| Abstract: | Blood pool radionuclide venography and contrast venography were carried out in 50 patients in whom deep vein thrombosis was suspected. The two procedures were compared on the basis of 198 individual vessel segments. The overall concordance between radionuclide and contrast venography was 92%. Using contrast venography as the standard, the sensitivity of radionuclide venography was 90% and specificity 96%. The predictive accuracy of a positive test was 77% and of a negative test was 97%. The performance of radionuclide venography improved if the calf veins were excluded from analysis. The main limiting factor affecting the performance of radionuclide venography was the moderately poor resolution of the technique. Technical manipulations to improve resolution were time consuming and added little to the accuracy of the test. Radionuclide will not replace contrast venography but may well be used to complement contrast venography when it is technically unsatisfactory or unequivocal, in patients with a history of intolerance to contrast media, and in bedbound patients. |
