Almost every knee replacement done in Australia is recorded in a national registry, and its 2026 report now follows more than a million of them. Most last a long time, and about 92 in 100 total knee replacements have not needed a revision (a further operation to change or add a part) 20 years later.[1] Your age, the type of replacement, the implant and a few surgical choices all change the odds, and this guide explains what that means for you.
At a Glance
- Most total knee replacements last a long time, and about 92 in 100 are still in place after 20 years.
- Your age at surgery has a big effect on your chance of ever needing a revision.
- Partial knee replacements are revised more often than total knee replacements, but they keep more of your own knee and suit some people well.
- You can ask any surgeon which implant they use and look up its results in the registry’s free reports.
On this page
What is the Australian joint registry?
How long does a total knee replacement last?
Why are knee replacements revised?
How does the knee Dr Vishal Pai uses perform in the registry?
Does resurfacing the kneecap make a difference?
Does robotic assistance change the results?
How do partial knee replacements compare?
What is the Australian joint registry?
The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) is run by the Australian Orthopaedic Association and funded by the Australian Government. It began recording hip and knee replacements in 1999, reached Victoria in July 2000 and has been national since 2002.[1]
All 327 public and private hospitals that perform joint replacements send their data, so the registry captures almost every operation in Australia. You are included automatically unless you choose to opt out, and no names are published.[1]
The registry tracks one main outcome, which is whether a joint is later revised, meaning a second operation to change, remove or add a part. It publishes a free annual report every year, along with a plain-language lay summary, on the registry website. The 2026 report covers operations up to 31 December 2025.[1]
How long does a total knee replacement last?
Of all total knee replacements for osteoarthritis (wear-and-tear arthritis), about 3 in 100 had been revised by 5 years, 5 in 100 by 10 years, 6 in 100 by 15 years and 8 in 100 by 20 years.[1] See our guide to how long a knee replacement lasts.
Age makes a big difference. People under 55 had about 16 in 100 knees revised by 20 years, compared with about 6 in 100 for people aged 65 to 74.[1] The report’s new lifetime-risk chapter estimates the chance of ever needing a revision. The most common age for a total knee replacement is 66 to 70. At that age, 86.1% of men and 88.6% of women are expected never to need a revision in their lifetime.[1] In other words, nearly 9 in 10 people having a total knee replacement at this age will never need a revision on that knee. Men have a slightly higher lifetime risk than women at every age, and the chart below shows the figures for each age group.[1] Your general health matters too. In patients recorded since 2012, people with more serious health problems (an anaesthetic fitness grade, called ASA, of 3 or 4) had a higher lifetime risk than very fit people (ASA 1), for example about 14 compared with 10 in 100 at 61 to 65.[1]
Lifetime chance of ever needing a revision after total knee replacement
Estimated by age at surgery, men and women, osteoarthritis
- Men
- Women
Show the numbers
| Age at surgery | Men | Women |
|---|---|---|
| 46–50 | 32.9% | 28.8% |
| 51–55 | 26.9% | 23.3% |
| 56–60 | 22.0% | 18.8% |
| 61–65 | 17.5% | 14.8% |
| 66–70 | 13.9% | 11.4% |
| 71–75 | 10.6% | 8.5% |
| 76–80 | 7.6% | 6.1% |
| 81–85 | 4.9% | 3.8% |
| 86+ | 2.4% | 1.8% |
Younger people live with their knee for longer, a pattern also seen overseas.[2] These figures are group averages, not a prediction for any one person.
Why are knee replacements revised?
The three most common reasons for revising a total knee were infection (29 in 100 revisions), loosening of the implant from the bone (21 in 100) and instability, where the knee feels wobbly or gives way (11 in 100).[1]
A revision is not always a complete redo. In about a third of revisions only the plastic insert between the metal parts was changed, and in about 1 in 6 a plastic button was added to the kneecap. Both main metal parts were replaced in about 1 in 4.[1] The size of the operation and the recovery vary a lot. Read more about revision knee replacement.
