT allows visualization of much smaller abnormal air collections than radiography. Despite this lack of sensitivity and specificity of chest films, studies have shown that up to 65% of daily films in the ICU reveal significant and/or unsuspected abnormalities that may change the patient's diagnosis or management. Based on these results, the consensus opinion of the ACR Expert Panel found that daily chest radiographs are indicated on patients with acute cardiopulmonary problems and those receiving mechanical ventilation. Patients who require cardiac monitoring but are otherwise stable require only an initial admission film. Additional radiographs are indicated only when a new device is placed or when there is a specific question regarding cardiopulmonary status. It is also noteworthy that despite the chest film being the most commonly ordered radiologic examination for inpatients, there are no comprehensive studies evaluating its cost-effectiveness. Although several studies have done a very limited cost accounting of the potential savings by eliminating routine films in the evaluation of specific subsets of patients, overall impact on patient outcome has not been investigated. Thus, a true assessment of cost-effectiveness has yet to be determined.