Rotator Cuff Surgery or More Physical Therapy? Questions That Clarify the Decision

The story below is a composite of real experiences; details changed.

Halfway to the top shelf, the deep ache shows up, right on schedule. Greg Palmer sets the 40-pound part back on the bench.

He’s 52 and owns the machine shop. After four months of physical therapy, he can run the mill and do the paperwork. Better, but not fixed: the hardest place to make a decision from.

His MRI shows a 2 cm full-thickness tear. His surgeon says repair is reasonable. His therapist says keep going.

His wife asks, “So which one is right?”

Maybe both. Here’s how to decide by what your shoulder has to be able to do, and which path gets it there.

Both paths help. They just help differently.

The American Academy of Orthopaedic Surgeons’ 2025 guideline gives a strong recommendation on this. For small-to-medium full-thickness tears that hurt, both physical therapy and surgery improve patient-reported outcomes. (AAOS, 2025)

So this isn’t a choice between a real treatment and a fake one. It’s a choice between two real options with different tradeoffs.

What each path looks like over time

The case for therapy. The MOON study followed 452 people with painful full-thickness tears who started with therapy.

After ten years, roughly one in four had chosen surgery. The rest kept their improvement on patient-reported scores. (MOON Shoulder Group, 2024)

The limit: success meant not having surgery. It doesn’t prove those tendons healed.

The case for repair. A Norwegian trial randomly assigned 103 people with tears up to 3 cm to repair or physiotherapy.

After 15 years, the repair group scored better on shoulder function, by about 12 points on a 100-point scale. About three in ten of the therapy group eventually had surgery anyway. (Moosmayer et al., 2024)

The limit: one hospital and repairable tears only.

The cost of waiting. The AAOS guideline warns, with a strong recommendation, that without repair, tear size and muscle quality may worsen over 5 to 10 years. That can happen even while symptoms improve.

The guideline also found that healed repairs did better than therapy and than repairs that didn’t heal. (AAOS, 2025)

Already had the repair and still hurting months later? Read what’s normal four months after rotator cuff repair.

When surgery usually comes up sooner

The AAOS patient guide lists situations where surgery may be recommended (AAOS OrthoInfo):

• Symptoms have lasted 6 to 12 months despite nonsurgical care.
• A large tear (more than 3 cm) with good surrounding tissue.
• Significant weakness and loss of function.
• The tear came from a recent, acute injury.

Greg checks two of those boxes partway. Four months, not six. Weakness, but not severe. That’s why his team is split.

Your decision worksheet

If you haven’t yet, learn how to read your MRI report line by line, and plan for time away from work in case surgery wins.

The rule on the fridge

Greg and his surgeon agree on something simple. Eight more weeks of targeted therapy. Then a strength test and a check-in.

If he can lift the 40-pound part to the top shelf without that ache, he stays the course. If not, they schedule the repair.

He writes it on a sticky note and puts it on the fridge. 40 lbs, top shelf, by Nov 15.

He didn’t pick a side. He picked a rule.

Your next step: tonight, write down the one thing your shoulder has to do. At your next visit, ask, “Which path gets me there, and how will we know?”

---

Greg is a composite, drawn from the kinds of decisions people describe with rotator cuff tears. The details are illustrative. The pattern is real.