Diagnostic accuracy of phleborheography in deep venous thrombosis

The clinical diagnosis of acute deep thrombophlebitis of the lower extremities is notoriously grossly inaccurate, even in experienced hands. Phleborheography (PRG) is one of the noninvasive diagnostic methods developed to make diagnosis accurate, noninvasive, and easily reproducible. The records of the first 272 patients studied by this method at South Fulton Hospital were examined, beginning June 1977 and ending August 1979. In those patients who had PRG and radiologic phlebography, the two methods were compared. Phlebography was considered as the standard for comparison. There were 36 patients who had both studies performed. Of the 36 patients studied, the PRG and the x-ray studies agreed in 34. Both were normal in 21 patients; and in 13 patients both were abnormal and in agreement. The studies did not agree in two of the 36 patients. There were none where the PRG was normal and the x-ray abnormal, and two where the PRG was abnormal and the x-ray normal. The diagnostic accuracy of a normal PRG in this study was 100 per cent, with no false-negatives. There were two false-positives with an accuracy of 87 per cent. The overall accuracy was 94 per cent. There were no complications.