RESUMEN
INTRODUCTION: The Constant score, allows an objective and subjective assessment of the shoulder function. It has been proven to have a poor interobserver reliability for some of its aspects and is not usable as a remote assessment tool. HYPOTHESIS: The Constant-Murley functional shoulder score can be assessed with a self-administered questionnaire. METHODS: We conducted a prospective continuous study in a shoulder-specialized service. For each patient seen in consultation or hospitalized for a shoulder pathology, a self-administered questionnaire was delivered, and a clinical examination was performed by a surgeon. The questionnaire, in French language, was composed of checkboxes only, with pictures preferred over text for most items. Correlations with surgeon examination were assessed with the intraclass correlation coefficients, differences with the paired t-test. RESULTS: One hundred consecutive patients were analyzed. Correlation between the two scores was excellent (0.87), as were the range of motion and the pain subscores (0.85 and 0.78), good for the activity (0.69) and fair for the strength (0.57). The mean total score was 3 points lower for the self-administered questionnaire (CI95 [-5; -1]; p<0.01). Activity and pain were not significantly different (-0.4/20 and -0.3/40; p>0.05) but pain and force were slightly different (+0.8/15; -3.0/25; p<0.01). CONCLUSION: The Auto-Constant questionnaire in French is an excellent estimator of the Constant score, and of its pain and mobility sub-scores. It is less accurate for the evaluation of the strength, but differences between sub-scores compensate and allow its use in daily practice. LEVEL OF PROOF: II, Prospective continuous clinical series.