Pulmonary emboli are often difficult both to diagnose and to confirm. Sudden dyspnea is the most common symptom; but clinically there may be any combination of chest pain, dyspnea, and possibly hemoptysis. Diseases that must be also considered are acute myocardial infarction (MI) and pneumonia. Pulmonary embolism is often associated with chronic congestive heart failure, cor pulmonale, postoperative complications of major surgery, and fractures of the pelvis or lower extremities, all situations in which MI itself is more likely. The classic x-ray finding of a wedge-shaped lung shadow is often absent or late in developing, because not all cases of embolism develop actual pulmonary infarction, even when the embolus is large.

Laboratory tests in pulmonary embolism have not been very helpful. Initial reports of a characteristic test triad (elevated total bilirubin and lactic dehydrogenase [LDH] values with normal aspartate aminotransferase [AST] proved disappointing, because only 20%-25% of patients display this combination. Reports that LDH values are elevated in 80% of patients are probably optimistic. In addition, LDH values may be elevated in MI or liver passive congestion, conditions that could mimic or be associated with embolism. Theoretically, LDH isoenzyme fractionation should help, since the classic isoenzyme pattern of pulmonary embolism is a fraction 3 increase. Unfortunately this technique also has proved disappointing, since a variety of patterns have been found in embolization (some due to complication of embolization, such as liver congestion), and fraction 3 may be normal. Total creatine phosphokinase (CK) initially was advocated to differentiate embolization (normal CK) from MI (elevated CK value), but later reports indicate that the total CK value may become elevated in some patients with embolism. The CK isoenzymes, however, are reliable in confirming MI, and normal CK-MB values plus normal LDH-1/LDH-2 ratios (obtained at proper times) is also reliable in ruling out MI.

Arterial oxygen saturation has been proposed as a screening test for pulmonary embolism, since most patients with embolism develop arterial oxygen saturation values less than 80 mm Hg. However, 15%-20% (range, 10%-26%) of patients with pulmonary embolization have oxygen saturation greater than 80 mm Hg, and 5%-6% have values greater than 90%. Conversely, many patients have chronic lung disease or other reasons for decreased oxygen saturation, so that in many patients one would need a previous normal test result to interpret the value after a possible embolism.

The most useful screening procedure for pulmonary embolism is the lung scan. Serum albumin is tagged with a radioisotope, and the tagged albumin molecules are treated in such a way as to cause aggregation into larger molecular groups (50-100 µm). This material is injected into a vein, passes through the right side of the heart, and is sent into the pulmonary artery. The molecules then are trapped in small arterioles of the pulmonary artery circulation, so that a radiation detector scan of the lungs shows a diffuse radioactive uptake throughout both lungs from these trapped radioactive molecules. A scan is a visual chart of the radioactivity counts over a specified area that are received by the radiation detector. The isotope solution is too dilute to cause any difficulty by its partial occlusion of the pulmonary circulation; only a small percentage of the arterioles are affected, and the albumin is metabolized in 3-4 hours. If a part of the pulmonary artery circulation is already occluded by a thrombus, the isotope does not reach that part of the lung, and the portion of lung affected does not show any uptake on the scan (abnormal scan).

The lung scan becomes abnormal immediately after total occlusion of the pulmonary artery or any branches of the pulmonary artery that are of significant size. There does not have to be actual pulmonary infarction, since the scan results do not depend on tissue necrosis, only on mechanical vessel occlusion. However, in conditions that temporarily or permanently occlude or cut down lung vascularity, there will be varying degrees of abnormality on lung scan; these conditions include cysts, abscesses, many cases of carcinoma, scars, and a considerable number of pneumonias, especially when necrotizing. However, many of these conditions may be at least tentatively ruled out by comparison of the scan results with a chest x-ray film. A chest x-ray film should therefore be obtained with the lung scan.

Asthma in the acute phase may also produce focal perfusion defects due to bronchial obstruction. These disappear after treatment and therefore can mimic emboli. Congestive heart failure or pulmonary emphysema often cause multiple perfusion abnormalities on the lung scan. This is a major problem in the elderly, since dyspnea is one of the symptoms associated with embolization or may be a source of confusion when emphysema and emboli coexist. Emphysema abnormality can be differentiated from that of embolization by a follow-up lung scan after 6-8 days. Defects due to emphysema persist unaltered, whereas those due to emboli tend to change configuration. The repeat study could be performed earlier but with increased risk of insufficient time lapse to permit diagnostic changes.

The lung scan, like the chest x-ray, is nonspecific; that is, a variety of conditions produce abnormality. Certain findings increase the probability of embolization and serial studies provide the best information. In some cases, a xenon isotope lung ventilation study may help differentiate emboli from other etiologies of perfusion defect; but when congestive heart failure is present, when the defect is small, and when embolization is superimposed on severe emphysema, the xenon study may not be reliable. Pulmonary artery angiography provides a more definitive answer than the lung scan, but it is a relatively complicated invasive procedure, entails some risk, and may miss small peripheral clots. The lung scan, therefore, is more useful than angiography as a screening procedure. A normal lung scan effectively rules out pulmonary embolization. A minimum of four lung scan views (anterior, posterior, and both lateral projections) is required to constitute an adequate perfusion lung scan study.