Conditions & surgery  ·  should I…?

Should you have arthroscopic surgery?

Arthroscopic knee or shoulder surgery, or physical therapy first? A physical therapist walks through what the research shows, through one patient’s decision.

Rick Olderman, MSPT
 ·  22 min read

Licensed orthopedic physical therapist, 30 years of clinical experience. Former clinic owner and
author of Pain Patterns, winner of the 2025 NIEA Award, and creator of a systems-based
approach to chronic pain used in his APTA-approved continuing education course.

Right now I’m working with a massage therapist who tore the meniscus in his knee. He’s very active, and when he came to me, he’d spent six months doing everything right: resting it, following my recommendations and those of another physical therapist, and reading up on his own. The pain hadn’t gone away. He was going a little crazy. He couldn’t do the things he loves, and surgery was starting to look like an attractive solution.

I’m in a similar boat myself. A few weeks ago I slipped on a wet rock while trail running and twisted my knee. Since then I’ve had pain in one spot, right on the inner joint line, which I’m assuming means a meniscus tear to some degree. I haven’t had an MRI.

At first blush, this looks simple. He has a meniscus tear. A surgeon trims off the torn piece, and the problem is solved.

Not so fast.

Arthroscopic surgery is one of the most common orthopedic operations there is. But some of the best studies we have randomly assigned patients to real surgery or a fake one, and nobody could tell the difference in how people recovered.^{1, 2}

I’m not going to tell him, or you, whether to have surgery. What I can do is lay out what the research shows, so your decision gets made with the whole picture in mind.

The short answer, based on the research

• Degenerative knee problems (wear-and-tear meniscus tears and knee arthritis): multiple studies, including some that compared real surgery with fake surgery, found no meaningful difference between surgery and not having it.¹
• Shoulder impingement pain (pain lifting the arm, without a full-thickness rotator cuff tear): the same pattern. Real surgery did no better than fake surgery, and still showed no benefit ten years later.²
• Traumatic knee tears (a specific fall, twist, or injury): surgery has a stronger case, but one major trial found that starting with physical therapy worked just as well as immediate surgery, and about 6 in 10 still hadn’t needed surgery two years later.³
• Rotator cuff repair (stitching a torn rotator cuff tendon) is a different operation from the shoulder surgery in those trials. One study found 97% of repair patients would have it again two years later, though it left out anyone who had a complication or a second surgery.⁴ A Norwegian trial that randomly assigned people to repair or physical therapy found the two did about the same at two years, but repair pulled ahead after that for small and medium tears, and it was still ahead 15 years later.⁵
• In most of the trials that compared them, conservative care performed about as well as surgery. Rotator cuff repair is the exception. So the question isn’t “surgery or nothing.” It’s: was the conservative care you tried actually addressing the right problems?

Was your conservative care actually meaningful?

Not all conservative care is created equal. Most people who end up in a surgeon’s office already tried physical therapy, and when it didn’t work, it was easy to conclude conservative care just wasn’t the answer.

I’d push back on that conclusion.

Not on the research. Not on physical therapists. But on the standard approach they’re trained to follow. It’s focused on the joint that hurts, and often doesn’t look at problems above or below it that may be setting that joint up for failure. That may be why you were disappointed with the process.

What standard physical therapy looks like

See for yourself. These are the typical goals of physical therapy for these problems, from the patient sites of the American Physical Therapy Association and the American Academy of Orthopaedic Surgeons:^{6, 7}

• Reduce pain and swelling
• Restore range of motion
• Strengthen the muscles around the joint
• Walk without favoring the leg, and return to daily activities, work, and sport

The surgeons’ Rotator Cuff and Shoulder Conditioning Program is a good standard example, and it does include exercises for the muscles that move the shoulder blade.⁷ The APTA’s guideline for meniscus and cartilage problems fits the same picture: it recommends strengthening the knee and hip muscles, and its recommended exam focuses on the knee.⁸

