Diagnosis of Subacromial Impingement
Begin with standard shoulder radiographs (AP views in internal and external rotation plus axillary or scapula-Y view), followed by MRI shoulder without IV contrast for soft tissue evaluation, or ultrasound if MRI is unavailable or contraindicated. 1
Initial Imaging Approach
Radiographs serve as the mandatory first step in evaluating shoulder pain, including suspected subacromial impingement 1. While radiographs primarily assess bony anatomy and alignment, they can reveal:
- Acromion morphology (particularly type III hooked acromion strongly associated with impingement) 2
- Acromioclavicular joint pathology
- Calcific tendinitis
- Bone spurs contributing to impingement
The supraspinatus outlet X-ray view demonstrates the highest diagnostic accuracy for acromion classification in subacromial impingement patients (88.7% overall consistency, Kappa=0.77), particularly for identifying type III hooked acromion (AUC=0.939). 2 This specialized view should be specifically requested when impingement is suspected.
Advanced Imaging Selection
MRI Without IV Contrast (Preferred)
MRI shoulder without IV contrast is the best modality for comprehensive soft tissue assessment 1, providing:
- High sensitivity and specificity for full-thickness rotator cuff tears (90-91% sensitivity, 93-95% specificity) 1
- Evaluation of subacromial bursa inflammation
- Assessment of supraspinatus tendon pathology
- Detection of bone marrow edema
- Identification of associated labral pathology
The key limitation is lower sensitivity for partial-thickness tears compared to MR arthrography 1.
Ultrasound (Alternative Option)
Ultrasound performs comparably to MRI for full-thickness rotator cuff tears (90-91% sensitivity, 93-95% specificity) 1, but has important caveats:
Use ultrasound preferentially when:
- Patient has proximal humeral hardware causing MRI artifacts 1
- MRI is contraindicated or unavailable
- Cost is a limiting factor
Avoid ultrasound when:
- Large body habitus limits visualization 1
- Restricted range of motion from acute pain 1
- Suspicion of intra-articular pathology beyond rotator cuff 1
- Partial-thickness tears are suspected (conflicting evidence on accuracy) 1
A practical ultrasound finding: subacromial impingement ratio ≥1.4 (bursa thickness during abduction divided by thickness during adduction) has 88.2% sensitivity and 96.3% specificity for impingement syndrome 3. The absolute bursa thickness cut-off of 1.3mm during abduction has lower accuracy (70.6% sensitivity, 85.2% specificity) 3.
Physical Examination Tests
While imaging confirms the diagnosis, physical examination helps guide the diagnostic workup:
Most useful individual tests for ruling in impingement:
- External rotation resistance test (positive likelihood ratio 4.39) 4
- Empty can/Jobe test (positive likelihood ratio 3.90) 4
- Painful arc test (positive likelihood ratio 2.25) 4
Most useful for ruling out impingement:
The diagnostic threshold of ≥3 positive tests out of 5 standard tests (Neer, Hawkins-Kennedy, painful arc, empty can, external rotation resistance) confirms subacromial impingement with an AUC of 0.79 4. However, recent data shows Neer's test alone has only 79% sensitivity and 63.2% specificity against MRI 5, emphasizing that physical examination cannot replace imaging.
The Yocum maneuver demonstrates the best overall sensitivity and accuracy for clinical diagnosis 6, though specificity remains limited across all physical examination maneuvers 6.
Common Pitfalls
- Do not rely on physical examination alone - most tests have low specificity despite acceptable sensitivity 6
- Do not order MRI with and without IV contrast - no evidence supports this for impingement evaluation 1
- Do not skip radiographs - they are essential for identifying bony causes and acromion morphology 1, 2
- Do not assume ultrasound can replace MRI for partial tears - evidence is conflicting and interobserver reliability is poor 1
- Recognize that MSCT and standard MRI have lower accuracy than specialized outlet views for acromion classification (62.3% and 58.5% consistency respectively vs. 88.7% for outlet radiographs) 2