Rotator Cuff Surgery vs Physical Therapy: What the Trials Actually Show

The Evidence: What the Major Trials Found

The rotator cuff surgery-versus-PT debate has more randomized trial data than almost any orthopedic question. Three landmark studies shape the current consensus.

The MOON group (2013) followed patients with atraumatic full-thickness tears. At two years, surgical patients had statistically better shoulder scores, but the gap was modest for tears under 1 cm. The Finnish trial (Kukkonen et al., 2015) randomized 167 patients with non-traumatic supraspinatus tears and found no meaningful difference at five years between repair, acromioplasty alone, and PT alone. And the UK CSAW trial (2018) showed that subacromial decompression added nothing over PT for impingement-range symptoms.

The takeaway is not that surgery is useless. It is that the benefit of surgery scales with tear severity and patient demands.

Who Benefits Most From Surgery

Full-thickness tears in patients under 60 who are physically active show the strongest surgical benefit. Traumatic tears from an injury respond better than degenerative tears that developed over years. Tears involving two or more tendons (massive tears) are more likely to progress without repair. And any tear with early fatty infiltration on MRI is on a clock: the longer you wait, the less repairable the muscle becomes.

Who Does Well With PT Alone

Partial tears of any size in patients over 65 frequently respond well to a structured 12-week PT program. Patients with low physical demands, those who primarily need to reach countertop height, often function well without surgery. Even some full-thickness tears in sedentary older adults show no functional difference between surgical and PT cohorts at two years.

The Crossover Question

Many trials allow patients who fail PT to cross over to surgery. Crossover rates range from 15% to 25% in most studies, meaning three out of four patients who try PT first never need the operation. For patients uncertain about surgery, a 6-to-12-week PT trial is a reasonable diagnostic step, as long as imaging does not show progressive tear enlargement or advancing fatty infiltration.

Cost and Recovery Comparison

In the US, arthroscopic rotator cuff repair ranges from $6,000 to $15,000 out of pocket depending on facility, with a 4-to-6-month total recovery arc. A 12-week PT program typically runs $1,500 to $3,000 for 2 to 3 sessions per week. If PT succeeds, the cost and time savings are significant. If PT fails and surgery follows, the patient has spent money on PT and still faces the surgical recovery window.

This is why the imaging and clinical picture matter upfront: a large tear with early muscle wasting is unlikely to respond to PT, and delaying surgery may worsen the outcome. A small partial tear in a retiree has low surgical urgency and high PT success probability.

The Bottom Line

There is no universal answer. The right choice depends on tear morphology, muscle quality, age, activity demands, and the patient's willingness to commit to rehabilitation either way. Surgery is not always better, and PT is not always enough. The imaging tells you more than the pain level does.

What the Imaging Tells Your Surgeon

An MRI is not a decision on its own, but it provides the data that makes the decision possible. Three findings matter most. First, tear type: partial-thickness tears involve only part of the tendon, while full-thickness tears extend all the way through. Full-thickness tears are further classified by retraction, how far the torn edge has pulled away from the bone. Second, muscle quality: the Goutallier classification grades fatty infiltration of the rotator cuff muscles on a 0-to-4 scale. Grade 0 is normal muscle. Grade 2 shows equal fat and muscle. Grade 3 and above means the muscle has been replaced by fat to a degree that surgical repair may not restore function, even if the tendon is reattached. Third, the acromion shape: a hooked acromion (Type III) mechanically impinges on the rotator cuff during overhead motion and may contribute to tear progression.

A surgeon who recommends PT for a partial tear with good muscle quality is not being dismissive. A surgeon who recommends repair for a full-thickness tear with early fatty infiltration in a 55-year-old is not being aggressive. Both are reading the imaging correctly for the patient in front of them.

The Rehabilitation Reality, Either Way

Whether a patient chooses surgery or PT, rehabilitation is the actual treatment. Surgery reattaches the tendon; rehabilitation restores function. Without dedicated rehab, a perfect surgical repair can still result in a stiff, weak shoulder. The PT protocol after rotator cuff repair follows a strict timeline: immobilization in a sling for 4 to 6 weeks, passive range of motion starting at 6 weeks, active motion at 8 to 10 weeks, strengthening at 12 weeks, and return to sport or overhead work at 16 to 24 weeks. Compliance with this protocol is one of the strongest predictors of surgical success. Patients who cannot commit to 4 to 6 months of structured rehabilitation should weigh that heavily in their decision.

For the PT-only pathway, a structured 12-week program focusing on scapular stabilization, posterior capsule stretching, and progressive rotator cuff strengthening produces the best results. Generic "shoulder exercises" from an internet search are not equivalent. A therapist experienced with rotator cuff pathology matters as much as a skilled surgeon does in the surgical pathway.

Age as a Factor, Not a Rule

Age influences the decision but does not dictate it. A 70-year-old recreational golfer with a chronic full-thickness tear and good muscle quality may do perfectly well with PT and never need surgery. A 45-year-old construction worker with the same tear will likely fail PT because their daily demands exceed what a compensated shoulder can deliver. The question is not "how old is the patient" but "what does the patient need the shoulder to do." A patient whose life requires overhead strength and endurance has different surgical indications than one whose life does not.