Abstract
Total knee arthroplasty (TKA) is a cost effective and extremely successful operation. As longevity increases, the demand for primary TKA will continue to rise. The success and survivorship of TKAs are dependent on the demographics of the patient, surgical technique and implant-related factors.
Currently the risk of failure of a TKA requiring revision surgery ten years post-operatively is 5%.
The most common indications for revision include aseptic loosening (29.8%), infection (14.8%), and pain (9.5%). Revision surgery poses considerable clinical burdens on patients and financial burdens on healthcare systems.
We present a current concepts review on the epidemiology of failed TKAs using data from worldwide National Joint Registries.
Cite this article: Bone Joint J 2016;98-B(1 Suppl A):105–12.
Total knee arthroplasty (TKA) is a cost effective and successful operation for patients with end-stage arthritis of the knee.1,2 In the United States, the demand for TKA is projected to rise by 673% to 3.48 million procedures by 2030, and revision TKA by 601%.3 Furthermore, the demand for TKA among patients aged < 65 years is predicted to be > 50% of the TKAs which are performed by 2016.3 An ageing population, a rising prevalence of risk factors such as obesity, and the necessity to maintain active lifestyles will continue to feed this trend in developed countries.3,4 A consequential rise in failed TKAs is inevitable, and they present a significant financial and health-related burden to patients and healthcare systems.5
We present a current concepts review on the epidemiology of failed TKAs using data from worldwide National Joint Registries (NJRs).
Defining knee arthroplasty failure
Revision surgery remains the benchmark outcome by which survival of orthopaedic implants is assessed.6 However, using revision surgery as the sole endpoint underestimates the volume of patients with pain or poor function following the primary procedure, and who do not qualify for revision surgery. Price et al6 highlighted this problem in a long-term review of 60 TKAs undertaken in 53 patients with primary osteoarthritis, aged < 60 years at the time of surgery, between 1987 and 1993. They reported a minimal 12-year cumulative survival of 82.2% using revision as an end point. This decreased dramatically to 59% if moderate or severe pain was included as an end point.
The rate of poor outcome following primary TKA ranges from 7% to 20% due to pain, stiffness, and poor function despite the absence of a discernible objective reason for revision.7-14 The use of different outcome measures substantially alters the proportion of TKAs which are deemed successful: 90-day mortality after TKA is 0.4%,15 3.8% of patients will undergo a non-revision re-operation within four years,16 5.6% of patients have worse Patient Reported Outcome Measures (PROMs) six months post-operatively,17 or a Knee Society Score (KSS) of < 70.18
PROMs are becoming popular in the assessment of satisfaction and health-gain post-operatively. Data from the Health and Social Care Information Centre (HSCIC) England show high post-operative PROM scores in the EQ-5D Index (82.1%), EQ-VAS (57.2%), and Oxford Knee Score (94.4%) for 2014 to 2015 following TKA, confirming the success of the operation.19
NJRs provide a wealth of community-based comparative data including patient characteristics, implant factors, and surgical techniques. They highlight trends in the variation of outcomes whilst remaining sensitive to the impact of changing practice and allowing the identification of best practice.20 There is mounting recognition that registries are an excellent adjunct to randomised controlled trials in determining the safety and effectiveness of different interventions in a timely and cost effective manner.21-23 The Swedish Knee replacement project was established in 1975 and was the first NJR of its kind.24,25 With surgical advancement, the focus of the registry shifted from evaluating short-term complications of TKA to time to first revision. There are currently 11 NJRs worldwide which all use time to first revision as the primary endpoint.20 Presently, no other benchmark exists for the assessment of clinical outcome. Furthermore, monitoring revision rates permits the identification of factors influencing the outcome.
