Treatment of proximal humerus fractures (PHF) is controversial in many respects, including the choice of surgical approach for fixation when using a locking plate. The classic deltopectoral (DP) approach is believed to increase the risk of avascular necrosis while making access to the greater tuberosity more difficult. The deltoid split (DS) approach was developed to respect minimally invasive surgery principles. The purpose of the present study (NCT-00612391) was to compare outcomes of PHF treated by DP and DS approaches in terms of function (Q-DASH, Constant score), quality of life (SF12), and complications in a prospective randomized multicenter study.

From 2007 to 2016, all patients meeting the inclusion/exclusion criteria in two University Trauma Centers were invited to participate in the study. Inclusion criteria were: PHF Neer II/III, isolated injury, skeletal maturity, speaking French or English, available for follow-up (FU), and ability to fill questionnaires. Exclusion criteria: Pre-existing pathology to the limb, patient-refusing or too ill to undergo surgery, patient needing another type of treatment (nail, arthroplasty), axillary nerve impairment, open fracture. After consent, patients were randomized to one of the two treatments using the dark envelope method. Pre-injury status was documented by questionnaires (SF12, Q-DASH, Constant score). Range of motion was assessed. Patients were followed at two weeks, six weeks, 3-6-12-18-24 months. Power calculation was done with primary outcome: Q-DASH.

A total of 92 patients were randomised in the study and 83 patients were followed for a minimum of 12 months. The mean age was 62 y.o. (+- 14 y.) and 77% were females. There was an equivalent number of Neer II and III, 53% and 47% respectively. Mean FU was of 26 months. Forty-four patients were randomized to the DS and 39 to the DP approach. Groups were equivalent in terms of age, gender, BMI, severity of fracture and pre-injury scores. All clinical outcome measures were in favor of the deltopectoral approach. Primary outcome measure, Q-DASH, was better statistically and clinically in the DP group (12 vs 26, p=0,003). Patients with DP had less pain and better quality of life scores than with DS (VAS 1/10 vs 2/10 p=0,019 and SF12M 56 vs 51, p=0,049, respectively). Constant-Murley score was higher in the DP group (73 vs 60, p=0,014). However, active external rotation was better with the DS approach (45° vs 35°). There were more complications in DS patients, with four screw cut-outs vs zero, four avascular necrosis vs one, and five reoperations vs two. Calcar screws were used for a majority of DP fixations (57%) vs a minority of DS (27%) (p=0,012).

The primary hypothesis on the superiority of the deltoid split incision was rebutted. Functional outcome, quality of life, pain, and risk of complication favoured the classic deltopectoral approach. Active external rotation was the only outcome better with DS. We believe that the difficulty of adding calcar screws and intramuscular dissection in the DS approach were partly responsible for this difference. The DP approach should be used during Neer II and III PHF fixation.