Shoulder Impingement

The VerdictMODERATE CONVICTION

"Shoulder impingement" is a 50-year-old label for a real pain pattern, and surgery to fix it does not beat sham surgery.

Stand sideways to a mirror. Slowly raise your affected arm out to the side from your hip toward overhead. If pain reproducibly appears between roughly shoulder height (about 60 degrees) and just above eye level (about 120 degrees) and eases above and below — that is a painful arc, the most efficient single shoulder self-screen.

What this actually is: a pain pattern in the rotator-cuff region with movement and load consequences, not a single mechanical pinch.
What most people get wrong: thinking that a positive single test (Hawkins-Kennedy, Neer's) is a diagnosis. It is a screening clue, not a verdict.
The first thing to start doing: graded loaded rehabilitation for 6 to 12 weeks under criteria-based progression — strength of the rotator cuff and scapula is the medicine.

A shoulder is not a hinge being pinched by a bone. It is a load-managed system. The pain is closer to a thermostat that has reset itself low after weeks of overload — not a pebble being pinched in a door.

What works

Graded loaded rehabilitation for the rotator cuff and scapula

Tier 1 — Graded loaded rehab as first-lineHIGH

Two to three sessions per week supervised, plus daily home work. Progress from isometric or low-load through graded rotator-cuff and scapular loading. Target eight to twelve reps at RPE seven to eight in the non-irritable phase. Six to twelve weeks for meaningful change.

Side-lying external rotation (light dumbbell) 3 × 12 · three times per week · effort in the back of the shoulder, no sharp pain

Resistance-band external rotation, elbow tucked 3 × 12 · three times per week · slow eccentric on the return

Scapular plane elevation, "thumb up" (light dumbbell) 3 × 10 · three times per week · stay below the painful arc until tolerated

Wall slide / Y-T-W (no weight) 3 × 8 each shape · daily · effort in the upper back, no pinching

Push-up plus (knees as needed) 3 × 8 · three times per week · scapular protraction at top of the rep

Tier 1 — Cluster-based examination + treatment-response trialHIGH

Run a minimum of three from {Hawkins-Kennedy, Neer's, painful arc, empty can} as a screening cluster. Add ER lag at 90° and drop-arm if a structural tear is suspected. Reassess at four to six weeks — the response to treatment is itself the most reliable diagnostic.

Tier 1 — Reserve imagingHIGH

Do not order routine early MRI in primary care. Reserve imaging for failed six- to twelve-week conservative trials, suspected acute massive cuff tear, or red-flag indication.

Tier 2 — Manual therapy as adjunctMODERATE

One to two sessions per week during the first four to six weeks, integrated with exercise. Strong as an add-on. Not effective as monotherapy.

Tier 2 — Scapular-stabilization protocolMODERATE

Add lower-trapezius and serratus-anterior loading two to three times per week when the Scapular Assistance or Scapular Retraction Test is positive. Yuksel 2024 RCT (N=64) showed superior outcomes on dyskinesis, pain, strength, and disability when added to standard care.

Tier 2 — Eccentric or concentric-eccentric loadingMODERATE

Twelve-week daily home program for tendinopathy-pattern presentations. Top-ranked alongside scapular protocols in Zhang 2025 NMA.

Tier 3 — BFR low-load loading bridgeEMERGING

30-15-15-15 protocol at 30 percent of one-rep max as a Phase 2 bridge when irritability or load tolerance prevents full-load cuff training. Cross-engine evidence base from BFR research; not yet shoulder-specific RCT base.

What does not work

Routine subacromial decompression / acromioplasty. Two sham-surgery RCTs (FIMPACT 2018/2021, CSAW 2018) show surgery does not outperform placebo at 24 months or 5 years.
Therapeutic ultrasound, laser, shockwave, and pulsed electromagnetic energy as monotherapy. Moderate evidence of no effect (Pieters 2020 JOSPT umbrella review).
Kinesio taping as monotherapy. Sham-controlled trials show no biomechanical effect.
Acromial-morphology-driven prescription. Type II/III hooked acromion does not predict pain or surgical benefit.
Serial cortisone injections. No durable benefit; harm signal at high frequency.

Return to training

The criteria below replace time-based progressions. Tick them off, then progress.