Indications for revision
More than 76 000 TKAs are performed annually in the United Kingdom and risk of revision following primary TKA for osteoarthritis (OA) ten years post-operatively is < 5%.26 This figure is consistent across a number of registries including 4% in Sweden, 5% in New Zealand, and 6.8% in Australia.26-30 Pooled data from registries worldwide identifies that the most common indication for revision surgery is aseptic loosening (29.8%), followed by infection (14.8%) and pain (9.5%) (Table I).31-34 Infection and pain as indications for revision have increased compared with aseptic loosening which has fallen since the NJRs first began (Fig. 1).27 The American Joint Replacement Registry (AJRR) highlights the fact that 68% of patients who underwent revision TKA between 2012 and 2013 did so within three months of the initial procedure.30 Patellofemoral pain, instability, and stiffness comprise the other main indications for revision surgery.
Fig. 1
Graph showing the cumulative incidence of the indication for revision of primary total knee arthroplasty. (Reproduced from Australian Orthopaedic Association NJR: Annual Report 2014).
Table I
Variation trends for the indication of revision TKA for the most commonly referred to National Joint Registries (NJR)
| NJR England, Wales, and Northern Ireland | Swedish knee arthroplasty register | Australian orthopaedic association NJR | New Zealand NJR | American NJR | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1st annual report 200432 (%) | 11th annual report 201426(%) | 1st annual report 199933(%) | 13th annual report 201429 (%) | 1st annual report 200034(%) | 14th annual report 2014 (%) | 1st annual report 199928(%) | 15th annual report 201328(%) | 1st annual report 201330 (< 3 months to revision) | |||||
| Aseptic loosening | 41.4 | 32.0 | 43.6 | 26.0 | 40.3 | 32.7 | 0 | 34.6 | N/a | ||||
| Deep infection | 18.4 | 22.0 | 11.2 | 22.0 | 9.1 | 17.4 | 50 | 27.4 | 45.7 | ||||
| Pain | 23.0 | 15.0 | n/a | n/a | 2.6 | 9.6 | 0 | 29.3 | n/a | ||||
| TKA, total knee arthroplasty | |||||||||||||
Large non-registry cohort studies echo similar findings with aseptic loosening and infection being the most common indication for revision TKA (Table II).35-46 Studies published from 2010 onwards clearly demonstrate that infection is the most common indication for revision within two years of the primary procedure, and aseptic loosening is the most common indication for late revision.36,38,40-43 This trend has also been observed in NJRs.26-30 Polyethylene wear and its sequelae of osteolysis and late instability are now uncommon. However, this was the dominant indication for revision in studies published prior to 2006, accounting for almost 25% of all revisions (Table II).35,44-47 Advancement in surgical techniques, tribology and polyethylene manufacturing (gamma sterilisation in inert environments and the use of highly cross-linked polyethylene) may have contributed to fewer polyethylene- and implant-related failures.37 Fehring et al46 concluded that the total number of early revision TKAs and the overall rate of revision could be reduced by 40% and 25%, respectively, if all TKAs were routinely cemented with careful balancing of the ligaments.
