Purpose of the study: The articulated external fixator of the elbow joint is often useful for the treatment of trauma victims. It can neutralise dislocation forces and protect osteosynthesis assemblies and ligament repairs while authorising early mobilisation. This work reports our indications and application of this type of fixator, as well as the expected clinical and radiographic outcomes.

Material and methods: From 1995 to 2008, 34 patients had an external fixator of the elbow in our unit, in combination with classical treatment. Two groups of patients were distinguished, those with a traumatic injury requiring emergency care (n=15, group 1: six dislocations, two fracture-dislocations, and seven complex fractures), and those treated outside an emergency context (n=10, group 2: ten chronic dislocations or subluxations, four stiff joints, one infection, four material disassemblies). Eighteen patient were reviewed retrospectively, clinically and radiographically. The DASH score and the Mayo Elbow Performance Score (MEPS) were noted. The Broberg and Morrey classification was used for osteoarthritis.

Results: At mean 4.3 years follow-up, for groups 1 and 2, the DASH scores were 35 and 25 points and the MEPS scores 74 and 74 points respectively. In group 1, the range of motion was 63° for flexion-extension; the elbow was centred and stable in all cases except 2 (one posterior subluxation). Six elbows presented moderate to severe osteoarthritis. In group 2, the range of motion was 80° flexion-extension; the elbow was centred and stable in all cases except one (one posterior subluxation). Moderate to severe osteoarthritis was noted in five elbows. There were four complications: two cases of transient (ulnar and radial) paralysis, one fracture of the humerus on a pin track, and one superficial pin track infection.

Discussion: This study demonstrates that the articulated external fixation can maintain the reduction during the healing process for complex elbow trauma where stability is compromised. The morbidity is low and functional outcomes favourable. Early mobilization of these injured elbows can limit secondary stiffness. The prognosis remains a function of the initial injury and the quality of the associated treatments.

Purpose of the study: Tibial osteotomy to correct for varus deformity is a well defined procedure. Survival has reached 80% at ten years. Nevertheless, a number of early failures are related to inadequate initial correction. Computer assisted surgery has demonstrated its efficacy for knee arthroplasty. We hypothesised that it could also improve the reliability of correction for tibial osteotomy.

Material and method: From 2007, in a prospective case-control study, 34 tibial wedge osteotomies were performed, 17 were computer assisted (Navitrack, Orthosoft) with plate fixation (Tomofix, Synthès) without wedge insertion; the objective was valgus measuring 2 to 5°.

Results: The two series were comparable for age (54.2±6 and 55.7±4.5), body mass index (28.9±6.2 and 28.7±5.7), and varus deformity (7.2±3 and 6.2±6) respectively in the standard and navigated groups. Osteoarthritis was more severe in the navigated group, with five patients stage 2 and 12 stage 2 versus one stage 1, 12 stage 2 and 4 stage 3 in the standard group (p=0.0152). The duration of the operation was not longer in the navigation group (p)0.2779). Comparisons were made for alignment at three months, between the groups and in relation to the preoperative data. There was no significant difference between the intraoperative navigation alignment and the alignment measured at 3 months: 3.6±6 and 2.5±3 at 3 months (p=0.2187). At 3 months, there was no significant difference in alignment between the two groups with 3.22 and 2.5±1.6 valgus in the standard and navigation groups respectively (p=0.2136). The objective was achieved in 25 patients: 12 in the standard group and 13 in the navigated group. In the navigation group, there were four failures, no cases of over correction, two cases of insufficient valgus at 1.5, one neutral alignment, and one recurrent varus. In the standard group, there were five failures with two over corrections at 7 and 8, two under corrections at 0 and 1, and 1 recurrent varus at 4.

Discussion: We were unable to prove that navigation improves the reliability of the correction but it did appear to avoid important errors, particularly over correction. Few series have compared standard varus navigated osteotomies, and all published series have been small. Our study has the advantage of being monocentric with two comparable series of patients. The sample size nevertheless remains small and the follow-up short.

