IntroductionTerrible triad (TT) of the elbow almost always requires surgery to prevent progression to chronic instability and more or less inevitable osteoarthritis. Progression toward osteoarthritis after well-conducted surgery and associated risk factors have been little reported. We performed a retrospective study: (1) to assess rates of post-traumatic elbow osteoarthritis after surgical treatment of TT; (2) to assess functional impact; and (3) to identify prognostic factors. HypothesisPrevalence of osteoarthritis after surgical treatment of TT is high, impairing functional results. Material and methodA single-center retrospective study included 53 patients, with a mean age of 50±17.8 years (range, 21–84 years), undergoing surgery for acute TT in our department. All received clinical examination with ranges of motion and Mayo Elbow Performance Index (MEPI) and radiographic assessment at a minimum 1 year's follow-up. Osteoarthritis at last follow-up was assessed on elbow X-ray in the humero-ulnar and radio-condylar compartments on the Broberg–Morrey classification. Functional impact on range of motion and MEPI and prognostic factors were assessed on Student test or ANOVA and Chi2 or Fisher test. ResultsPrevalence of Broberg–Morrey grade 2 or 3 osteoarthritis was 45.3% (24/53) in the humero-ulnar compartment and 50% (25/50) in the radio-condylar compartment. Humero-ulnar osteoarthritis impaired MEPI (76.3 points with versus 88.4 points without; p=0.003), flexion-extension (102.3° versus 115.2°; p=0.043) and pronation-supination (138.8° versus 159.3°; p=0.006). Radio-condylar osteoarthritis had no significant impact on MEPI (81.4 points with and 84.4 points without; p=0.47), flexion-extension (104.8° and 113°; p=0.23) or pronation-supination (141.8° and 156.4°; p=0.2). Humero-ulnar osteoarthritis at last follow-up was associated with dislocation or subluxation on immediate postoperative lateral view (45.8% with versus 10.3% without; p=0.004) and at last follow-up (20.8% versus 3.4%; p=0.047) and with postoperative complications (54.2% and 27.6%; p=0.049). Radio-condylar osteoarthritis at last follow-up was associated with radial head replacement rather than internal fixation (respectively, 92% and 48%; p=0.0007) and excessively high radial head implant positioning (47.8% versus 0%; p=0.023). ConclusionPrevalence of traumatic osteoarthritis after TT surgery was high, at 45.3% in the humero-ulnar compartment and 50% in the radio-condylar compartment, with clinical impact in humero-ulnar involvement. Level of evidenceIV; cohort study without control group.