Table II
Indications for revision surgery over the last decade from retrospective cohort studies
| Study | Study Design | No. of revision TKAs | Male / female (No. of patients) | Mean age at RKA (yrs) | Years of operations | Mean follow-up times | Mean time to revision | Ratio of early*/ late† revision | Causes of revision knee procedure |
|---|---|---|---|---|---|---|---|---|---|
| Sharkey et al, 201435 US | Single centre RR | 781 | 291 / 453 | Male: 65.4 (37 to 96): Female: 65.1 (34 to 96) | 2003 to 2012 | NA | 0.84* (1day to 1.97 yrs) 6.9† (2 to 30.36) | 37.6% / 62.4% | Aseptic loosening 39.9% Infection 27.4% Instability 7.5% Fracture 4.7% Arthrofibrosis 4.5% PE wear: 3.5% |
| Le et al, 201437 US | Single centre RR | 253 | 103 / 143 | 64 | 2001 to 2011 | N/a | 35 mths (9 to 51) | 46% / 54% | Infection 24%*, 25%† Instability 26%*, 18%† Stiffness 18%*, 14%† PE wear 2%*, 9%† |
| Koh et al, 201438 Korea | Multi centre RR | 634 | NA | NA | 2008 to 2012 | 5 yrs | 76.5 mths (1 to 312) | 27% / 73% | Infection 38% Aseptic Loosening 33% Wear 13% Instability 7% Stiffness 3% |
| Kasahara et al, 201339 Japan | Multi centre RR | 140 | 27 / 113 | 72.9 (71.5 to 74.3) | 2006 to 2011 | 35 months (12 to 132) | 73 mth (2 to 420) | NA | Aseptic loosening 40% Infection 24% Wear 9% Instability 9% Implant failure: 6% Fracture: 4% |
| Dalury et al, 201340 US | Multi centre RR | 820 | 283 / 410 | 69 (24 to 94) | 2000 to 2011 | NA | 6 yrs (0.1 to 26) | 49% / 51%‡ | Aseptic loosening 23.1% Infection 18.4% PE wear 18.1% Instability 17.7% Pain/stiffness 9.3% Osteolysis 4.5% |
| Schroer et al, 201341 US | Multi centre RR | 844 | 313 / 531 | 65 | 2010 to 2011 | NA | 5.9 yrs (10 days to 31 yrs) | 35.3% / 64.7% | Aseptic Loosening 31.2% Instability 18.7% Infection 16.2% PE Wear 10% Arthrofibrosis 7% |
| Haasper et al47 Germany | Single Centre RR | 150 | NA | NA | 2009 | NA | 55 mths (0.5 to 125 mths) | NA | Aseptic loosening 30.7% Instability 24.7% Stiffness 17.3% Component malposition 6% |
| Bozic et al, 201042 US | Nationwide Inpatient Sample RR | 60 355 | 25 711 / 34 644 | 65.8 | 2005 to 2006 | NA | NA | NA | Infection 25.2% Aseptic loosening 16.1% Implant failure 9.7% |
| Hossain et al, 201043 UK | Single centre RR | 349 | 204 / 139 | 67.8 (32 to 94) | 1999 to 2008 | 57.7 mths | 84.3 mths (1 to 167) | 32% / 68% | Infection 35.7% Aseptic loosening 14.9% PE wear 12.3% Conversion from UKA 8.3% |
| Mulhall et al, 200644 US | Multicentre PCS 318 | 318 | NA | 68.7 (34 to 85) | NA | 1 yr | 7.9 yrs (0.5 to 27 yrs) | 69% / 31% | Osteolysis: 59.4% Aseptic loosening: 41.3% Instability 28.9% PE Wear: 24.9% Failed PE insert 18.1% Infection 10.4% |
| Sheng et al 200445 | MA on 33 studies & 1356 patients | NA | 429 / 611 | 67 (45 to 90) | 1990 to 2002 | 57 mths (6 to 108) | N/A | N/A | Aseptic loosening 55% PE wear 11% Instability 10% Disease progression 4% |
| Sharkey et al, 200235 US | Single centre RR | 212 | 75 / 128 | Male: 68.7 (31.6 to 88.7) Female: 68.4 (36.1 to 96.9) | 1997 to 2000 | NA | 1.1* (9 days to 2 yrs)7† (2.2 to 28) | 55.6% / 42.4% | PE wear: 25% Loosening: 24.1% Instability: 21.2% Infection: 17.5% Arthrofibrosis: 14.6% Malposition: 11.8% |
| Fehring et al, 200146 US | Single centre RR | 440 | 1986 to 1999 | 5 years | 19 months | 63% within 5 years of index procedure | Infection: 38% Instability: 27% Failure of ingrowth (cementless fixation) 13% Patellofemoral problems 8% PE wear 7% | ||
-
* Early revision (< 2 years post primary total knee arthroplasty) † Late revision (> 2 years post primary total knee arthroplasty) ‡ Defined early revision < 5 years post index procedure RR, Retrospective review, PCS: prospective cohort study, MA: meta-analysis, NA: not available; US, United States, UK, United Kingdom
Pain, instability, and stiffness are more likely to be reported now as causes of failure after TKA in NJRs26-29 and retrospective studies undertaken in the US,36,37,40,41 in contrast to multicentre studies from Korea and Japan.38,39 This observation may be partially explained by the fact that European NJRs and data series from the United States would include operations on a much younger cohort of patients. Such patients are more physically active with higher demands and expectations, therefore placing greater strain through the prostheses leading to earlier failure. Operating on younger patients highlights the importance of meticulous ligament balancing.