Full active and passive range of motion symmetric to the contralateral side within 10–15°.
No painful arc reproduction at submaximal load.
Strength symmetry within 10–15% on handheld dynamometry (external rotation, internal rotation, abduction).
SPADI or QuickDASH improvement at or above the minimum clinically important difference and at or near pre-injury baseline.
Tolerates progressive loading without 24-hour flare for two consecutive weeks.
For overhead athletes: tolerates submaximal sport-specific drills (throwing, racquet) without symptom reproduction.
Sleep no longer disturbed by shoulder symptoms.

Safety first

Stop and refer if any of these are present

Unexplained mass, night pain unrelated to position, unintended weight loss, or prior cancer history.
Fever, a hot swollen joint, or systemic illness — especially after a recent invasive procedure.
Recent trauma followed by sudden severe pain and inability to elevate the arm at all.
Sudden weakness post-trauma with a positive drop-arm and external-rotation lag — this is an acute massive cuff tear in the 4–6 week surgical window.
New neurological deficits, dermatomal pattern, or progressive weakness or numbness — screen the cervical spine.
Bilateral shoulder pain with morning stiffness over 60 minutes, systemic features, age over 50 — refer for inflammatory or rheumatology workup.
Atypical pain pattern that is not mechanical or position-dependent — consider visceral referral and book medical workup.

Refer to: GP for medical workup. Orthopedics for suspected acute massive cuff tear with ER lag and drop arm. ED for suspected septic joint or fracture. Neurology for progressive neurological deficit.

The Takeaway

Try the painful arc, right now.

Stand sideways to a mirror. Slowly raise your affected arm out to the side from your hip toward overhead. If pain reproducibly appears between roughly shoulder height (about 60°) and just above eye level (about 120°) and eases above and below, that is a painful arc — the most efficient single shoulder self-screen. It does not diagnose anything on its own. It tells you whether the rotator-cuff and subacromial region pattern is in play and worth a proper assessment.

The Verdict

The plain-English version

"Shoulder impingement" is a 50-year-old label for a real pain pattern, and surgery to fix it does not beat sham surgery.

The analogy. A shoulder is not a hinge being pinched by a bone. It is a load-managed system. The pain is closer to a thermostat that has reset itself low after weeks of overload than to a pebble being pinched in a door. Strengthening the rotator cuff and scapula resets the thermostat. Cutting bone away does not — that is what the sham-surgery trials proved.

Three things you need to know

What this actually isA pain pattern in the rotator-cuff and subacromial region with movement and load consequences, not a single mechanical pinch under the acromion.
What most people get wrongTreating one positive test (Hawkins-Kennedy, Neer's) as a diagnosis — it is a screening clue, not a verdict, and clusters of three or more shift probability much more than any single test.
The first thing to start doingGraded loaded rehabilitation for six to twelve weeks under criteria-based progression — strength of the rotator cuff and scapula is the medicine.

Best for

Adults with insidious shoulder pain in the cuff region, no clear traumatic full-thickness tear, no red-flag pattern, who can commit to a 6–12 week supervised rehab block.

Skip if

You have any red-flag pattern listed above, sudden severe weakness after a fall, bilateral systemic stiffness, or new neurological symptoms — go straight to a clinician for screening, not a self-rehab plan.

Want the full evidence and assessment cluster?▾

Conviction: Moderate

HIGH for cluster-based diagnostic strategy and graded loaded rehab as first-line. HIGH for sham-surgery null result and routine MRI not indicated. MODERATE for specific scapular protocols outperforming general loading. LOW (replaced) for "impingement" as a unitary mechanical diagnosis.

What would change this protocol: a sham-controlled RCT of acromioplasty restricted to the narrowly mechanical sub-population (Type III hooked acromion + isolated supraspinatus involvement + symptomatic AHD narrowing) showing decompression > sham at MID would re-open the mechanical model for that specific phenotype. No such trial currently exists.

What would change my mind on the diagnostic side

A primary-care or sports-medicine pragmatic RCT (N ≥ 400) randomizing adults with a positive cluster of three or more PETS to (a) standard structured cluster-based examination plus treatment-response trial vs (b) early MRI-driven targeted treatment, with a prespecified 12-month SPADI improvement primary outcome favoring the imaging-first arm by at least the MID, would shift the recommendation toward earlier imaging.

