Primary total knee replacement in many studies offers greater than 90% survival of the implants at 25 years (De Puy PFC). 

Roughly 55,000 primary knee operations are performed in the UK each year (National Joint Registry) with numbers rising year on year. Over 3,000 knee replacements need revision each year. As the number of knee replacements increases, inevitably the number of revision operations will also increase.

Revision knee replacement usually involves removing all the old components, cleaning out any debris and scar tissue and then rebuilding the joint using metal stems and possibly metal augments to support the new knee replacement. Rarely only a portion of the original knee replacement will require revision with either replacement of the Femoral (thigh bone) or tibial (leg bone) component as well as the polyethylene.

Revision knee replacement utilises significantly greater hospital resources than primary replacement:

• Operative time is longer (typically 2 to 3 hours),
• There is a longer length of hospital stay (typically 3-5 days longer than primary replacement),
• There is greater blood loss and
• Post operatively there is diminished function compared to a primary replacement.

The number of revision knee replacements continues to increase each year. It is anticipated that there will be a massive expected demand for revision knee surgery in the next two decades.

The cumulative survival rate following revision total knee arthroplasty is 95% at two years, 89% at five years and 79% at ten years. (Pu-Yi Sheng et al).  The most significant predictors of prosthetic survival was age of the patient (greater than 70 years) and the life of service of the primary total knee replacement (greater than five years or more between primary surgery and revision).

The average age at revision is 70 with no difference between men and women

Causes and Treatment of Failure

There are three main reasons why initial knee replacement surgery can fail. These are infection, stiffness and loosening. These causes and their treatments are covered in the following sections


Infection in a prosthetic joint is a major cause of early failure and significant suffering for patients. Infection can be broadly divided into two categories.

• Firstly, and counting for the vast majority, is infection at the time of the original surgery.

 • Secondly and much less common is late infection caused by organisms in the blood stream (bacteraemia) settling on a joint replacement.

Introduction of organisms at the time of surgery can never be totally excluded. There are, however, many precautions which can lower the risk significantly. These include:

 • experienced specialist surgeon,

• operations performed within a ‘clean air enclosure’,

• dedicated theatre and ward staff, • dedicated orthopaedic ward

surgical time less than one hour

 • minimal disturbance of the post-operative dressings

• antibiotic cover at the time of surgery

Late infection resulting from blood borne infection, again, can never be totally excluded. The risk can be minimised by antibiotic cover at the time of other significant surgical interventions e.g. abdominal or bladder surgery. Antibiotics are also indicated if there is a proven infection either in the same limb as the knee replacement or elsewhere.

Infected knee replacements present with pain, swelling, loss of movement, localised redness and warmth and occasionally a discharging sinus (a link between the joint and the skin that drains pus.)

In the first few weeks after a primary knee replacement there is often swelling and warmth as a result of the original surgery and this should not be confused with infection.

Infection should always be excluded as a cause of failed joint replacement


An infected total knee replacement cannot be resolved by antibiotics alone. Rarely in an acute early infection the joint can be salvaged by surgical debridement, lavage and exchange of the polyethylene liner by further surgery.

Usually an infected knee replacement will require revision in two operations to eradicate the infection.

 In the first operation the knee replacement and all infected tissue are removed.  A temporary joint replacement (or spacer) is constructed to allow the patient home, mobilising freely whilst the wounds and soft tissues settle. During this stage tissues collected at the time of the first stage are sent to a laboratory and an attempt is made to identify the infecting organism. Once an organism is identified the patient can then be treated with an appropriate antibiotic to eradicate the organism. Antibiotics will need to be administered intravenously through a special intravenous line.

The second stage is then planned to permanently rebuild the joint. The timing of the second operation will depend on the type of temporary joint used and the confidence that the infection has been cleared. Typically a preformed spacer is used. These spacers are infiltrated with antibiotics and are designed to allow movement of the knee.

Patients can usually weight bear on their joint immediately after surgery and discharge from hospital is typically 1 week after the first stage. The second operation can then be timed to maximise the clearance of infection. This will vary from patient to patient but is typically 8 to 12 weeks after the first stage.


Stiffness results from two causes.

• Firstly, Arthrofibrosis (excessive scar formation within the knee joint). This can be a natural phenomenon but is exacerbated by poor patient compliance with post-operative physiotherapy.

 • Secondly a poorly balanced replacement i.e. the implants are poorly sized and/or positioned with inappropriate tension of the knee ligaments and capsule.

Stiffness is often, but not always, accompanied by pain and significant loss of function (range of movement)


Stiffness of a knee resulting from arthrofibrosis may respond to physiotherapy, hard work and time. Failure to improve over 6 to 12 months should be considered for surgical excision of scar tissue with significant improvement expected.

Stiffness resulting form malpositioning of the knee replacement will not improve with physiotherapy over any length of time. Malpositioning can only be improved by removing the replacement and revising the knee to establish soft tissue balance thus allowing the knee to move.


With time all implants will loosen i.e. break the bond between the cement or implant and bone. This can occur late (10 to 15 years) merely as the result of time and activity as polyethylene wear debris leads to loosening. Unfortunately some joints loosen early (2 to 5 years).

Poorly positioned or balanced implants will result in excessive wear of the polyethylene liner and abnormal loading of the fixation to bone. The result will be loosening of the implant which may present with a variety of symptoms. These include: Pain from the loss of fixation to bone; Instability due to loss of polyethylene thickness and/or subsidence of the implants; Swelling caused by polyethylene and/or metal debris within the joint.


A loose knee replacement can only be resolved by removal of the implants and revising the knee. Revision knee replacements typically need stems as well as other augments to increase the surface area over which fixation to bone can be achieved.

Significant loosening may have resulted in loss of bone around the old knee components. In such cases augments will be required to restore the joint line and stabilise the soft tissues and ligaments. The cause of joint loosening should be established before surgery. Infection in particular should be excluded.


Revision knee replacement is challenging both for the surgeon and the patient. Successful surgery will restore the patient to mobility and remove their pain. There is often a loss of function compared to the primary knee replacement and expectations and end results must be clear.

 Whilst technically challenging the benefits to patients are clear and revision knee replacement is often a highly rewarding procedure.