The incidence of post-operative peri-prosthetic fractures is increasing. This is a consequence of the larger number of revision cases being undertaken, the increase in the use of cementless implants and a number of patients who develop undetected osteolysis as a result of poor follow up. The Vancouver classification has been shown previously, in North America, to be a valid and reliable method for determining the configuration of a periprosthetic fracture. This is essential in directing the fractures further management appropriately and the classification system has been adopted by surgeons throughout the world.

The reliability of any classification system depends on the reproducibility between the clinicians who are making the management decisions. We have revalidated this classification system, independently from the original authors, at a centre in Europe. The radiographs from 30 patients with peri-prosthetic fractures were reviewed by 6 expert consultant surgeons, 6 non-experts at registrar level and 6 medical students, who had received no specialist teaching in this area, in order to assess intra and inter-observer reliability and reproducibility. Each observer read the radiographs on 2 separate occasions and classified the fracture according to its type (A, B1, B2, B3, and C). The results were subjected to weighted kappa analysis and were: 0.76 (substantial agreement) for experts; 0.68 (substantial agreement) for non-experts; and 0.61 (substantial agreement) for medical students.

Our results confirm the reliability and reproducibility of this classification system. In addition we have shown that substantial agreement can even be found between individuals with no specialist training. This is a classification system that can be reliably used by non-experts, between centres and across continents.