[Prognostic factors during rehabilitation after shoulder prostheses for fracture]

To evaluate the role, the difficulties of rehabilitation and to diagnose the eventual surgical complications after shoulder prosthesis for 4-part fractures.

Forty three patients (46 shoulders) who underwent shoulder arthroplasty after fracture of the proximal humerus underwent rehabilitation and follow-up at a special reeducation center for an average of 3 months (1 to 6). There were 42 four-part fractures (with 22 fracture-dislocation) and 4 three-part fractures. The patients were send by five different hospitals and have all been operated by senior surgeons. Three types of implants were used: the Modular Shoulder prosthesis (27 cases), the Global prosthesis (2 cases), and the Aequalis prosthesis (17 cases). The rehabilitation followed the protocol recommended by Neer (recovery of passive joint movements, muscular strengthening and stretching) to which were added hydrotherapy, physiotherapy and occupational therapy. Forty patients (43 épaules) were reviewed and radiographed with an average follow-up of 29 months (18 to 72 months).

The functional results were disappointing with a normalised Constant score of only 60.2 per cent and an average active elevation of only 96 degrees. There appeared to be two factors which explained these poor results. Firstly, the advanced age of the population (52 per cent older than 70) who was often poorly or non-motivated (22 per cent) and debilitated (21 per cent chronic alcoholics) and who had significant medical and neuro-psychiatric histories. Secondly, incompletely resolved anatomical and surgical problems: damage to the circumflex nerve (6.5 per cent), early migration of the greater tuberosity (6.5 per cent), secondary migration with malunion (15 per cent) and/or nonunion (11 per cent) of the greater tuberosity. Migration of the greater tuberosity should be suspected clinically in three circumstances: 1) in patients who have an abnormally painful shoulder in the immediate post-operative period (16 cases in our series); 2) when there is no progression (24 per cent) or regression (9 per cent) of active shoulder mobility after three months of correct supervised rehabilitation; 3) later, if there is a dissociation between active anterior elevation (deficient) and passive anterior elevation (preserved).

The age and poor general condition of the patients as well as the difficulty of the surgical technique more than the rehabilitation, explain the disappointing results observed after shoulder prosthesis for four-part fractures. The discrepancy between active and passive elevation suggests that limited motion is not caused by a stiff shoulder because of glenohumeral scarring but instead by weakness of the deltoid (because of axillary lesion) and/or of the external rotators (because of greater tuberosity migration). There is some discordance between the necessity to early mobilise the shoulder and the high rate of tuberosity migration.