What would change my mind on surgical indication

FIMPACT and CSAW did not stratify by acromial morphology or by isolated supraspinatus pattern. A sham-controlled trial in that narrower phenotype is the door that is technically still open.

Tired of guessing what's actually wrong with your shoulder?

The Verdict cuts through the noise — evidence-based protocols every week, free.

Join The Verdict

The Full Picture — Anatomy, Diagnosis & Evidence

What's actually going on

Anatomy of the rotator cuff and subacromial region

The rotator cuff (supraspinatus, infraspinatus, subscapularis, teres minor) sits between the scapula and the humerus. The long head of biceps tendon shares the anchor seam at the upper labrum. The subacromial-subdeltoid bursa sits between the cuff and the underside of the acromion. The original 1972 Neer model proposed that the supraspinatus tendon was mechanically pinched by the anterior acromion during arm elevation. That model produced a generation of "impingement tests" and a generation of acromioplasty surgery.

The mechanism story has been revised on three fronts. Asymptomatic adults show high prevalence of cuff tears, AC hypertrophy, and acromiohumeral narrowing on imaging — the structural finding does not predict pain. Sham-surgery trials (FIMPACT, CSAW) demonstrate that decompression does not outperform placebo. And exercise produces outcomes equivalent to surgery in conservative trials and umbrella reviews.

What replaces the mechanical model is a multifactorial frame: pain in this region reflects load tolerance of the rotator cuff and surrounding tissue, scapular and kinetic-chain control, glenohumeral kinematics, central sensitization in chronic cases, and contextual factors. The pragmatic working term is "subacromial pain syndrome" (SAPS) or "rotator cuff related shoulder pain" (RCRSP).

How to identify it

Cluster-based shoulder examination — clinical scene

The honest framing first: across four meta-analyses (2012–2024), no single physical examination test for shoulder impingement reaches the Sn ≥0.80 + Sp ≥0.80 threshold needed for standalone rule-in or rule-out. Cluster them — three or more positive of {Hawkins-Kennedy, Neer's, painful arc, empty can} meaningfully shifts probability toward the SAPS/RCRSP pattern.

Test	Sensitivity	Specificity
Hawkins-Kennedy	0.69–0.78	0.57–0.62
Neer's sign	0.62–0.79	0.36–0.59
Painful arc 60–120°	~0.74	~0.81
Empty can / Jobe	~0.69	~0.62
External Rotation Lag at 90°	High DOR for tear	High
Drop-arm	Low–moderate	0.77–0.92
Lift-off (subscapularis)	0.40–0.60	>0.90

Pair the cluster with a differential screen (Spurling and distraction for cervical referral, apprehension for instability, passive ROM loss for capsulitis, cross-body adduction for AC joint, O'Brien for biceps anchor) and with the irritability stage. The goal is a working clinical pattern, not a single-test diagnosis.

The Debate

"Hawkins-Kennedy and Neer reliably diagnose impingement"

Both are screening tools (Sn 0.69–0.78, Sp 0.57–0.62). They cannot rule in or rule out alone. Use as probability shifters in a cluster.

"Acromial morphology drives shoulder pain"

Type II/III hooked acromion does not predict pain or surgical benefit. Sham surgery matches decompression at 24 months and 5 years.

"Decompression after 6 weeks of failed conservative care"

FIMPACT and CSAW show decompression does not outperform sham. Karjalainen 2019 Cochrane shows RC repair vs structured exercise: no clinically important difference at 1 year. Surgery is reserved for failed criteria-based progression with a clear structural indication.

"Routine MRI early in workup"

Imaging-pain mismatch is the dominant epistemic issue. Reserve imaging for failed 6–12 week trial or red-flag indication.

"Specific scapular-targeted exercise required for SAPS"

Specific protocols outperform general only in stratified cohorts. Default to graded loaded rehab; add scapular-specific work when SAT/SRT is positive.

Honest limitations

Inter-rater reliability is lower in primary care than in research

Diagnostic studies report kappa values for shoulder PETS in the 0.4–0.7 range. Studies are typically performed by experienced clinicians; primary-care practice will sit at the lower end. Train clusters, not individual tests; let the treatment-response trial be the ultimate diagnostic.

