Interobserver Agreement of Centrally Adjudicated Venous Ultrasound Video Documents in Asymptomatic Patients after Hip and Knee Replacement.

Feasibility of venous ultrasound as an endpoint measure in multicenter trials critically depends on the type and quality of documents which can be assessed by central adjudication. We evaluated the offline assessability of standardized video documents of complete compression ultrasound (CCUS) in asymptomatic patients 7 days after hip or knee replacement. The study was performed at a single study center. Consecutive asymptomatic patients one week after hip or knee replacement were screened for deep vein thrombosis by complete compression ultrasound of proximal and distal veins including muscle veins by one trained sonographer. Examinations were recorded on video tape in a standardized format. Video sequences with an approximate duration of 60 seconds demonstrated femoral veins, popliteal veins, peroneal veins and tibial posterior veins, respectively. The video documents were digitized and electronically stored in a central adjudication unit. All videos were read by two independent readers, and veins were scored as normal, thrombotic or nonevaluable. All thrombotic findings were re-assessed by two senior readers according to the same protocol. Analysis of inter-observer agreement was performed by Cohen’s Kappa coefficient. Video documents of 300 legs were recorded in 150 patients. Documents of 10 patients (6.7%) or 13 legs (4.3%) were scored as non-evaluable by at least one reader. Reading A revealed 5.5% proximal DVT and 18.7% distal DVT including muscle vein thrombosis. Reading B revealed 4.1% and 23.6%, respectively. Kappa values for different types of DVT are given in the table.

Conclusion: In a single center setting, interobserver agreement of centrally adjudicated standardized CCUS documents in the diagnosis of asymptomatic postoperative DVT is strong for all categories of DVT. It remains to be established how these figures will be affected by a multicenter setting.