From January 2007 to December 2009, all consecutive patients that were treated in our institution with the Polarus nail for stabilisation of fractures of the proximal humerus were included in this retrospective study. Institutional board approval was obtained. Using true anteroposterior (AP) and y-views radiographs of the shoulder and, when necessary, CT scan to accurately assess the fracture pattern (in cases of comminuted fracture patterns), fractures were classified according to Neer’s classification . Inclusion criteria were displaced 2-part and 3-part fractures as well as some 4-part fractures where satisfactory reduction was possible by closed means. Exclusion criteria for the use of the Polarus nail were minimally displaced or undisplaced fractures of the proximal humerus, severely comminuted 4-part or 3-part fractures (especially in elderly patients with poor bone quality), head split fractures, fracture-dislocations, and impaction fractures with involvement of more than 40% of the articular surface. Osteoporosis and arthritis of the glenohumeral joint were also factors that were taken into account. In these cases, we used other fixation devices or hemiarthroplasty.
In all cases, a short Polarus nail was used. This device is a cannulated locked antegrade intramedullary humeral rod with a tapered profile to reduce distal stress concentration, and it has a spiral array of four 5-mm proximal screw-holes and two 3.5-mm distal interlocking holes [10, 14, 16, 17].
Passive physiotherapy was initiated from the 1st postoperative day, including pendulum motion and passive elevation and rotation. At the same time, all patients were encouraged to actively exercise the wrist and elbow joints. Subsequently, active assisted exercises after the 2nd week and active motion of the shoulder joint at 4 weeks was encouraged. Patients were trained in self-assisted shoulder abduction and elevation using the uninjured extremity as an assistant. Outpatient physiotherapy was initiated 2 weeks postoperatively. Exercises under resistance were restrained until the 6th postoperatively week or until fracture union was radiologically confirmed.
Following discharge from the hospital, patients were followed up in the outpatient clinic at 2 and 6 weeks, and at 3, 6, 12 months and yearly thereafter (maximum 3 years) as indicated. Anteroposterior and axillary radiographs were taken at each follow-up in order to evaluate the progress in fracture union and the presence of implant related complications. The time to union as well as all minor or major complications and their management were documented. Numerous other parameters were recorded and analysed including patient demographics, mechanism of injury, associated injuries, days to surgery and operative time. The mean follow-up time was 36 months (range of 27–43).
The functional outcome was assessed using the Constant shoulder scoring system  at the final follow up. The scoring system is constituted from 4 categories of interest: pain, activities of daily leaving, range of motion and strength. Pain is scored with a maximum of 15 points (no pain) whereas the activity of daily living maximum score is 20 points (no limitations in daily living and recreational activities, no night sleep disturbance and above head usage of the arm). The maximum score for the range of movement parameter is 40 points (evaluates forward flexion, abduction, internal and external rotation) whilst power is estimated by doubling the average kilos from 5 pulls and has a maximum score of 25 points. Scores below 50% were considered to be a poor result, between 50% and 75% a satisfactory result, and scores above 75% to be an excellent result [18, 20]. The strength in the shoulder was measured with an isometric tensiometer. Additionally, a comparative evaluation of the functional outcomes in patients older than 60 years in relation to the younger ones was performed using the Mann–Whitney U-test. Statistical analysis was performed using SPSS version 13.00 (Statistical Package for the Social Sciences, SPSS Inc., Chicago, Ill., USA) and a p-value of 0.05 or less was considered as statistically significant.
Finally, the current literature was reviewed and all relevant studies reporting on the clinical use of this fixation method were collected. Data regarding total number and mean age of patients, fracture patterns, union, functional outcome and scoring system used as well as complications rate were documented and analysed.
Institutional board approval was obtained from the local Trauma & Orthopaedic Directorate and the research carried out was in compliance with the Helsinki Declaration. Every patient consented to participate in this study for publication of this report and any accompanying images.