The role of digital radiography has evolved in recent years. In many hospitals, radiographs have been completely digitised and moved to Picture Archiving and Communications System (PACS). Pre-operative templating for arthroplasty has been a major problem as a result.

We investigate the accuracy of Orthoview™ software in templating for hip and knee arthroplasty. A retrospective review of 20 Stryker-Exter hip and 20 Biomet-Vanguard knee arthroplasties was conducted. Anonymised preoperative radiographs were reviewed by experienced orthopaedic surgeons. Templated component sizes were compared with actual implanted component sizes. All radiographs were digitised on Kodak Carestream PACS. Five surgeons were asked to separately review the radiographs to avoid intra-observer error.

In templating for hip arthroplasty, Orthoview™ was 80% accurate in predicting the femoral stem size within one size of the actual component used. It predicted the offset with 100% accuracy. In 90% of patients, the actual head implant was within one size of the templated head. The system was able to predict the acetabular component size in only 30%. In knee arthroplasty, Orthoview™ was 80% accurate within one size of the actual component used for the femur and 90% for the tibia.

Orthoview™ enables the flexibility of digitised films to be used for pre-operative templating. It is reasonably accurate in prediction of femoral sizing in both hip and knee arthroplasty and tibial size in knee arthroplasty. It is considerably less useful for acetabular sizing. Surgeons should keep this variability in mind until more accurate systems are available.