Introduction: The Constant- and the Neer-Score are widely used to assess shoulder function after trauma or shoulder diseases. The objective of this study was to compare the correlation of score result with the patient subjective assessment. We hypothesized that there is a clinically relevant difference between the score result and the patient assessment, especially for highly selective patient groups.
Methods: 51 patients were followed up after the surgical treatment of a proximal humeral fracture. For each patient the Constant- and the Neer-Score was calculated. Further, the patients were asked for a subjective assessment of their shoulder function ("excellent", "good", "fair", "poor").
Results: For both score-systems a good linear correlation (r = 0.97) is shown. 57% of the patients assessed their shoulder function as "excellent" or "good", but only 37% (Constant) vs. 43% (Neer) of the patients were classified as "excellent" or "good" based on their score results. The Spearman correlation of subjective and score-based assessment was just fair with r = 0.50 (Constant) and r = 0.55 (Neer). When comparing the score results with the patient subjective assessment for the groups: "excellent", "good", "fair" and "poor", there was a positive correlation for the Neer-Score in 20 cases and for the Constant-Score in 15 cases. Of all elderly patients (> 60 years, n = 20), 14 (Neer-Score) vs. 16 (Constant-Score) assessed their shoulder function as better than the score did.
Conclusion: The hypothesis of a clinically relevant difference between the subjective and the score-based assessment of shoulder function can be confirmed. Our results suggest that for clinical practice, statements and therapy strategies recommended in the literature, that are just based on score results might not be valid for all patient-groups (e.g. elderly patients).