What a total knee revision involved
Share of all total knee revisions for osteoarthritis, by the part or parts changed
Show the numbers
| Type of revision | Share of revisions |
|---|---|
| Plastic insert only | 31.5% |
| Both main metal parts | 24.5% |
| Kneecap button only | 16.6% |
| Insert and kneecap button | 10.9% |
| Shin bone part only | 7.0% |
| Thigh bone part only | 4.5% |
| Temporary cement spacer | 4.2% |
How does the knee Dr Vishal Pai uses perform in the registry?
Dr Vishal Pai usually uses the Stryker Triathlon cruciate-retaining (CR) knee, which keeps the knee’s own posterior cruciate ligament. He places it with Mako robotic assistance and fixes it either with bone cement or, when bone quality is good, with a cementless (press-fit) Tritanium base that bone grows onto.
In the registry, 2.3 in 100 cemented Triathlon CR knees had been revised at 5 years, 3.3 at 10 years, 4.4 at 15 years and 7.0 at 20 years. With the cementless Tritanium base, 2.4 in 100 had been revised at 5 years, so far similar to the cemented version, although longer results are not yet available.[1]
Percentage of Triathlon CR knees revised in the registry
All Australian surgeons and hospitals, all diagnoses
| Fixation | 5 years | 10 years | 15 years | 20 years |
|---|---|---|---|---|
| Cemented | 2.3% | 3.3% | 4.4% | 7.0% |
| Cementless, Tritanium base | 2.4% | – | – | – |
Of the 8 knee designs with 20-year results, Triathlon CR had the lowest 20-year revision figure, and it met the registry’s highest benchmark for performance at 10 and 15 years.[1] The cruciate-retaining and condylar-stabilised insert options performed almost the same, with 4.7 in 100 revised at 15 years for each.[1]
These are averages across many surgeons, not one surgeon’s results. Several other modern knees also perform well, and the implant is only one part of a good outcome.
Does resurfacing the kneecap make a difference?
The back of the kneecap (patella) can be left as it is or resurfaced with a plastic button. Resurfaced knees had fewer revisions from about 6 months onwards. At 10 years, 3.7 compared with 5.2 in 100 were revised for the most common knee design, and 4.5 compared with 8.0 in 100 for posterior-stabilised knees.[1] The lifetime chance of revision was lower at every age.[1]
Lifetime chance of revision, with and without kneecap resurfacing
Total knee replacement for osteoarthritis, estimated by age at surgery
- Kneecap resurfaced
- Kneecap not resurfaced
Show the numbers
| Age at surgery | Resurfaced | Not resurfaced |
|---|---|---|
| 46–50 | 24.9% | 41.9% |
| 51–55 | 20.2% | 34.7% |
| 56–60 | 16.4% | 28.6% |
| 61–65 | 13.0% | 22.9% |
| 66–70 | 10.1% | 18.0% |
| 71–75 | 7.6% | 13.6% |
| 76–80 | 5.4% | 9.6% |
| 81–85 | 3.5% | 6.0% |
| 86+ | 1.7% | 2.9% |
Part of the difference is that an unresurfaced kneecap can be resurfaced later if it becomes painful, which counts as a revision. Pooled results from randomised trials also show fewer re-operations when the kneecap is resurfaced.[3] The difference was largest with the cementless Tritanium base. At 5 years, 1.8 in 100 knees with the kneecap resurfaced had been revised, compared with 4.4 in 100 without.[1] Dr Vishal Pai routinely resurfaces the kneecap unless there is a reason not to, such as a kneecap that is too thin or worn, where the risk of fracture is higher. This is his preference, and some surgeons resurface selectively.
Does robotic assistance change the results?
Robotic assistance is now used in 45% of total knee replacements and 57% of partial knee replacements in Australia.[1] Robotic systems help the surgeon plan and place the implant precisely.
