AJR:195, September 2010
Kaushik Shahir, Lawrence R. Goodman, Ardita Tali, Kristin M. Thorsen, Robert S. Hellman
OBJECTIVE. The purpose of this study was to evaluate the equivalence of CT pulmonary angiography and perfusion scanning in terms of diagnostic quality and negative predictive value in the imaging of pulmonary embolism (PE) in pregnancy.
MATERIALS AND METHODS. Between 2000 and 2007 at a university hospital and a large private hospital, 199 pregnant patients under-went 106 CT pulmonary angiographic examinations and 99 perfusion scans. Image quality was evaluated, and the findings were reread by radi-ologists and compared with the original clinical readings. Three-month follow-up findings of PE and deep venous thrombosis were recorded.
RESULTS. PE was found in four of the 106 patients (3.7%) who underwent CT pulmonary angiography. The overall image quality was poor in 5.6% of cases, acceptable in 17.9%, and good in 76.4%. Fourteen CT and nine radiographic studies showed other clinically significant abnormalities. Six patients had indeterminate CT pulmonary angiographic findings, three had normal perfusion scans, and none underwent anticoagulation. All perfusion scan findings were nortnal. There was one incomplete study, and follow-up CT pulmonary angiography per-formed the same day showed PE. Two of 99 studies (2.02%) showed intermediate probability of the presence of PE; PE was not found at CT pulmonary angiography, but pneumonia was found. PE was found in one postpartum patient 9 weeks after she had undergone CT pulmonary angiography and ultrasound with normal findings. None of the patients died.
CONCLUSION. CT pulmonary angiography and perfusion scanning have equivalent clinical negative predictive value (99% for CT pul-monary angiography; 100% for perfusion scanning) and image quality in the care of pregnant patients. Therefore, the choice of study should be based on other considerations, such as radiation concern, radiographic results, alternative diagnosis, and equipment availability. Reducing the amount of radiation to the maternal breast favors use of perfusion scanning when the radiographic findings are normal and there is no clinical suspicion of an alternative diagnosis.