Venography of the lower limbs: Pitfalls of the diagnostic standard

Rationale and objectives: Phlebography is considered the diagnostic standard for suspected deep venous thrombosis. The authors studied the inter-observer variability of phlebogram interpretation in the setting of a multicenter therapeutic trial of the thrombolytic agent alteplase. Methods: The interpretation of 31 pairs of venograms (before and after thrombolytic therapy) was studied by comparing the quantitative Marder’s scores which were computed by three experts and the qualitative assessment of phlebographic changes induced by thrombolysis by the panel of experts and by the investigators. Results: Although the scores of the three experts correlated fairly well (r = .67-.82; P < .001), they differed significantly from each other (P < .0001). Substantial differences also were found between local (by investigators) qualitative evaluation of the venographic changes induced by the treatment and central evaluation by the panel of experts (coefficient of agreement kappa = 0.19), local assessment being significantly more optimistic (P = .002) than central judgment. Conclusion: Significant differences were observed between assessment of changes in venographic scores after thrombolytic treatment both among three expert radiologists, and between the panel of experts and the local investigators of the multicenter trial. This observation points to the need for an a priori definition of well-characterized decision criteria to allow a valid interpretation of the effects of the therapeutic intervention.