How "gold" is the standard? Interobservers' variation on venograms.

ABSTRACT

BACKGROUND: The aim of this study was to determine the degree of agreement between radiologists having different levels of experience, in reporting 151 venograms independently. It was also aimed to assess whether the degree of disagreement would have influenced the final outcome of a thromboprophylactic study and the decision to anticoagulate the patients. METHODS: Seventy-eight patients undergoing total hip replacement had bilateral venography on postoperative days 8-12. Patients were allocated to receive either a low molecular weight heparin (enoxaparin) with or without TED stockings or a placebo Each of the 151 venograms obtained has been reported on four times. RESULTS: The overall incidence of deep venous thrombosis was 42% (33 of 78 patients). The range of the Kappa value of each radiologist versus the others was 0.568 to 0.669. There was a significant decrease in the incidence of deep venous thrombosis in the treatment versus the placebo groups. On an intention to treat all diagnosed thrombi, a difference of up to 16% was demonstrated between the report of individual radiologists and the panel. This difference was reduced to 8% when only thrombi proximal to the calf were considered clinically significant. CONCLUSIONS: These differences reflect the radiologist's experience and frequency of reporting on venograms and should be taken into account when studies for thromboprophylaxis are planned.