Inter-observer variation. An alternative method of assessing the role of ultrasonic imaging in clinical decision-making in lower limb arterial disease

Background. The aim of the study is to determine the role of duplex scannin g in deciding on the final treatment for patients with lower limb arterial disease compared to intra-arterial digital subtraction arteriography (IA DS A). Methods. Eighty-two patients (55 males, 27 females, mean age 68 years) with lower limb arterial disease had both duplex and arteriography performed. T he findings of both modalities were reported by sonographer and radiologist in 164 anonymous reports. Five vascular surgeons (consultant grade) were a sked to plan their treatment on the data provided in these reports. For the purposes of statistical analysis, the actual treatment the patient receive d was used as the reference standard. Both reports and treatment decisions were coded (double blinded). Results. The overall accuracy of duplex scanning to assess arterial disease of the lower limbs showed a sensitivity of 92%, and specificity of 99%. Th e accuracy of the decisions based on duplex was 84% and kappa with k of 0.6 2 (95% CI+0.14) compared to the reference standard, while the accuracy of t he decision based on arteriography was 85% and kappa 0.63 (95% CI+0.14). Th e accuracy of decisions based on duplex compared to those based on arteriog raphy (arteriography used as reference standard) was 91% and a kappa of 0.7 7 (95% CI+0.14). Conclusions. Duplex scanning is an accurate diagnostic modality, and clinic al decisions can be safely based on its findings.