Objective: To assess intraobserver reliability of angular and linear clinical measurements of scapular position.
Design: Test-retest analyses.
Setting: Outpatient department in National Health Service teaching hospital in the United Kingdom.
Participants: Subjects (n=45) without symptoms (21 men, 24 women; age range, 23-56y) and 45 subjects (22 men, 23 women; age range, 19-84y) with shoulder symptoms (defined as pain in the C5-6 dermatome reproduced by shoulder movement and not reproduced with cervical movement).
Interventions: Not applicable.
Main outcome measures: Intraclass correlation coefficient (ICC) models 2,1 and 2,3, 95% confidence intervals (CIs), and SE of measurements for 68% confidence and 2 SEs of measurement (for the 95% CI) for the bilateral angular measurements of scapular rotation and tilt, and the bilateral linear measurements of lateral scapular displacement (protraction) and vertical displacement (elevation).
Results: For subjects without symptoms, ICC(2,3) results ranged from .75 to .98. The 2 SE results for the angular movements ranged from 1.8 degrees to 2.4 degrees and from 0.4 to 1.0 cm for the direct linear measurements. Subjects with symptoms: ICC(2,3) results ranged from .61 to .98. The 2 SE results for the angular movements ranged from 1.4 degrees to 2 degrees and from 0.6 to 1 cm for the direct linear measurements.
Conclusions: Repeated-measure (ICC(2,3)) results were more reliable than single-measure (ICC(2,1)) results. Very good to excellent intraobserver reliability was demonstrated for the angular and linear measurements of interest in both shoulders of subjects with and without symptoms. The 2 SE results provide guidance about the error associated with the individual measurements and will assist the clinician determining whether a change in the static position of the scapula has occurred as a result of intervention or over time. (National Research Register identifier N0060148286.).