BACKGROUND: The Vancouver classification of periprosthetic femur fractures divides B1 and B2 subtypes based on the stability of the femoral stem. However, this classification was described and validated with cemented femoral stems. We sought to assess reliability and validity of the Vancouver classification in patients with cementless femoral stems.

METHODS: This is a blinded radiographic study which included patients treated for Vancouver B cementless periprosthetic femur fractures between February 2007 and December 2017. Adult reconstruction-trained and trauma fellowship-trained orthopedic surgeons graded all preoperative radiographs using the Vancouver classification on 3 separate occasions. Interobserver and intraobserver reliability was assessed via the Fleiss' kappa statistic. Validity was assessed via accuracy between radiographic and intraoperative assessments. The Landis and Koch criteria were used to interpret the kappa values.

RESULTS: Fifty-three patients with Vancouver B fractures (B1, 8; B2, 45) around a cementless femoral stem were included in the study. Five reconstruction-trained and 5 trauma-trained orthopedic surgeons graded all radiographs. The interobserver reliability kappa value was 0.45 (moderate agreement), with all raters agreeing on only 43% of radiographs. Validity analysis showed demonstrated 79% agreement. Overall, 20% (range, 14%-24%) of unstable B2 fractures were misread as B1 fractures. Intraobserver reliability was 0.71 between readings.

CONCLUSION: The reliability of the Vancouver classification for cementless total hip arthroplasty is lower than previously described in cemented femoral stems. Radiographic assessment alone may be inadequate for determination of stability of cementless stems in periprosthetic femur fractures.

LEVEL OF EVIDENCE: Level III therapeutic study: retrospective comparative study.