The ABC classification has recently been proposed as a comprehensive classification system for posterior shoulder instability (PSI). The purpose of this study was to analyze the comprehensiveness as well as inter-rater and intrarater reliability of the ABC classification. All consecutive patients presenting with unidirectional PSI from June 2019 to June 2021 were included in a prospective study. No patients were excluded, leaving a consecutive series of 100 cases of PSI in 91 patients. All recorded clinical and imaging data were used to create anonymized clinical case vignettes, which were evaluated twice according to the ABC classification at the end of the recruitment period in random sequential order by 4 independent raters (2 experienced shoulder surgeons and 2 orthopedic residents) to analyze the comprehensiveness as well as inter-rater and intrarater reliability of the ABC classification for PSI and to describe differences in characteristics among subtypes. Group A was defined as a first-time singular PSI event <3 months in the past regardless of etiology and is further subdivided into type 1 and type 2 depending on the occurrence of a subluxation (A1) or dislocation (A2). Group B comprises recurrent dynamic PSI regardless of time since onset and is further subdivided by the cause of instability into functional (B1) and structural (B2) dynamic PSI. Group C includes chronic static PSI with posterior humeral decentering that can be either constitutional (C1) or acquired (C2). None of the cases was deemed unsuitable to be classified based on the proposed system by the observers. After consensus agreement between the 2 expert raters, 16 cases were attributed to group A (8 type A1 and 8 type A2); 64, to group B (33 type B1 and 31 type B2); and 20, to group C (11 type C1 and 9 type C2). The expert raters agreed on the classification subtypes in 99% and 96% of the cases during the first rating and second rating, respectively (intraclass correlation coefficients [ICCs], 0.998 and 0.99, respectively). The intraobserver reliability was excellent for both raters. The beginners reached the same conclusion as the consensus agreement in 94% of the cases (ICC, 0.99) and 89% of the cases (ICC, 0.97) during the first round and 94% each (ICC, 0.97) during the second round. The intraobserver reliability was excellent for both beginners. Overall, discrepancies between raters were found between groups B1 and B2 (n = 14), groups B2 and C2 (n = 4), groups B1 and C1 (n = 1), and groups A1 and B2 (n = 1). In general, each subtype showed distinctive clinical and imaging characteristics that facilitated the diagnosis. The presented ABC classification for PSI is a comprehensive classification with a high reliability and reproducibility. However, a gradual transition and potential progression between the subtypes of PSI must be considered. The reliable distinction between different subtypes of PSI based on etiology and pathomechanism provides a standardized basis for future investigations on treatment recommendations.
Copyright remains with the authors and publisher.
The ABC classification has recently been proposed as a comprehensive classification system for posterior shoulder instability (PSI). The purpose of this study was to analyze the comprehensiveness as well as inter-rater and intrarater reliability of the ABC classification. All consecutive patients presenting with unidirectional PSI from June 2019 to June 2021 were included in a prospective study. No patients were excluded, leaving a consecutive series of 100 cases of PSI in 91 patients. All recorded clinical and imaging data were used to create anonymized clinical case vignettes, which were evaluated twice according to the ABC classification at the end of the recruitment period in random sequential order by 4 independent raters (2 experienced shoulder surgeons and 2 orthopedic residents) to analyze the comprehensiveness as well as inter-rater and intrarater reliability of the ABC classification for PSI and to describe differences in characteristics among subtypes. Group A was defined as a first-time singular PSI event <3 months in the past regardless of etiology and is further subdivided into type 1 and type 2 depending on the occurrence of a subluxation (A1) or dislocation (A2). Group B comprises recurrent dynamic PSI regardless of time since onset and is further subdivided by the cause of instability into functional (B1) and structural (B2) dynamic PSI. Group C includes chronic static PSI with posterior humeral decentering that can be either constitutional (C1) or acquired (C2). None of the cases was deemed unsuitable to be classified based on the proposed system by the observers. After consensus agreement between the 2 expert raters, 16 cases were attributed to group A (8 type A1 and 8 type A2); 64, to group B (33 type B1 and 31 type B2); and 20, to group C (11 type C1 and 9 type C2). The expert raters agreed on the classification subtypes in 99% and 96% of the cases during the first rating and second rating, respectively (intraclass correlation coefficients [ICCs], 0.998 and 0.99, respectively). The intraobserver reliability was excellent for both raters. The beginners reached the same conclusion as the consensus agreement in 94% of the cases (ICC, 0.99) and 89% of the cases (ICC, 0.97) during the first round and 94% each (ICC, 0.97) during the second round. The intraobserver reliability was excellent for both beginners. Overall, discrepancies between raters were found between groups B1 and B2 (n = 14), groups B2 and C2 (n = 4), groups B1 and C1 (n = 1), and groups A1 and B2 (n = 1). In general, each subtype showed distinctive clinical and imaging characteristics that facilitated the diagnosis. The presented ABC classification for PSI is a comprehensive classification with a high reliability and reproducibility. However, a gradual transition and potential progression between the subtypes of PSI must be considered. The reliable distinction between different subtypes of PSI based on etiology and pathomechanism provides a standardized basis for future investigations on treatment recommendations.
Copyright remains with the authors and publisher.