These are reasonable goals that are measurable, and that’s important so the PT knows improvement is happening and for insurance reimbursement. In my experience, these programs help most people. No argument there. But look at what they’re built around: the joint that hurts. What’s missing is an assessment of why that joint may be overloaded, leading to the tear or pain in the first place:

• How the hip, ankle, shoulder blade, and mid-back are positioned and moving
• How the joint works when you walk, bend, climb stairs, reach into a cabinet: how you move, not just how fast or how far
• Getting up from a chair, vacuuming, getting out of the car

That’s the gap. I understand that many of these things are also included in many PTs’ care, but that leaves it up to the therapist to decide, which, in my mind, introduces doubt that everything that should be done, was done. The guidelines say as much themselves: they aren’t meant to be a standard of care, and the final judgment rests with the clinician.^{8, 9} My sciatic patient wasn’t unhappy with bad care; she was unhappy with the standard approach, followed correctly.

There’s also a second question to consider: who actually delivered your care?

• Physical therapists now graduate with a doctorate.
• Physical therapist assistants have an associate’s degree and work under a PT’s supervision.
• In busy clinics, you might see the PT once, for the evaluation, and an assistant for nearly every visit after that.

This may come into play when so much of the plan depends on the PT’s judgment. In one large study, patients whose care was delivered mostly by assistants or aides averaged about two more visits and finished with lower function scores.¹⁰

I tried this model myself, in my own clinic. Scaling with PTAs made financial sense (I’d even considered aides too, since they’re cheaper still), but I was never satisfied with the level of care being delivered, and that might say more about my hiring skills than about their ability to do the job well. Either way, I couldn’t make the model work financially while holding the standard I wanted, and it’s one of the biggest reasons I sold my clinic.

Component thinking vs. systems thinking

I think of the standard above as an example of component thinking. Treat the joint in front of you: get the knee or shoulder to bend and straighten well, and make sure there’s decent strength. In my experience, this works very well for acute injuries, but not so well for chronic problems. And many degenerative cases are actually chronic problems.

Instead, I use systems thinking, which incorporates how each joint affects the ones above and below it.

Schools don’t seem to teach this in a meaningful way, and there’s a reason. Medical research proves or disproves an effect by isolating one variable. It’s great at “does strengthening this muscle help this joint,” and not very good at how ten things happening at once across the body add up to someone’s pain. Many of the studies cited here left out people who had other problems at the same time. My massage therapist patient is a good example. Along with his tear, he had a stress fracture in the bone just beneath the joint surface.

That’s why I read the surgery-vs-conservative-care trials cited earlier with one question in mind: what did the conservative care actually include? Where the trials describe it, it was solid, standard care:^*

• In a Norwegian trial, everyone did the same 12-week strength and balance program, two to three supervised sessions a week.¹
• In a US trial, physical therapy followed a knee-focused protocol of swelling control, hands-on work, strengthening, and balance, for about six weeks.¹
• In the trial of younger people with injury-related tears, physical therapy was a knee-focused three-phase program, with a median of 8.5 sessions.³
• In the Finnish fake-surgery trial, both groups got the same home exercise program after surgery.¹

Several of these programs included hip strengthening. None described assessing what was driving each person’s knee problem: the hip, the foot, or the way they walk.

In one of the newest trials, adding in-clinic physical therapy (hands-on work around the knee and more strengthening of the same muscles) to a home exercise program didn’t reduce pain any further.¹¹ The researchers were careful about what that means:

In my view, more of the same isn’t the same as targeting what’s driving the problem. And yes, sometimes knee pain from a meniscus tear is just that: a torn meniscus. But what I’m talking about is figuring out if there were problems that led to that meniscus being vulnerable to a tear in the first place.

So the accurate reading is that these specific programs did about as well as surgery. That’s good news. My honest belief is that a systems-thinking approach might tip those results further toward conservative care.