For total knee replacements overall, there was no difference in revision rates between robotic and traditional instruments once patient, implant and surgeon factors were taken into account.[1] Under 65, robotic knees had fewer revisions at 5 years (2.9 compared with 3.8 in 100), but the difference was not statistically significant, meaning it could be due to chance. In people aged 65 and over, the robotic group had a slightly higher revision rate.[1]
For the Triathlon CR knee, robotic knees had fewer revisions at 5 years (1.9 compared with 2.4 in 100). One analysis found a lower rate after the first 18 months, while a second analysis, limited to one type of plastic insert and also allowing for how the base was fixed, found no significant difference.[1] The effect also varied between implants, with a benefit seen for some implant designs and not for others.[1] For partial knee replacements, robotic procedures had a lower revision rate after 3.5 years. Only two partial knee designs can be used both with and without the robot, so the effect of the robot is hard to separate from the effect of the implant.[1] Infection made up a larger share of revisions after robotic partial knee replacements (17 compared with 7 in 100 revisions), but the actual rate of revision for infection was low and similar in both groups, about 5 compared with 4 in 1,000 operations.[1]
Function one year after surgery is similar with robotic and conventional total knee replacement.[4] Robotic follow-up in the registry is still short. You can read more about robotic knee replacement.
How do partial knee replacements compare?
A partial knee replacement (unicompartmental knee replacement) replaces only the worn part of the knee. It is revised more often than a total knee replacement, with about 11 in 100 revised by 10 years and about 1 in 4 by 20 years.[1] The lifetime chance of revision is roughly double that of a total knee at every age, for example 37 compared with 16 in 100 for people aged 61 to 65.[1]
Total and partial knee replacements revised over time
Cumulative percentage revised, all primary procedures for osteoarthritis in Australia
- Total knee replacement
- Partial knee replacement
Show the numbers
| Years since surgery | Total knee | Partial knee |
|---|---|---|
| 1 | 1.0% | 1.9% |
| 3 | 2.3% | 4.3% |
| 5 | 3.1% | 6.0% |
| 7 | – | 7.9% |
| 10 | 4.6% | 11.1% |
| 15 | 6.2% | 17.9% |
| 20 | 8.0% | 26.3% |
The most common reasons were arthritis wearing out another part of the knee (39 in 100 revisions) and loosening (30 in 100), and most revisions converted to a total knee.[1] Younger people had higher revision rates, and women had a higher lifetime risk than men, 42 compared with 33 in 100 at 61 to 65. Body weight made no significant difference.[1]
The design matters. Fixed-bearing partial knee replacements, where the plastic is locked to the metal base, had lower revision rates than mobile-bearing designs, where the plastic can slide, with 8.4 compared with 13.0 in 100 at 10 years.[1]
Fixed and mobile bearing partial knee replacements revised over time
Cumulative percentage revised, partial knee replacement for osteoarthritis
- Fixed bearing
- Mobile bearing
Show the numbers
| Years since surgery | Fixed | Mobile |
|---|---|---|
| 1 | 1.4% | 2.3% |
| 3 | 3.5% | 5.2% |
| 5 | 4.8% | 7.2% |
| 10 | 8.4% | 13.0% |
| 15 | 12.6% | 20.6% |
| 20 | 19.6% | 29.1% |
Dr Vishal Pai uses a fixed-bearing partial knee, the Stryker Restoris MCK, with Mako robotic assistance. It is now the most used partial knee in Australia, and 4.2 in 100 had been revised at 5 years, alongside the best-performing designs (4.3 to 8.1 in 100 for other commonly used designs).[1]
Revision is only one measure. A partial knee keeps more of your own knee, and whether it suits you depends on where your arthritis is.
What should you ask your surgeon?
It is reasonable to ask your surgeon these questions.
- Which implant do you plan to use?
- How does that implant perform in the Australian registry?
- Will you resurface my kneecap?
- What is my own estimated chance of a revision, given my age and health?
