Long-term outcome of a proximal humerus fracture predicted after 1 year

Abstract

Background Earlier reports of results after proximal humerus fractures have often been short- or medium-term studies. The aim of this investigation was to follow an unselected group of patients prospectively and assess the long-term results in relation to fracture type and health, and to evaluate the possibility of predicting the long-term results after 1 year. Method 258 consecutive urban patients who sustained a proximal humerus fracture in 1987 were followed prospectively and examined after 1 and 13 years. 158 patients had died by the year 2000 and 47 of the surviving patients were available for clinical and radiographic evaluation. The Constant-Murley score (CM-score) was used in the clinical assessment. In the final follow-up group, all patients except 1 were treated non-operatively. Results 21 patients had asymptomatic, pain-free shoulders. In 26 patients with shoulder symptoms, 11 had severe pain and 10 had moderate pain, and only 2 of these 21patients had normal shoulder function. The remaining 5 patients in the symptomatic group had no shoulder pain but had reduced shoulder function. Shoulders with a displaced fracture and also injured shoulders of patients with a chronic disease were significantly worse at the long-term follow-up. The 1-year examination had been able to predict the long-term results with 71% sensitivity for protracted pain and 88% sensitivity for persistent shoulder dysfunction. Interpretation Our results suggest that there is a substantial mortality in patients with a proximal humerus fracture, as we have previously reported, and that surviving patients frequently have persistent symptoms that can be predicted as early as after 1 year.