This isn’t a knock on physical therapists. It’s about training, which is why I built a 30-hour CEU course that teaches this systems approach to all pain in the body.

What I actually check

(and my YouTube video links if you want to check yourself)

Knee

• Gait: how you actually walk.
• Hip and glute firing pattern: people with knee arthritis tend to have weaker hips,¹² and in my experience weak or poorly-timed glutes overload the knee (see Figure 3 · read more).
• Foot and ankle mechanics: changes how force travels up before it hits the knee.
• Femoral version: thigh bone angle in the hip socket (watch the test · read more).
• Popliteus tightness: a small forgotten muscle behind the knee that’s a commonly overlooked source of ongoing knee pain (watch · read more).
• Pelvis and rib cage level: uneven often means one leg isn’t working right (watch the test · read more).
• Compensation: is the painful knee the problem, or is the other leg forcing it to overload? (watch the test · read more).
• Standing habits: equal weight on both legs? Does one knee lock more?
• Anterior femoral glide: does the thigh bone move correctly in the socket? True AFGS testing needs a practitioner, but weak glute function, shown in the video above, is usually the giveaway that it’s present.

The hip research measured how hard the hip muscles could push in a testing position. A muscle can test strong and still not turn on when you walk or climb stairs, depending on how you’re using your body. In my experience, that’s the more important question, along with how the pelvis and trunk are positioned. I go deeper into what these studies measured, and what they missed, in my hip pain and knee pain articles.

Shoulder

• Shoulder blade position and mechanics: there are rules about how the shoulder blade rests and moves (watch below · read more).
• Serratus anterior function: a key muscle moving and stabilizing the shoulder blade; when it’s not doing its job, the whole shoulder mechanism suffers.
• Sidebending pattern: the rib cage matters because the scapula rests on it (watch the test · read more).
• Trapezius function: precise, or just present?

WatchShoulder Pain: 5 Landmarks & 2 ExercisesShoulder blade landmarks and resting position

I measure where the shoulder blade sits at rest and where it ends up once the arm is overhead. Early research supports focusing on the shoulder blade, though the evidence is low certainty.⁹

When my patient came to me, I ran him through all of this (except for the shoulder blade screen). He’d been my patient years earlier, and I hadn’t seen him since. In that time he’d built my recommendations into his own training, and it showed:

• Glutes: fired nearly perfectly
• Strength: strong almost across the board, without the weakness I usually find after an injury
• Compensation: none
• Structure: neutral thigh bones, not hypermobile, so no built-in factors his training needed to work around
• Popliteus (the small muscle behind the knee I find involved in so much stubborn knee pain): no tenderness at all (read more)
• Calf and soleus: a little tighter on the injured side

That dashed my hopes for a miraculous recovery.

I was happy to see a lower body system working that well. It also made my job harder, because there was very little left for me to change. That’s part of figuring these things out: sometimes there’s no hidden driver to find, and an injury is just an injury. My recommendations came down to a knee brace for stability, and a night splint to gently lengthen his calf and soleus. I wanted to address everything I possibly could.

Over the months, his program included clamshells, standing leg lifts, and two of my own exercises, butt pumps and Neanderthal lunges. The lunges, which I’d given him years earlier for a different problem and he’d kept doing ever since, made his pain worse. He also tried stationary biking, complete rest, taping, the brace, and the splint.

Nothing exotic. Just looking under every possible rock for clues.

If you have an inkling that your conservative care could have been more complete, try the tests in the videos above before you decide about surgery.

If you’ve decided to move forward with surgery

If you’ve already decided to have surgery, this part is about doing it well, starting with your pre-op consultation.

What happened to me

It just goes to show that you can do everything right and still have pain. But knowing I did everything right beforehand gives me peace of mind: this is the best outcome available to me. That matters. I can compartmentalize it.

