Background: The clinical follow-up of patients for degeneration of the supraspinatus tendon is limited by the lack of objective assessment of pain evolution over time. We therefore tested a new method to collect follow-up data on patients treated either by surgical cuff repair or rehabilitation. Objectives: We report the feasibility this method in terms of recruitment of clinicians and patients and their compliance. Methods: In this multicenter longitudinal observational study, between September 2015 and March 2019, patients consulting either for surgical repair or rehabilitation were examined at baseline and after twelve months by their clinician, including the Mini-DASH questionnaire. Fortnightlys, during one year, patients were asked about number of days their shoulder problem affected their daily life, number of nights woken up from shoulder pain, and present pain score, using text-messages for sending and responding to questions. A system administrator supervised responses and non-compliant subjects were contacted and assisted with the procedure. The CONSORT statement for pilot studies was followed. Results: Four of 11 invited clinicians accepted participation and collected data till the end. Of the 410 patients we originally planned for, 252 were included in the study, but complete data for the clinicians’ follow-up at 12 months were missing for 30. Of the 222 subjects with SMS data files, 190 (85%) provided at least 80% of their fortnightly messages. All three SMS messages were answered equally often. In total, 160 study subjects answered at least 80% of times and had clinical data at twelve months, i.e. 39% of the intended study sample and 72% of the 222 subjects with SMS data. Conclusion: The most important difficulty of this study was the enrolment and compliance of clinicians. The collection of SMS data was less successful than in previous studies, but French people accepted well this new method which is much easier and specific than collecting data through clinical records. The quality of the SMS data was acceptable. However, because of the limited number of complete datasets, only a limited number of questions from the original study protocol can be answered.