• 26
    Feb
  • Clinical Validity of a Normal Perfusion Lung Scan in Patients: METHODS

Rtiients

Between 1981 and 1988,1,420 consecutive patients with clinically suspected pulmonary embolism were referred to the Thromboem­bolism Service at Chedoke-McMaster Hospitals and were studied prospectively. The 515 (36 percent) of these patients who had normal perfusion lung scans are the subject of this report. The findings in the patients with abnormal perfusion scans have been described previously.

Study Protocol

Each patient was examined on the day of referral by the consultant physician and the study nurse who confirmed and recorded (on special forms) the clinical history and physical findings listed in Table 1. All patients had an electrocardiogram and a chest x-ray film and underwent impedance plethysmography (to detect proxi­mal deep-vein thrombosis). Impedance plethysmography is sensitive and specific for proximal-vein thrombosis in patients with suspected pulmonary embolism (sensitivity, 86 percent; specificity, 97 percent).
canada drugs online

Perfusion lung scanning was then performed. After intravenous injection of 3 mCi of technetium 99m-macroaggregated albumin, six anatomic views were obtained by perfusion scanning (anterior, posterior, right and left lateral, right and left posterior oblique).

One million counts were obtained per view. For each anatomic view, the perfusion images were presented on roentgenographs film in a 5 x 4.5-cm hexagonal field. The perfusion lung scan was classified as normal if perfusion defects were absent in all of the views. The perfusion scan was considered abnormal if a perfusion defect was present in any one or more views.

Anticoagulant therapy was withheld or withdrawn immediately in all patients with normal perfusion scans, regardless of the clinical findings (except in the occasional patient with associated deep-vein thrombosis). The correctness of this approach was tested by long- term follow-up.

All patients with normal lung scans were reviewed by the consultant physician to establish the cause for the clinical findings, either by clinical association (eg, pneumonia and positive sputum or blood cultures) or by the evolution of their disease (eg, cancer).

All patients were assessed by long-term follow-up for three months. They were asked to return at once if they developed symptoms or signs suggesting pulmonary embolism or deep-vein thrombosis. Those with clinically suspected pulmonary embolism underwent ventilation-perfusion lung scanning; pulmonary angiog­raphy was performed in patients with new perfusion defects that did not have a high probability for pulmonary embolism. Patients with clinically suspected deep-vein thrombosis underwent impe­dance plethysmography, leg scanning and venography, according to previously described protocols.

All patients were reevaluated routinely at three months. An interim history was taken which addressed general health, specific symptoms (chest pain, dyspnea, hemoptysis, syncope, as well as leg pain, tenderness or swelling), hospital admission and the use of anticoagulants. The majority of patients were seen in the clinic and testing with impedance plethysmography was repeated; those who were geographically inaccessible were followed up by telephone. For each patient who died, the cause of death was documented from autopsy findings, coroners report or independent clinical review. Apcalis Oral Jelly

The criteria for pulmonary embolism on follow-up were: (a) a high probability ventilation-perfusion lung scan (segmental or larger perfusion defect with ventilation mismatch), (b) a positive pulmonary angiogram (intraluminal filling defects that were constant on all films or sharp cutoff of a vessel greater than 2.5 mm in diameter) or (c) pulmonary embolism found at autopsy.

The diagnostic criteria for deep-vein thrombosis by objective testing with impedance plethysmography, leg scanning, and venog­raphy have been described previously.

Statistical Considerations

The 95 percent confidence limits for the true frequency of venous thromboembolism on follow-up were calculated from the binomial distribution.

Online Pharmacy