Surgery is the last item on this page deliberately. For most people with hip or knee arthritis, the treatments below (used properly and given time) control symptoms well. They also make surgery safer and recovery faster if an operation is needed later. Nothing here is wasted effort.
Most people asking about alternatives to knee replacement, or to hip replacement, are asking two things at once: what actually works, and how long it might postpone surgery. The measures below are ordered by how well they are supported, not by how often they are sold.
Exercise and physiotherapy, the strongest evidence here
Exercise is the single best-supported treatment for arthritis of the hip and knee. That surprises many people, a worn joint feels like something to protect, not to work. But strengthening the muscles around a joint reduces the load going through it, and moving the joint keeps it supple. Australian and international guidelines put exercise first for good reason.
The size of the effect is worth knowing, because it puts everything else on this page in context. A 2024 Cochrane review pooled 139 trials and 12,468 people and found exercise improved pain by about 13 points on a 100-point scale compared with no treatment, and physical function by a similar amount. The reviewers rated the certainty of that evidence as low to moderate, and the measurements were short-term.
For comparison, hyaluronic acid injections (widely sold for arthritic knees) shift pain by about 2 mm on a 100 mm scale. On comparable scales, exercise is worth roughly six times as much as the injection most heavily marketed for this problem. That is not an argument for enthusiasm; it is an argument for spending your effort where the evidence is.
A physiotherapist can build you a program suited to your joint and your fitness, and progress it as you get stronger. It takes weeks to feel the benefit, not days. Expect some muscle soreness at the start; that is different from harm.
Weight management
Each kilogram of body weight puts several kilograms of force through the knee with every step. Losing even a modest amount of weight measurably reduces pain in weight-bearing joints. If you are carrying extra weight, this is likely to be the most powerful thing you can do for your joint, and if surgery is needed later, it also lowers the risk of complications from both the anaesthetic and the operation.
This is easy to say and hard to do. Your GP can help, and there are structured programs that work. It is raised not as a judgement, but because the evidence is strong.
Medicines
- Paracetamol helps some people with milder pain and is safe for most when taken as directed.
- Anti-inflammatory tablets (NSAIDs) such as ibuprofen are often more effective for arthritis pain. They are not suitable for everyone (they can affect the stomach, kidneys and blood pressure) so check with your GP or pharmacist, especially if you take other medicines.
- Anti-inflammatory creams and gels are worth trying for the knee, where the joint is close to the skin.
- Stronger pain medicines (opioids) have a poor record in arthritis: modest benefit, significant harms, and real dependence risk. They are not a long-term treatment for joint pain.
Walking aids and simple measures
A walking stick used in the opposite hand takes meaningful load off a painful hip or knee. Firm, supportive shoes help. So does pacing, planning the day so the joint is not asked to do everything at once. None of this is giving in; it is managing a mechanical problem mechanically.
Injections
Several injections are offered for knee arthritis, and the evidence behind them is uneven. Because some are marketed directly to patients and paid for out of pocket, it is worth setting out plainly what the better trials show. None of these treatments slows or reverses arthritis; the question in each case is whether it relieves symptoms better than a placebo injection.
Corticosteroid (“cortisone”)
A corticosteroid injection can settle a painful, inflamed knee, and of the injections on this page it has the best support for short-term symptom control. A Cochrane review of the trials found the benefit was moderate at one to two weeks, smaller by four to six weeks, small at three months, and gone by six. The reviewers rated the certainty of that evidence as low, and were careful to say it is unclear how much of the early benefit is large enough to matter to a patient.
So it is a reasonable option for calming a flare or getting through a particular period, with realistic expectations: useful early, fading over weeks to a few months, and not something that changes the arthritis itself.
What the evidence argues against is injecting on a routine schedule. In a randomised trial, patients received either corticosteroid or salt water into the knee every twelve weeks for two years. The corticosteroid group lost slightly more cartilage (0.21 mm against 0.10 mm of thinning) and their pain was no better than the salt-water group’s. The sensible conclusion is that a single injection with a clear purpose is defensible, and an indefinite programme of them is not.