Reference-standard drift across decades

Older diagnostic-accuracy studies used arthroscopy-confirmed bursal contact as the reference standard. Modern evidence treats SAPS as a clinical syndrome. Pooled DORs across decades mix incompatible references.

Surgical-indication drift

Pre-FIMPACT and pre-CSAW practice indicated decompression liberally after a brief conservative trial. Practice has lagged the evidence; document the supervised loaded rehab dose explicitly before any surgical referral.

The Nuance

The conservative vs surgical decision in shoulder pain

The default escalation pathway from "shoulder pain" to "subacromial decompression" should be considered closed for the routine adult patient with SAPS/RCRSP. Two highest-quality sham-surgery trials and the largest Cochrane on cuff repair tell a consistent story: most adults do as well or better with structured loaded rehab as with surgery. Surgery is reserved for specific indications — acute traumatic full-thickness cuff tear with marked ER lag and drop-arm in a high-functional-demand patient inside the surgical window, failed 12-week supervised conservative trial with persistent functional deficit and clear structural indication, or specific anatomic lesions.

The assessment masterclass takeaway is not better tests. It is asking better questions of the tests we already have. A positive Hawkins-Kennedy is a probability shifter, not a verdict.

Sources

Alqunaee M, Galvin R, Fahey T (2012). Diagnostic accuracy of clinical tests for subacromial impingement syndrome — a systematic review and meta-analysis. Arch Phys Med Rehabil. N=1,684. Hawkins-Kennedy and Neer Sn 0.69–0.78, Sp 0.57–0.62.
Hanchard NC, Lenza M, Handoll HH, Takwoingi Y (2013). Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum. Cochrane Database Syst Rev. No single test reaches Sn ≥0.80 + Sp ≥0.80; clusters slightly outperform singles.
Gismervik SØ, Drogset JO, Granviken F, et al. (2017). Physical examination tests of the shoulder — DOR ranking. BMC Musculoskelet Disord. Supraspinatus test for full-thickness RC tear DOR 9.24.
Zhao Q, Palani P, Kassab NS, et al. (2024). Evidence-based approach to the shoulder examination for subacromial bursitis and rotator cuff tears. BMC Musculoskelet Disord. N=3,438. ER Lag Sign at 90° highest DOR for RC tear.
Paavola M, Malmivaara A, Taimela S, et al. (2018). FIMPACT trial. BMJ. N=210; ASD vs sham at 24 months below MID.
Paavola M, Kanto K, Ranstam J, et al. (2021). FIMPACT 5-year follow-up. Br J Sports Med. N=175 at 5 years; ASD vs sham below MID.
Beard DJ, Rees JL, Cook JA, et al. (2018). CSAW trial. Lancet. N=313; surgery did not exceed sham arthroscopy at MID.
Karjalainen TV, Jain NB, Heikkinen J, et al. (2019). Surgery for rotator cuff tears. Cochrane Database Syst Rev. 9 RCTs N=1,007; RC repair vs non-operative care no clinically important differences at 1 year.
Pieters L, Lewis J, Kuppens K, et al. (2020). Conservative PT for subacromial shoulder pain — umbrella review. J Orthop Sports Phys Ther. STRONG recommendation for exercise; STRONG recommendation manual therapy as adjunct.
McCreesh KM, Crotty JM, Lewis JS (2015). AHD measurement reliability — SR. Br J Sports Med. Strong evidence for ultrasound reliability; clinical utility unestablished.
Lewis J (2016). Rotator cuff related shoulder pain — assessment, management and uncertainties. Manual Therapy. Paradigm-shift narrative review proposing the RCRSP umbrella term.
Yuksel E, Yesilyaprak SS (2024). Scapular stabilization exercise training in SAPS — RCT N=64. J Bodyw Mov Ther. Scapular-stabilization-added group superior on dyskinesis, pain, strength, disability.
Zhang W, Du M, Xia L, et al. (2025). Seven types of exercise in RCRSP — Bayesian NMA. J Orthop Surg Res. Concentric-eccentric and scapular protocols rank highest.