It also exposed something in my own thinking. When I got that diagnosis, having surgery wasn’t really a decision I weighed. It just felt like what I needed to do. If a PT who’s spent three decades studying this didn’t know the research on arthroscopy’s risks going in, most people don’t either. That’s the reason this article, and the YouTube video I made about it, exist. With my current meniscus problem, I’m applying the wisdom from my previous shoulder surgery, along with this research, and waiting it out instead of jumping into surgery.

How to improve your outcome

The difference between component thinking and systems thinking doesn’t disappear just because you’re now post-surgical. If you’ve decided to move forward, here’s my advice: get together with your physical therapist beforehand and address every weakness and tightness issue affecting your function, as thoroughly as you can. Maybe even try those tests in my videos above to measure your function more broadly.

This serves two purposes. After surgery, the muscles around the joint can temporarily “switch off.” Researchers call it arthrogenic muscle inhibition, and it makes those muscles harder to activate. If you’ve already practiced activating those muscles beforehand, in my experience that connection tends to come back faster afterward, which speeds up your recovery.

Second, the more you fix your function before surgery, the clearer a baseline you have. That makes it much easier to tell, afterward, whether your function actually improved, and how much that mattered for your pain, instead of guessing.

Where this leaves you

Back to my patient. His decision comes down to three options, and they’re the same ones many readers face:

• Surgery now, knowing it might turn out to be a repair or a trim.
• More time, continuing the conservative care he’s already doing. For injury tears, one trial of people aged 18 to 45 found that the group who started with physical therapy, including the 4 in 10 who went on to have surgery, had the same knee function at two years as the group who had surgery right away.³ Trying conservative care first didn’t leave them worse off.
• Changing what he does for a while, so the knee pain can settle down without giving up activity entirely.

That third choice is the one I’ve made for my own knee. Instead of running, I’ve taken up swimming, which I’m very bad at. It keeps my cardio going while protecting the joint, and it gives my knee time to work things out.

None of this is about anyone hiding the long-term picture. Surgeons want to help their patients, and patients want relief now. I feel the same pull. Most of us, me included, aren’t great at weighing a long-term risk against a problem that’s bothering us today, especially when it’s our own body. Part of me wants to tell my patient to just get the surgery, because it might make him happy soonest. But I know that trimming the meniscus may speed up arthritis over the years that follow, and possibly lead to a knee replacement down the road.¹ That knowledge comes with a responsibility to put the long-term picture on the table.

When my patient read a draft of this article, he asked me whether his tear was degenerative or traumatic. My guess is both. In my experience, a lot of tears in active adults happen this way: years of gradual wear, then one load too many. For him, it may have been heavy farmer’s carries on top of a lot of training. The stress fracture is the clue pointing me in this direction. That’s often overtraining, and I’m guilty of it myself. I didn’t adjust my training to my age, and my bones and soft tissue paid for it. I’ve since changed my program. My body is happier about that than my ego is.

If your situation is degenerative, the honest takeaway is that surgery probably isn’t your best first move, and a trial of meaningful conservative care should come first. If it’s a traumatic injury with a mechanical block, or repair is possible and conservative care has been tried, surgery becomes a far more reasonable option.

And sometimes, even after you’ve read all of this, rested, and given conservative care a thorough try, your short-term goals come first. Maybe you have a race coming up, or an event that means a lot to you. Then surgery may be the right choice for you at this point in your life, made knowing the potential long-term trade-off. Just know that faster relief isn’t guaranteed either: in one fake-surgery trial of knee arthritis, the real operation never relieved pain better than the fake one at any point, while another trial found a small early advantage that faded within a few months.¹

At first blush, my patient’s decision looked simple: trim the tear, problem solved. Once you see what the research shows, it isn’t. It’s now been more than nine months. He has a surgery date on the calendar, but he’s recently started feeling better, and he’s back on the fence. It’s been a roller coaster. Whichever way he goes, he’s deciding with the whole picture, short term and long term. That’s what I want for you too.