(Golden Oldie) Observational scapular dyskinesis: known-groups validity in patients with and without shoulder pain

Key points

1. 01Scapular dyskinesis prevalence was not significantly different between the shoulder pain group and healthy controls during either flexion or abduction SDT.
2. 02Unblinded examiners (who knew the patient had shoulder pain) rated dyskinesis as present about 12% more often on the painful shoulder than blinded examiners did.
3. 03This confirmation bias suggests clinicians may over-identify scapular dyskinesis when they already know a patient is in pain.
4. 04Scapular dyskinesis may represent normal movement variability rather than a disorder-specific impairment.
5. 05Clinical decisions to treat scapular dyskinesis should not be based solely on a positive SDT result.

How it was conducted

Design

Cross-sectional known-groups validity study

Participants

135 adults: 67 with shoulder pain and 68 pain-free controls, recruited from 4 physical therapy clinics in Virginia and California

Examiners

21 physical therapists; one unblinded (treating clinician) and one blinded to pain status per patient

Test performed

Scapular Dyskinesis Test (SDT): 5 repetitions each of weighted shoulder flexion and abduction, rated as normal, subtle, or obvious dyskinesis; collapsed to dyskinesis vs. no dyskinesis for analysis

Primary outcomes

Prevalence of scapular dyskinesis in shoulder pain vs. control groups (Aim 1); prevalence differences between blinded and unblinded examiners in the shoulder pain group (Aim 2)

Sample size target

108 participants required (a priori power analysis, 80% power, alpha 0.05)

What they found

• Flexion SDT prevalence - blinded examiner: shoulder pain group 67.2% (95% CI: 0.55, 0.77) vs. control group 61.8% (95% CI: 0.50, 0.72); p=0.51 (not significant).
• Abduction SDT prevalence - blinded examiner: shoulder pain group 67.2% (95% CI: 0.55, 0.77) vs. control group 52.9% (95% CI: 0.41, 0.64); p=0.09 (not significant).
• Involved shoulder - flexion SDT: blinded examiner 67.7% (44/65; 95% CI: 0.56, 0.78) vs. unblinded examiner 80.0% (52/65; 95% CI: 0.69, 0.88); p<0.001.
• Involved shoulder - abduction SDT: blinded examiner 66.2% (43/65; 95% CI: 0.54, 0.76) vs. unblinded examiner 78.5% (51/65; 95% CI: 0.67, 0.87); p=0.001.
• Uninvolved shoulder - abduction SDT: unblinded 66.2% (43/65) vs. blinded 67.7% (44/65); p=0.008.
• Uninvolved shoulder - flexion SDT: unblinded 60.0% (39/65) vs. blinded 56.9% (37/65); p=0.222 (not significant).
• Sensitivity analysis using obvious-only as the dyskinesis threshold showed no significant differences between groups or between blinded and unblinded examiners.
• Pre-study inter-rater reliability on a subset of 5 clinicians: kappa=0.69 (95% CI: 0.36, 1.00), percent agreement 85%.

Limitations

• Shoulder diagnoses were not controlled for, so it is unclear whether specific pathologies (e.g., full-thickness rotator cuff tear) would show different dyskinesis prevalence.
• Participants had low pain and disability levels, limiting generalizability to more severely affected patients.
• 21 different physical therapists performed the SDT and variation between them was assumed to be random rather than measured systematically.
• A tally of screened but ineligible participants was not recorded, making selection bias difficult to assess.

Why it matters

For patients

If your clinician tells you that your scapula moves abnormally, this finding alone does not confirm it is causing your shoulder pain, as similar movement patterns are common in people without pain.

For clinicians

The SDT should not be used in isolation to drive scapular-focused treatment; consider blinded or structured assessment protocols and supplement with scapular symptom alteration tests to determine whether scapular movement is truly contributing to a patient's pain.

For readers

This study challenges the diagnostic value of observational scapular dyskinesis testing and highlights that examiner knowledge of patient symptoms can meaningfully inflate positive test rates, an important methodological concern for all observational clinical tests.