Knee Injections: What Works, What Doesn’t

Steroid injection is the most commonly offered treatment for knee arthritis in this country, and it is the one I have moved away from. I should be clear that this is a considered position rather than the standard one, so here is the reasoning.

What it does

Cortisone is a powerful anti-inflammatory. Injected into a swollen, inflamed knee it can settle the pain quickly — often within a few days. That part is real and I am not disputing it.

Why that is not enough

It treats the inflammation and leaves the arthritis exactly as it was. The cartilage is unchanged, the alignment is unchanged, the joint is unchanged. What you get is a quieter knee for a period that typically runs a few weeks to a few months — and that period gets shorter each time the injection is repeated.

Three things concern me about using it as a routine treatment:

• Repetition is not harmless. Repeated steroid injections into the same joint have been associated with cartilage loss. Using a treatment that may accelerate the underlying problem in order to mask its symptoms is a poor trade in a knee you want to keep for years.
• Every injection carries an infection risk. Small, but a septic knee is a serious problem, and the risk is being taken for temporary relief.
• It postpones the real conversation. This is the one that matters most in practice. A patient gets an injection, feels better for two months, comes back, gets another, and two years pass in which nothing has been decided and the knee has quietly got worse. Time is not a neutral thing to spend here — a stiffer, more deformed knee makes for a harder operation later.

There is also a practical consequence people are rarely warned about: a steroid injection shortly before joint replacement raises the risk of infection after surgery, so it forces a waiting period — commonly around three months — before you can be operated on.

Where steroid injection genuinely is the right treatment

None of this means steroid injections are bad medicine. In the right condition they are excellent, and I use them. The difference is that in these conditions inflammation is the disease, rather than a symptom of a worn-out joint:

• Frozen shoulder — where injection combined with physiotherapy can shorten a long and miserable illness considerably
• Tennis elbow and golfer’s elbow
• Trigger finger — often resolves the problem outright
• De Quervain’s tenosynovitis
• Carpal tunnel syndrome, in selected patients
• Plantar fasciitis that has not responded to other measures

So if you have been offered a steroid injection, the question worth asking is not whether steroid works — it does — but whether it is being used to treat a condition it can resolve, or to quieten a joint that needs a decision.

Marketed as lubricating or cushioning the joint. The physiological story is appealing; the clinical evidence is genuinely mixed, and major international guidelines disagree with each other on whether to recommend it at all.

My reading of it, in practice: a subgroup of patients with mild to moderate arthritis get worthwhile relief for several months. Others notice nothing. In advanced bone-on-bone arthritis, it rarely does much — there is no longer a surface for it to protect.

It is reasonable to try in the right knee. It is not reasonable to be sold a course of them as an alternative to a replacement in a severely arthritic knee, which happens more than it should. If you try it and feel nothing after the course, do not repeat it.

Your own blood is spun down to concentrate the platelets, and that concentrate is injected into the knee. The theory is that the growth factors in platelets damp inflammation and support tissue repair.

Where I think it stands

PRP is the most interesting of the four, and also the most inconsistently studied. Several trials show it outperforming both hyaluronic acid and placebo in early to moderate arthritis, with relief lasting six to twelve months in responders. Other trials show no meaningful benefit at all.

Part of the inconsistency is that “PRP” is not one thing. Preparation methods, platelet concentration, whether white cells are included, the number of doses — all vary between clinics, and the product you receive at one centre may bear little resemblance to the one studied in a trial.

A fair summary

Worth considering if your arthritis is early, you want to defer surgery, and you can afford it without strain. Not worth it in advanced bone-on-bone arthritis — it will not regrow cartilage that is gone, whatever is implied. And if two doses have produced nothing, a third is unlikely to.

Stem cell injection for knee arthritis is heavily marketed in India, frequently at costs approaching or exceeding a knee replacement. Patients are shown the prospect of regrowing cartilage and avoiding surgery altogether.

That is not something the current evidence supports. There is no good evidence that injected stem cells regenerate meaningful cartilage in an arthritic human knee. Studies to date are small, inconsistently designed, rarely blinded, and often conducted by the people selling the treatment. Some patients report improvement — and part of that is likely the same inflammatory damping that a far cheaper injection provides.

Regenerative medicine may well deliver something real for arthritis in the years ahead, and I hope it does. It has not delivered it yet. Paying two lakh rupees today for a treatment that cannot show it works, on the strength of what it might one day become, is not a good trade.

Roughly, and adjusted for the individual knee:

• Early to moderate arthritis, patient wants to defer surgery — a genuine physiotherapy programme first, then PRP or hyaluronic acid if that plateaus. Not steroid.
• A knee that keeps flaring — the repeated flare is information, not something to be silenced. It usually means the load on the joint needs addressing, or that the arthritis has moved on and the conversation should move with it.
• Advanced bone-on-bone arthritis, function badly affected — injections are a bridge, not a treatment. If someone is offering them as an alternative to replacement here, be sceptical.
• Every stage — weight, quadriceps strength and activity modification do more over a year than any injection, and cost nothing. They are unglamorous, which is why they are undersold.

An injection is worth having when it buys you something you actually want — a function, an event, a working physiotherapy programme, a few more years before surgery. It is not worth having as a way of not making a decision.