Purpose of the study: Propionibacterium acnes (PBA) is an anaerobic Gram-positive commensal bacillus of human skin which can cause bone and joint infections (Lutz 2005, Zeller 200, Levy 2008). The purpose of this work was to evaluate over a given period the frequency of PBA infections and the reality of its role as a pathogenic organism.

Material and methods: A retrospective survey of activity from 2006 to 2008 using bacteriology laboratory data identified 34 patients (22 male and 12 female) with at least one sample collected during an orthopaedic or traumatology surgical procedure that was positive for PBA. The mean number of positive samples was 3.6; 17 from the thoracic limb, 17 from the pelvic limb, involving 16 arthroplasties (4 hips, 6 knees, 5 shoulders, 1 elbow), 13 osteosynthesis procedures, 3 cuff repairs and 1 acromioclavicular procedure. Six of 20 histology samples showed a septic granuloma.

Results: The PBA was the only germ isolated in 18 cases; it was associated with other bacteria in 16 cases. Other blood tests were abnormal (WBC 6800 leukocytes, CRP 25mg/L en average). According to the Lutz classification, three groups of infection could be identified: certain infection with clinical signs and at least two positive samples (n=12), possible with clinical signs but only one positive sample (n=5), and probable without clinical signs and one or more positive samples (n=17). Of the 12 patients in the first group, eight had material (three shoulder prostheses, three hips, one knee and one femoral nail), which had to be removed for six with use of a cemented spacer in four. Mean duration of antibiotics was five weeks. Four patients in this group have not yet achieved cure. In the second group, all samples were taken from a thoracic limb and had another germ in four cases; all patients have achieved cure. In the third group, the samples were systematic (two shoulder arthroplasties, two repeated cuff repairs, five revision prostheses, four nonunions and four material removal); only one sample was positive in ten cases and only five patients were given antibiotics; all achieve cure.

Discussion: This series is in agreement with the literature: frequent localization on a thoracic limb, association with another germ, questionable attribution to PBA. Patients meeting the criteria of the first group should be treated. If a PBA infection is suspected, samples should be repeated, with prolonged culture; this attitude should be validated prospectively.

Purpose of the study: Locked centromedullary nailing (LCMN) has become the gold standard treatment for fractures of the femur (I. Kempf, Chirurgie 91 ; 117 : 478 / Borel R.C.O. 93 ; 79,553 / Wolinsky J. trauma 99, 46 : 382). Nevertheless, the SOFCOT 2004 round table emphasized the frequency of complications related to inappropriate material and techniques. The series studied had several biases: multicentric recruitment, materials with different designs, high percentage of patients lost to follow-up. In order to overcome these shortcomings we reviewed retrospectively a consecutive series of LCMN performed in a single center from 2001 to 2002, attempting to be as exhaustive as possible.

Material and methods: The study group included 78 patients (81 LCMN) aged 30 years on average (range 16–87 years) with male predominance (69%). A large proportion of patients were traffic accident victims (44% two-wheel vehicles, 42% four-wheel vehicles). The fracture was open in 8% and 65% of patients had multiple fractures (11% floating knees, 23% multiple trauma). The fractures were simple (43%), wedge (47%) and comminutive (10%). Time to operation was 7.4 hr on average for 83% of patients. Mean reaming was 12 mm (range 11–14 mm). An 11-mm (range 10–13 mm) static Grosse and Kempf (Stryker) nail was used in all patients. Intraoperative complications occurred in 8% of cases with no effect on bone healing.

Results: Three patients died from severe head trauma. Five patients were lost to follow-up. Written follow-up data were available for six patients and 64 patients were reviewed clinically. Among the 70 fractures with known outcome, four had not healed (with two screw failures and one nail failure). Knee motion was normal in all patients. Anteroposterior and lateral alignment was normal (±5°) in 94% with no leg length discrepancy (< 10 mm)in 87%. The nail was withdrawn in 84% of patients and the withdrawal procedure was complicated in three cases (hematoma, screw failure). Mean hospital stay for single-fracture patients was 9.7 days for nailing and 2.2 days for nail removal.

Discussion: LCMN is a reliable technique which provides constant clinical results when applied with rigorous technique. The logistics is resource intensive. Nonunion can be revised with the same method. A new nailing with second reaming should be performed early in the event of late healing.