As an example of what you can find, this table shows the five knee implants with the lowest revision figures among total knee replacement designs that have 20 years of registry results, along with the number of knees recorded for each.[1]
Lowest revision figures at 20 years, total knee replacement
Primary procedures for osteoarthritis, lowest figure first
| Implant | Knees recorded | Revised at 20 years |
|---|---|---|
| Triathlon CR | 160,652 | 7.0% |
| Genesis II CR | 26,966 | 7.4% |
| Genesis II PS | 21,165 | 7.4% |
| Nexgen LPS Flex | 31,637 | 8.0% |
| LCS CR | 8,351 | 9.3% |
The first four figures are close and their margins of error overlap, so the differences between them may be due to chance. Newer designs do not yet have 20 years of results, so they cannot appear in this table, and that does not mean they perform less well.
You can look up the registry’s free annual reports, including the lay summary, at registry.aoa.org.au/annual-reports-2026. A good surgeon will be happy to talk this through with you.
Summary
The 2026 registry report shows that most total knee replacements in Australia last a long time, with about 92 in 100 still in place after 20 years. Your age, the type of replacement, the implant and choices such as kneecap resurfacing all affect the chance of a revision, while robotic assistance has not changed revision rates for total knee replacements overall. Partial knee replacements are revised more often but keep more of your own knee. Asking which implant your surgeon uses, and checking its registry record, is a sensible step.
If you would like to discuss how this applies to your knee, you can book a consultation with Dr Vishal Pai.
References
Show all 4 referencesHide references
- Smith PN et al. 2026. Hip, Knee and Shoulder Arthroplasty: 2026 Annual Report. Australian Orthopaedic Association National Joint Replacement Registry.
- Bayliss L et al. 2017. The effect of patient age at intervention on risk of implant revision after total replacement of the hip or knee: a population-based cohort study. Lancet. Funded by the Oxford Musculoskeletal Biomedical Research Unit, National Institute for Health Research.
- Teel AJ et al. 2019. Patellar resurfacing in primary total knee arthroplasty: a meta-analysis of randomized controlled trials. Journal of Arthroplasty.
- Parsons H et al. 2026. Robotic-arm-assisted versus conventional total knee replacement (RACER-Knee): a pragmatic, multicentre, participant-masked and assessor-masked, superiority, randomised controlled trial. Lancet. Funded by the UK National Institute for Health and Care Research Health Technology Assessment Programme.
Frequently Asked Questions
Does the registry tell me how happy people are with their new knee?
No. The main report measures revisions only, and pain, function and satisfaction are reported separately in the registry’s patient-reported outcomes report.[1] A knee that has not been revised is not always a perfect knee, and a revision is not always a failure of the implant.
I’m in my 50s. Should I wait before having a knee replacement?
Younger people have a higher lifetime chance of revision, about 25 in 100 for a total knee done at 51 to 55, compared with 12 in 100 at 66 to 70.[1] That is a good reason to try non-surgical treatment first, such as exercise, weight management and physiotherapy. It is not a reason to live with severe, disabling arthritis, and the timing is a decision to make with your surgeon.
Does my weight affect how long a knee replacement lasts?
For total knee replacements, people with a body mass index (BMI) of 40 or more had a higher early revision rate than people of a healthy weight, about 4.6 compared with 3.3 in 100 at 9 years. Being moderately overweight made little difference.[1] For partial knee replacements, BMI made no significant difference to revision rates.[1]
Is a cementless knee replacement as reliable as a cemented one?
Across all brands, cementless total knee replacements had a slightly higher revision rate than cemented knees, 5.3 compared with 4.5 in 100 at 10 years.[1] Results vary by design, and the cementless Triathlon with a Tritanium base has so far performed similarly to the cemented version (2.4 compared with 2.3 in 100 at 5 years).[1] Dr Vishal Pai chooses cemented or cementless fixation according to your bone quality.
Wondering what these results mean for your knee?
Dr Vishal Pai performs partial and total knee replacement, including robotic knee replacement. A consultation is a chance to talk through which option may suit your knee and what the registry figures mean in your circumstances.
Written by Dr Vishal Pai
Vishal is an experienced knee surgeon based in Bayside, Melbourne. He consults at Sandringham Orthopaedic Specialists and operates at The Avenue Hospital and Calvary Holmesglen Private Hospital. His practice focuses on sports knee surgery, including ACL reconstruction and meniscal repair, as well as robotic knee replacement.