The risk factors for revision surgery
Patient factors
The age and gender of the patient are major factors affecting the outcome of primary TKA. All registries concur that the rate of revision increases with decreasing age.26-29,48 In Sweden, patients aged < 65 years have twice the risk of revision compared with those aged > 75 years, regardless of diagnosis or the type of implant.29 In Australia, four years post-operatively, patients with osteoarthritis aged < 55 years have > 4.5 times the risk of revision than do those aged > 75 years.27 Men have higher rates of revision than women (Fig. 2).26-29,38 In England, Wales, and Northern Ireland, the Cumulative Risk of Revision (CRR) in men aged > 75 years is 2% ten years post-operatively, whereas in those aged < 55 years it is 12%.26 A large multicentre study reported similar findings: for every ten years of increasing age, the overall risk of failure of TKA decreased by 60% (OR 0.41; 95% confidence interval (CI) 0.37 to 0.49).38 It is therefore imperative that younger patients are fully informed of the increased risk of revision by referencing registry data, especially as the demand for TKA in younger patients rises (Fig. 3). The reasons for this demand include higher patient expectations and an increased incidence of degenerative joint disease resulting from the obesity pandemic.49,50
Fig. 2
Kaplan–Meier estimates of the cumulative percentage probability of a first revision of primary knee arthroplasty broken down by age group and gender at increasing years after primary surgery. (Reproduced with permission from NJR England Wales Northern Ireland, Annual Report 2014).
Fig. 3
The relative distribution of primary arthroplasties among different age groups (%) (all types of implants) since 1975. (Reproduced from Swedish Knee Arthroplasty Register, Annual Report 2014).
Patients with rheumatoid arthritis who undergo TKA have the highest rate of revision secondary to infection, especially in men.29,51,52 However, with improved medical management the number of rheumatoid patients requiring TKA has fallen dramatically.29
Surgical technique
The year of primary operation affects the likelihood of revision. The cumulative percentage probability of a revision being undertaken increases slightly for each operative year between 2003 and 2007.26 Thus, the risk of revision five years post-operatively for patients who underwent TKA in 2008 was 3% compared with 1.7% for those whose operation was undertaken in 2003 (Fig. 4).26 This variation may reflect incomplete data in the NJR until about 2007/8. The Swedish Knee Arthroplasty Register (SKAR) has captured a large range of data over a much greater period and illustrates the reducing risk for revision over time (Fig. 5).
Fig. 4
Changes in cumulative percentage chance of implant failure by year of primary operation using Kaplan–Meier estimates. (Reproduced with permission from NJR England Wales Northern Ireland, Annual Report 2014).
Fig. 5
Cumulative rate of revision after total knee arthroplasty (%). (Reproduced with permission from Swedish Knee Arthroplasty Register, Annual Report 2014).
