BackgroundPeri-prosthetic hip fractures (PPHFs) are serious complications whose treatment is generally difficult due to their predominance in elderly patients with bone frailty and other comorbidities. The Vancouver classification is the most widely used and is helpful for assisting treatment decisions. However, its value for predicting morbidity and mortality has not been assessed. The objective of this retrospective study was to assess post-operative morbidity and mortality according to the fracture type in the Vancouver classification. HypothesisPost-operative morbidity and mortality vary across fracture types in the Vancouver classification. MethodsA single-centre retrospective study was conducted from 1st January 2010 to 31st December 2015. All patients who had surgery for a PPHF were included. There were 88 patients, including 66 (75%) females, and mean age was 82 years. The patients were re-evaluated at least 3 years after surgery. The distribution of the fracture types was as follows: Vancouver (V) A, n=7; VB, n=63 (VB1, n=30; VB2, n=23; and VB3, n=10); and VC, n=18. Data on the pre-operative status (self-sufficiency, comorbidities, ASA score, etc.) were extracted from the admission files. Morbidity and mortality were evaluated globally and according to the Vancouver type, using the patient files and telephone calls to determine self-sufficiency scores (Parker, Katz, and Lawton) and functional scores (Merle d’Aubigné-Postel [MAP] score and Harris Hip Score [HHS]). ResultsPost-operative medical complications were very common (33 [37.5%] patients) and correlated with the severity of the fracture. Similarly, the mortality rate at last follow-up varied significantly (p<0.05) with the severity of the fracture, as follows: VA, 28.5%; VB1, 40%; VB2, 47.8%; VC, 55.6%; and VB3, 66.7%). In the overall population, loss of self-sufficiency was 20%, 14%, and 26% according to Parker, Katz, and Lawton, respectively; loss of function was 13.9% and 13.3% according to the MAP score and HHS (p<0.05). All the self-sufficiency scores (Parker, Katz, and Lawton) and functional scores (MAP and HHS) decreased post-operatively in proportion to the severity of the fracture (very small losses for VA and greatest losses for VB3) (p<0.05). ConclusionThe short- and medium-term mortality rates in our cohort of patients with PPHFs were high and chiefly dependent on the severity of the fractures. The self-sufficiency and functional scores were better in the group with VA fractures than in the groups with VB1, VB2, VB3, and VC fractures. In any case, early weight-bearing is without doubt a key factor in limiting the impact of PPHFs on the functional outcome and on mortality. Level of evidenceIV, retrospective study.