Is surgeons' experience important on intra- and inter-observer reliability of classifications used for adult femoral neck fracture?

Purpose: To evaluate whether surgeons' experience affect inter- and intra-observer reliability among mostly used classification systems for femoral neck fractures.

Material and methods: A power point presentation was prepared with 107 slides which were anteroposterior radiographs of each femoral neck fracture. Five residents, 5 orthopaedic surgeons and 5 senior orthopaedic surgeons reviewed this presentation and classified the fractures according to Garden, Pauwels and AO classifications. The order of the slides was changed and reviews were repeated after 3 months. Fleiss kappa and intraclass correlation coefficient values were calculated to evaluate inter and intra-observer reliability.

Results: Garden and AO classifications' inter-observer reliabilities were similar and higher than Pauwels classification. Among three experience groups, the inter-observer reliability for Garden classification was highest in senior surgeon group, the interobserver reliability for AO classification was highest in surgeon group, and interobserver reliability of Pauwels classification was highest in low experienced groups (residents and surgeons). Intra-observer reliability was highest for Garden and lowest for Pauwels classifications. Surgical experience was found to be not effective for intraobserver reliability.

Conclusion: Both Garden and AO classifications were more reliable than Pauwels classification. Surgical experience was not significantly important on these three classification systems' evaluation.

Level of Evidence: Level IV, Diagnostic study.