Hyaluronic acid (“gel” injections, including Synvisc)
This is the largest evidence base of any of them: 169 trials and more than 21,000 patients, pooled together. Hyaluronic acid does reduce pain compared with a placebo injection, but by roughly 2 mm on a 100 mm pain scale, when the difference a person actually notices is about five times larger. The same analysis found serious side effects were somewhat more common than with placebo.
The researchers concluded that the evidence is now settled enough that further trials of the same kind are unlikely to change the answer, and that the findings do not support routine use.
Platelet-rich plasma (PRP)
PRP concentrates factors from your own blood and injects them into the joint. An Australian randomised trial gave 288 people with knee arthritis either PRP or salt water and followed them for a year. Pain improved by 2.1 points with PRP and 1.8 points with salt water, a difference small enough to be chance. Cartilage loss was the same in both groups. Of 31 other measurements, 29 showed no difference.
The important detail is that both groups improved. People who receive PRP often do feel better, which is why it seems to work and why studies without a placebo group are so unreliable. Compared against a placebo injection, no advantage was found.
Cooled radiofrequency ablation of the genicular nerves
This one is different in kind. Rather than treating the joint, it interrupts the nerves that carry pain signals from the knee. The arthritis is unchanged; the pain messaging is turned down, and the effect wears off as the nerves recover.
It is also the only treatment on this page where a placebo-controlled trial found a genuine benefit: in a study of 40 patients, 76.5% of those treated responded at three months against 33.3% of those given a sham procedure, with relief maintained at six months. A separate follow-up group still reported meaningful relief at two years. That said, this is one small trial of 40 patients. The evidence for radiofrequency ablation is far thinner than for the treatments above, and its place will not be settled until larger independent trials report.
The honest caveat is size. Forty patients is a small trial, and the two-year data describe 27 people. It is best understood as promising and genuinely supported, rather than established, and a large trial comparing it directly against injections is currently under way. It is generally considered where simpler measures have not worked and joint replacement is either not appropriate or not yet the right timing.
If a joint replacement may be coming, timing matters
There is one practical point that applies to all of the injections above, and it is worth raising before you have one.
Studies looking back at large numbers of patients have found that having an injection into the knee within about three months of a joint replacement is associated with a higher rate of infection around the new joint. For corticosteroid the increase is modest; for hyaluronic acid given within a month of surgery it appears larger. One study of knee replacements found no such association, and all of this research is observational; it can show a pattern but cannot prove the injection caused the infection. The absolute risk of infection remains low either way.
Infection in a replaced joint is nonetheless a serious problem, so the sensible response is simply about sequencing. If a knee is likely to need replacing in the foreseeable future, it is reasonable to avoid injecting it in the few months beforehand, and if you have had an injection, tell your surgeon when. It is a question worth asking before an injection rather than after.
Where that leaves things
Exercise, weight management where it applies, and appropriate pain relief remain the measures with the strongest support, and none of the injections above replaces them. If you are considering any of these treatments, two questions are worth asking: what is this injection meant to achieve, and what happens after it.
Arthroscopy, a note
Keyhole surgery to “clean out” an arthritic knee was common for decades. Good trials have since shown it does not help osteoarthritis, and it is one of the procedures Choosing Wisely Australia (the profession’s own campaign against low-value care) specifically advises against for arthritis. If someone offers to arthroscope your arthritic knee, it is reasonable to ask why.
When is it reasonable to consider a replacement?
There is no X-ray finding, score or age that makes the decision for you. A joint replacement becomes worth discussing when three things line up:
- The arthritis is advanced on examination and X-ray.
- The non-surgical options above have been genuinely tried and are no longer holding the line.
- The joint is taking things from your life you are not willing to give up, sleep, work, independence, the walk you have always done.
If that describes you, the next chapters explain what preparing for surgery involves, and what the operation actually is. Reading them commits you to nothing.