Rotator Cuff Tear Treatment: Surgery Versus Exercise

Why Rotator Cuff Tear Treatment Matters

Shoulder pain drives roughly 4.5 million clinician visits a year in the United States, and rotator cuff disease is its most common cause. Most patients first present in primary care, where the decision is rarely whether to treat but how.

For decades, a symptomatic tear on imaging pointed many patients toward the operating room. Rotator cuff repair and subacromial surgery became among the most frequently performed orthopaedic procedures in high-income countries.

Then a series of rigorous trials unsettled that logic. Placebo-controlled surgical studies asked a question few had tested directly: does the operation outperform a convincing sham?

In the CSAW trial, arthroscopic decompression produced Oxford Shoulder Scores no better than arthroscopy alone, the placebo (mean difference -1.3 points, 95% CI -3.9 to 1.3). Both surgical arms beat no treatment only by a margin the authors judged not clinically important.

The independent Finnish FIMPACT trial reached the same conclusion, and added a third arm: structured exercise. Decompression, placebo surgery, and exercise all converged, with no between-group difference exceeding the clinically important threshold at two or five years.

These results forced a reasonable question for the tear itself. If removing bone spurs adds nothing over placebo, how much does repairing the tendon add over rehabilitation? That is the comparison this review examines.

Pain Relief — How Surgery and Exercise Differ

On pain, the two approaches sit closer than many patients expect. The Cochrane review found repair reduced 12-month pain by 0.87 points on a 0-to-10 scale (95% CI 0.43 to 1.30), an improvement that falls short of what patients typically register as meaningful.

Kukkonen’s three-arm trial saw no significant pain difference across physiotherapy, decompression, and repair. Both surgical strategies caused improvement, but not beyond what exercise alone achieved.

Lambers Heerspink offers the counterpoint. In degenerative tears, repaired patients reported significantly lower pain scores at one year than those treated conservatively, even though function did not differ.

Tendon Healing and Rotator Cuff Tear Progression — How They Differ

Here the two approaches genuinely part ways. Surgery restores the tendon; exercise does not. Kukkonen documented this directly: mean tear size stayed near 11 mm in the non-repaired arms but shrank to about 4 mm where the tendon was fixed.

Left unrepaired, degenerative tears tend to progress. The AAOS 2025 guideline notes with strong-rated evidence that tear size, muscle atrophy, and fatty infiltration may advance over five to ten years under non-operative management.

That structural drift has a clinical echo. In Moosmayer’s cohort, tears that stayed unrepaired throughout the study enlarged substantially, and the associated shoulder scores declined — a plausible mechanism for the late functional gap favouring repair.

Crossover captures the practical balance. Across trials, a consistent minority of exercise-first patients — roughly 12% in the pooled meta-analysis, and about a quarter of Moosmayer’s physiotherapy arm over the long term — eventually elected surgery. Most did not.

Quality and Consistency of the Evidence

The body of evidence is unusually strong for a surgical question — multiple Level I trials and independent meta-analyses — but it carries real limitations.

• Moderate — short-term pain and function (repair vs exercise). Consistent across trials and pooled analyses, but the Cochrane review downgraded certainty for risk of bias, since no cuff-repair trial could blind participants.
• Low — long-term superiority of repair. Driven largely by a single well-conducted cohort (Moosmayer) and heterogeneous meta-analytic pooling; the direction is consistent but the effect size is uncertain.
• High — placebo-controlled decompression findings. Two independent, blinded trials (CSAW, FIMPACT) agree that decompression adds nothing over sham for impingement — though this addresses a different operation than tendon repair.
• Moderate — structural progression without repair. Imaging data and guideline consensus agree that unrepaired degenerative tears may enlarge over years, but the link to symptoms is inconsistent.

One structural gap dominates: no trial has ever compared cuff repair against placebo surgery. Every repair-versus-exercise estimate is therefore unblinded, leaving room for expectation effects that the decompression trials showed can be substantial.

Practical Implications

The evidence supports framing the choice as genuinely shared rather than default-surgical. For a small-to-medium degenerative tear in an older adult, a structured exercise trial is a well-founded first step that most patients will not need to escalate.

The AAOS 2025 guideline reflects this balance, endorsing either physical therapy or repair for symptomatic small-to-medium full-thickness tears, since both produce meaningful improvement in patient-reported outcomes.

Counselling can be concrete. Exercise offers substantial symptom relief without operative risk, but leaves the tendon torn and open to slow enlargement. Repair adds structural healing and a modest long-term functional edge, at the cost of surgery, recovery time, and a retear rate.

Because these are educational syntheses rather than individual advice, treatment should follow clinical assessment. Drug dosages, injection decisions, and surgical timing rest on the judgement of the treating clinician.

Article Information and References