Is routine magnetic resonance imaging necessary in patients with clinically diagnosed frozen shoulder? Utility of magnetic resonance imaging in frozen shoulder

Abstract

BackgroundShoulder magnetic resonance imaging (MRI) is commonly performed in patients with frozen shoulder (FS). However, the necessity of MRI and its diagnostic value is questionable. Therefore, the purpose of the present study was to clarify whether routine MRI could identify additional shoulder pathologies not previously suspected in the clinical examination and if any change in the treatment plan based on these additional MRI findings in FS patients was observed.Materials and methodsThe medical records of all patients who presented in our outpatient clinic with a diagnosis of FS from January 2017 to December 2018 were retrospectively reviewed. Patient demographics, the number of patients who received a shoulder MRI, changes in the diagnosis or identification of structural shoulder pathologies following MRI examination (if performed), as well as any alternation in the initially suggested treatment plan were recorded.ResultsA total of 609 patients (male: 241, female: 368) diagnosed with an FS and an average age of 52 ± 10 (range: 18 to 81) years were identified. In 403 of the 609 patients (66%), a shoulder MRI was performed. An additional structural shoulder pathology was identified in 89 of 403 (22%) patients following the shoulder MRI, mostly rotator cuff tears (partial: 46/403 [11.4%], full-thickness: 30/403 [7.4%], rerupture following reconstruction: 10/403 [2.5%]) and labrum tears (3/403 [0.7%]). At minimum 2-year follow-up, 11 of 403 (2.7%) patients were treated surgically for the additional pathology identified on the MRI scan consisting of an arthroscopic rotator cuff reconstruction in 10 patients and a labrum refixation in one patient. Five of the 609 (0.8%) patients were treated for refractory FS by arthroscopic capsulotomy.ConclusionsAlthough additional pathologies were identified in 22% of the patients, a change in treatment plan due to the MRI findings was only observed in 2.7% (37 MRIs needed to identify 1 patient with FS requiring surgery for the additional MRI findings). Therefore, routine use of shoulder MRI scans in patients with FS but without suspicion of an additional pathology may not be indicated.