Swimmer arm-to-shoulder test for early differentiation between shoulder and cervical spine pain

Key points

1. 01SAS test achieved 89.2% sensitivity and 96.1% specificity for identifying shoulder pathology versus cervical radiculopathy
2. 02Positive likelihood ratio of 22.6 makes a positive SAS test strongly confirmatory for shoulder pathology
3. 03Negative likelihood ratio of 0.11 makes a negative SAS test useful for ruling out shoulder pathology
4. 04The test is patient-performed without examiner interference, reducing examiner bias
5. 05Performance was consistent across age and sex subgroups in the study population

How it was conducted

Design

Prospective diagnostic accuracy study, single center, January 2018 to December 2021

Participants

718 consecutive patients aged 40-65 years with unilateral localized shoulder girdle pain lasting at least 12 weeks

Index test

Swimmer Arm-to-Shoulder (SAS) test: arm abduction at 90 degrees with elbow flexed, followed by horizontal adduction (swimmer strike) to touch opposite shoulder with hypothenar eminence; positive if shoulder pain VAS > 3/10

Reference standard

Clinical diagnosis confirmed by two independent rheumatologists (at least 10 years experience) using history, clinical examination, provocative testing, imaging (plain radiograph or MRI), and EMG where indicated

Primary outcome

Sensitivity, specificity, PPV, NPV, positive and negative likelihood ratios, and overall accuracy of SAS test versus reference diagnosis

What they found

• Clinical diagnosis identified shoulder pathology in 288 patients (40.1%) and cervical spine pathology in 430 patients (59.9%)
• SAS test was positive in 274 patients (38.2%): 257 true positives for shoulder pathology versus 17 false positives in cervical spine patients
• Sensitivity: 89.2% (95% CI: 85.0-92.6%); Specificity: 96.1% (95% CI: 93.8-97.7%)
• PPV: 93.8% (95% CI: 90.5-96.0%); NPV: 93.0% (95% CI: 90.5-94.9%)
• Positive likelihood ratio: 22.6 (95% CI: 14.1-36.0); Negative likelihood ratio: 0.11 (95% CI: 0.08-0.16)
• Overall accuracy: 93.3% (95% CI: 91.2-95.0%)
• Kappa agreement with gold standard diagnosis: 0.24, p < 0.001 (fair agreement by kappa scale)
• In pilot sample, test was positive in 31 of 50 shoulder pain patients (62%) versus 0 of 50 healthy controls

Limitations

• Single-center study limits generalizability to other clinical settings and populations
• Test result relies on subjective pain perception (VAS score), an inherent limitation of provocative clinical tests
• Kappa value of 0.24 indicates only fair agreement with the reference standard despite high sensitivity and specificity statistics
• Patients with concurrent shoulder and cervical pathology were excluded, which may overestimate real-world performance

Why it matters

For patients

A patient with shoulder pain can self-perform this simple arm movement test in the clinic to help the clinician quickly decide whether the pain is coming from the shoulder or the neck, potentially speeding up diagnosis.

For clinicians

The SAS test's high positive likelihood ratio (22.6) and low negative likelihood ratio (0.11) make it a practical triage tool to distinguish shoulder impingement from cervical radiculopathy before ordering imaging, complementing existing provocative tests.

For readers

This study introduces and validates a novel, examiner-independent clinical test with strong diagnostic accuracy in a large prospective cohort, adding a useful tool to the shoulder and cervical spine examination toolkit.