Cortisone Injections — What They Do and What They Cost You

A useful tool with a real bill

Corticosteroid injections are among the most commonly performed procedures in musculoskeletal medicine, and they are genuinely useful. They are also handed out more casually than they should be, and the costs are usually explained less carefully than the benefits.

My position is simple: a cortisone injection is worth it when it buys you something you can use. A window of relief that lets you sleep, or start rehabilitation, or get through a specific event, is a good trade. Relief that simply postpones addressing the actual problem for another three months is a bad one, and the trade gets worse each time you repeat it.

What it actually does

Cortisone is not a painkiller and it is not a lubricant. It is a potent anti-inflammatory that suppresses the local inflammatory cascade in the tissue where it is placed.

That means it works well where inflammation is genuinely driving the pain, and works poorly where it is not. This is the crux of a lot of disappointed expectations. A degenerative tendon with disorganized collagen and minimal inflammation is not a good target. An acutely inflamed bursa or an inflamed joint lining is.

Relief typically begins within a few days and peaks over one to two weeks. Many people get a couple of days of increased pain first, and a small number get a steroid flare from crystal formation, which settles on its own.

The evidence, condition by condition

Knee and hip osteoarthritis. Reasonable short-term benefit for pain, generally over weeks to a few months. Benefit past three months is unreliable. There is also evidence, from repeated injections given every three months over two years, of greater cartilage volume loss compared to saline — which is a genuine reason to be thoughtful about frequency.

Adhesive capsulitis (frozen shoulder). One of the better indications. Injection combined with a stretching program meaningfully accelerates recovery in the painful, inflammatory phase.

Subacromial pain and rotator cuff related shoulder pain. Short-term benefit, diminishing over time, and no clear advantage over other approaches at a year.

Carpal tunnel syndrome. Effective for temporary relief and useful diagnostically, though not a durable solution for moderate or severe compression.

Trigger finger and de Quervain's tenosynovitis. Genuinely good indications with high success rates.

Tendinopathy. This is where I am most cautious. For lateral epicondylitis, randomized evidence found that patients receiving corticosteroid did better at six weeks and worse at one year than those who did not. Short-term relief, long-term cost. Cortisone into or around a degenerative tendon can weaken the tissue, and near a weight-bearing tendon like the Achilles or patellar tendon there is a rupture concern serious enough that most of us will not inject there at all.

Plantar fasciitis. Effective short-term, with a small but real risk of fascial rupture and fat pad atrophy.

The costs worth knowing about

Tendon weakening and rupture near the injection site, which is why location matters so much.

Cartilage effects with repeated intra-articular use.

Skin and fat atrophy, sometimes with a permanent dimple or pale patch at the injection site. More noticeable in superficial injections and in darker skin.

Blood glucose elevation for days to a couple of weeks. If you have diabetes, this matters, and you should be monitoring more closely afterward.

Infection, rare but serious in a joint.

Systemic absorption, which can transiently suppress the adrenal axis and, in some people, disrupt the menstrual cycle or cause facial flushing.

Delay before joint replacement surgery. Injection within roughly three months of arthroplasty is associated with higher infection risk, so timing matters if surgery is on the horizon.

Where the "only three per year" rule comes from

It is a convention rather than a hard biological threshold, but it exists for a reason and I broadly follow it — no more than three or four injections into a given joint per year, spaced at least three months apart.

The more important question is not the count. It is what happened after the last one. If an injection bought six months of function and let you build strength, repeating it is reasonable. If the relief lasted three weeks and nothing about the underlying problem changed, a fourth injection is not treatment. It is avoidance, and it is time to reconsider the diagnosis.

Why I use ultrasound for these

Blind injection based on landmarks misses the intended target more often than most patients would guess, with accuracy varying considerably by site — and a steroid placed in the wrong tissue plane is at best wasted and at worst harmful.

Ultrasound guidance lets me see the needle enter the actual structure, avoid vessels and nerves, and confirm the medication is spreading where I intend. It also lets me see what I am treating, which sometimes changes the plan mid-procedure.

The conversation I would rather have

Before injecting anything, I want to know what you plan to do with the relief.

If the answer is that you will finally be able to start loading the tissue, sleep enough to function, or complete a course of rehabilitation, that is an excellent reason. If the answer is that it will let you keep doing the thing that is causing the problem without addressing it, we should talk about that instead — and the alternatives, including shockwave, PRP, manual treatment, and a properly structured strengthening program, are real options rather than consolation prizes.