اشتباهات شایه در بخش آی سی یو 8
Do not use a Normal Arterial Blood Gas to rule out a Pulmonary Embolism
David N. Hager MD
Venous thromboembolic disease affects 200,000 persons in the United States annually. Among affected persons, almost half suffer symptomatic pulmonary embolism (PE). Without treatment, 30% of patients will die within 1 year. However, despite its high incidence, PE is a difficult diagnosis to make. Clinical presentations vary and the symptoms are often nonspecific. Though an abnormal arterial blood gas can be informative, normal indices of oxygenation and ventilation do not rule out PE.
The major risk factors for PE are the same as those for venous thrombosis. Broadly defined, these include a prior thromboembolic event, neoplastic disease, immobility, recent surgery or trauma, estrogen replacement therapy (especially in the context of tobacco use), and a family history of hypercoaguability such as activated protein C resistance, hyperhomocysteinemia, proteins C and S deficiency, antithrombin III deficiency, and factor V Leiden.
Signs and Symptoms
PE most commonly presents as a triad of dyspnea (70% to 80%), pleurisy, and tachypnea (RR >20). Among those patients enrolled and proven to have PE in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study, 97% exhibited one of these three symptoms. A less common presentation is hemo-dynamic compromise, which may be severe enough to cause syncope. The elderly often present with symptoms consistent with unresolving pneumonia or heart failure despite appropriate therapy. On clinical exam an enhanced second heart sound, tachycardia, and jugular venous distention may be appreciated. A normal chest radiograph is consistent with PE, as are other findings such as pleural effusion, Westermark sign (a focal loss of peripheral vascular markings), and even focal infiltrates. Tachycardia and nonspecific ST-T wave changes are the most common abnormal finding on an electrocardiogram (ECG). The frequently discussed S1Q3T3 pattern is present in less than 12% of patients with PE.
Hypoxia (PaO2 <80 mm Hg), hypocapnia (PaCO2 <35 mm Hg), and an elevated alveolar-arterial oxygen difference (P[Aâ€“a]O2 >20mm Hg) are the most common arterial blood gas abnormalities in patients with PE. However, in a group of subjects suspected of having PE, these indices did not effectively discriminate between those ultimately proven to have PE and those who did not. Further, among individuals without prior lung disease, these indices will often fall within normal limits even in the presence of PE. For these reasons, arterial blood gas data contribute little to the diagnosis or exclusion of PE.
Because of the nonspecific nature of the clinical findings associated with PE and the risks and limited availability of invasive pulmonary angiography (the definitive test), approaches to diagnosis should include a combination of (a) a clinical estimation of the likelihood of PE prior to further radiographic testing, and (b) the results of further radiographic testing. As demonstrated in the PIOPED study, this combination approach is more successful at diagnosing and excluding PE than clinical assessment or radiographic assessment alone. Though the interpretation of ventilation/perfusion (V/Q) imaging and computed tomographic (CT) angiography is beyond the scope of this chapter, all clinicians should be able to determine if a patient's clinical presentation is consistent with a high, intermediate, or low probability of PE.
- High probability is defined as an 80% to 100% chance of having had a PE. Such patients have (a) unexplained dyspnea, tachypnea, and/or pleurisy, (b) an unexplained gas exchange abnormality or x-ray abnormality, and (c) a risk factor for hypercoaguability.
- Intermediate probability is defined as a 20% to 79% chance of having had a PE. Most simply, such patients are defined as not meeting criteria for high or low probability.
- Low probability is defined as a 1% to 19% chance of having had a PE. Such patients may have dyspnea, tachypnea, pleurisy, x-ray abnormalities, or gas exchange abnormalities. However, in these patients, such findings can be attributed to a different condition. Typical risk factors for hypercoaguability are not present.
The clinical impression can then be combined with the radiologist's interpretation of a V/Q scan, CT angiogram, or both. Each radiographic test has limitations. V/Q studies are very difficult to
interpret in the context of coexisting lung disease such as pneumonia, emphysema, or fibrosis and take a long time to complete compared with a CT angiogram. However, V/Q has been more thoroughly evaluated and, in otherwise normal lungs, is more sensitive to peripheral emboli. CT angiography can be performed quickly and usually detect emboli in the large central and segmental pulmonary arteries if present. However, some reports suggest that the sensitivity of CT angiography may be as low as 70%. Further, the resolution of CT scanner at different centers varies widely, intravenous contrast is required (which may pose risks) and the expertise of radiologists also varies.
These imaging limitations emphasize the important role of the clinical impression in the workup of a suspected PE. If the clinical impression and radiographic assessment both suggest a low probability of PE, the evaluation can be stopped. If both suggest a high probability of PE, anticoagulation should be continued or initiated. However, if both the clinical impression and radiographic assessment suggest intermediate probability, or if one or the other suggests high probability, invasive angiography should be pursued. For the patient in whom the clinical probability of PE is low and the radiographic assessment suggests intermediate probability (or vice versa), some investigators have advocated the use of lower-extremity duplex ultrasound imaging to rule out venous thrombosis. If such testing is negative, anticoagulation therapy may be withheld without further evaluation provided close follow-up is available.
There has been much emphasis on the use of d-dimer assays in the evaluation of deep venous thrombosis and PE in the last several years. It is frequently suggested that a positive d-dimer is nondiagnostic, but that a negative result rules out venous thrombosis and/or PE. However, studies assessing the utility of d-dimer testing in the diagnosis of thromboembolic disease show that the results of d-dimer tests are subject to not only the type of test used (latex agglutination or enzyme-linked immunosorbent assay [ELISA]) but also the prevalence of thromboembolic disease in the study population, patient age, length of hospitalization, and even location in the hospital (floor vs. emergency department). For these reasons, d-dimer testing remains of little utility, especially when the clinical impression suggests a high probability of PE.
Hyers TM. Venous thromboembolism. Am J Respir Crit Care Med. 1999;159:1â€“14.
Tapson VF, Carroll BA, Davidson BL, et al. The diagnostic approach to acute venous thromboembolism. Am J Respir Crit Care Med. 1999;160:1043â€“1066.
Value of the Ventilation/Perfusion Scan in Acute Pulmonary Embolism. Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA.