A pragmatic regional interdependence approach to primary frozen shoulder: a retrospective case series

Key points

1. 01All five patients regained near-normal shoulder passive range of motion (mean flexion 179 degrees, abduction 175 degrees, external rotation 89 degrees) by discharge
2. 02Effect sizes for range of motion gains were large: flexion d=5.7, abduction d=11.9, external rotation d=13.2
3. 03DASH disability score improved from a mean of 40.3 to 6.2 (large effect size d=2.4) at 32 weeks follow-up after discharge
4. 04Treatment addressed four upper quarter regions: shoulder joint, shoulder girdle, scapulothoracic and humerothoracic muscles, and spine
5. 05Current clinical practice guidelines only recommend glenohumeral-focused treatments; this regional approach adds a broader targeting strategy

How it was conducted

Design

Retrospective case series of de-identified data from a single outpatient physical therapy practice

Participants

5 consecutive patients (ages 40-66, mean 50.2 years) diagnosed with primary frozen shoulder, no prior or concurrent shoulder care

Symptom duration before diagnosis

2-30 weeks (mean 14.2 weeks)

Treatment

Pragmatic regional interdependence approach: joint mobilization, HVLAT, MWM, MET, SCS, soft tissue mobilization, stretching, strengthening, patient education to shoulder, shoulder girdle, scapulothoracic, and spinal regions

Sessions and duration

11-21 sessions (mean 14.8) over 5-10 weeks (mean 7.6 weeks)

Primary outcomes

Passive shoulder ROM at discharge; DASH functional disability score at follow-up (minimum 12 weeks post-discharge)

What they found

• Mean passive flexion improved from 117 +/- 10 degrees to 179 +/- 12 degrees (d=5.7, CI95%: 2.9-8.5)
• Mean passive abduction improved from 74 +/- 8 degrees to 175 +/- 9 degrees (d=11.9, CI95%: 6.6-17.3)
• Mean passive external rotation improved from 23 +/- 7 degrees to 89 +/- 2 degrees (d=13.2, CI95%: 7.3-19.0)
• All ROM gains exceeded the minimal clinically important difference (11-16 degrees) and standard error of measurement (4-7 degrees) for shoulder ROM
• DASH score improved from 40.0 +/- 19.4 at initial evaluation to 6.2 +/- 3.7 at follow-up (mean 32.4 weeks post-discharge; d=2.4, CI95%: 0.9-4.1)
• 4 of 5 patients had DASH improvements exceeding the SEM (4.6) and MCID (10.8)
• Follow-up DASH assessments were conducted at 13-56 weeks after discharge (mean 32.4 +/- 16.3 weeks)

Limitations

• Very small sample (n=5) with no control group, making it impossible to separate treatment effects from natural history or placebo
• Retrospective design limited access to clinical decision-making details and did not allow collection of missing data such as pain and internal rotation at all time points
• Pragmatic approach meant each patient received different specific techniques, reducing the ability to attribute outcomes to any single intervention
• Single treating clinician with >20 years of experience may limit generalizability; assessor bias cannot be excluded

Why it matters

For patients

People diagnosed with frozen shoulder may benefit from asking their physiotherapist about treatment that includes the neck, mid-back, and shoulder blade rather than only the shoulder joint, though this study is too small to confirm the approach works for everyone.

For clinicians

This case series suggests that a regional interdependence protocol addressing the cervicothoracic spine, shoulder girdle, and glenohumeral joint may restore near-normal passive ROM in 7-10 weeks, outperforming outcomes typical of glenohumeral-only interventions in the literature, and warrants a prospective controlled trial.

For readers

The study challenges the widespread assumption that frozen shoulder resolves on its own and provides a rationale and treatment template for a multi-region physical therapy approach, though evidence is currently at Level 4 only.