When considering all knee procedures, partial arthroplasties account for between 7% and 9%, primary TKAs for between 83% and 88%, and revision TKAs for between 5% and 8%.26-29 Partial arthroplasty including unicompartmental, patellofemoral, bicompartmental, unispacer, and partial resurfacing, has a higher rate of revision than TKA.26-29 Unicompartmental arthroplasty (UKA) comprises > 90% of all partial procedures.26-28 Proponents of UKA cite as advantages the preservation of normal kinematics, lower peri-operative morbidity, blood loss and risk of infection compared with TKA, and an accelerated rehabilitation and recovery.53-55 However, the volume of UKAs which are undertaken has diminished over time and continues to decline (Fig. 6).26,27,29 The Australian registry reports that the number of UKAs undertaken in 2013 had decreased by 2.7% compared with the previous year and by 49.5% compared with 2003.27 One reason for this decline is that UKA performed for osteoarthritis has a substantially higher risk of revision26,27,29 and for re-operation.55 The rate of revision for UKA has been reported to be twice as high as the equivalent group of TKAs at each period of review post-operatively, and three times higher at ten years post-operatively.26 The CRR for primary UKA ranges from 12.7% at ten years26 to 19.1% at 13 years27 post-operatively. One reason for the higher failure rates is offering UKA to younger patients.26,29 This delays the need for a TKA in younger patients with mild disease but often such patients are more active and put a greater amount of strain on their implants. This explains the high CRR ten years post-operatively in patients aged < 55 years who undergo UKA (21%) compared with that of patients aged > 75 years who undergo UKA (6%). It has been estimated that if 100 patients undergoing TKA received a UKA instead this would result in one fewer death and three more re-operations during the first four post-operative years.55
Fig. 6
The relative yearly distribution of types of implant used for primary surgery. (Reproduced with permission from Swedish Knee Arthroplasty Register, Annual Report 2014).
Finally, the risk of subsequent revision is twice as high in patients who have undergone conversion from a UKA to a TKA than for primary TKA.29 Revision of a failed UKA to another UKA is rare and reflects the higher risk of re-revision associated with this than if the UKA is revised to a TKA.27,29
Information from the New Zealand Joint Registry shows no difference in the rate of revision in relation to the operative approach (i.e. medial or lateral parapatellar approach).28 This registry also showed that patients were statistically twice as likely to develop periprosthetic infection requiring revision surgery within six months of the initial procedure if it had been performed in a laminar flow theatre (revision risk (RR) 0.24%) rather than a conventional operating theatre (RR 0.11%). On further analysis, the rates of revision for deep infection within six months were 2.9 times higher if the operation was undertaken in a conventional theatre using space suits (RR 0.26%) and 2.8 times higher if it was undertaken in a laminar flow theatre using space suits (RR 0.25%), compared with operating in a conventional theatre without space suits (RR 0.09%). Furthermore, the rate of revision rate doubled with the use of spacesuits in either conventional or laminar flow theatres.28 Such findings pose an interesting predicament and warrant further research and validation.
Historically, there has not been overwhelming evidence to support the use of bone cement or resurfacing of the patella during primary TKA; the choice lying with the surgeon. Data from NJRs show that the risk of revision is substantially increased with the use of uncemented compared with cemented implants.26-29 They also confirm that > 80% of TKAs are performed with cement. Cox regression analysis, adjusting for age, gender, year of operation, and use of patellar component, showed that the risk of revision for TKA with an uncemented tibial component was 1.7 (1.4 to 1.9) times higher than for that if a cemented component was used.29
Resurfacing the patella is extremely variably undertaken across countries and remains an area of controversy (Table III).56,57 A patellar button was used in > 70% of TKA cases in Sweden during the 1980s, but its use fell to 2.3% in 2013 (Fig. 7).29 TKA performed without patellar resurfacing in Sweden between 1991 and 2000, however, had a significantly higher rate of revision than those performed with patellar resurfacing (RR × 1.3, 95% CI 1.1 to 1.4). Conversely, TKA performed with patellar resurfacing between 1999 and 2008 had a higher risk of revision (RR × 1.2, 95% CI 1.1 to 1.4). In contrast, the Australian registry reports an increase in patellar resurfacing from 41% of all TKAs in 2005 to 56.8% in 2013.27 Estimating the impact of patellar resurfacing on the risk of revision is extremely complex and cannot explain the variations in the findings worldwide. Furthermore, registry data cannot accurately interpret the impact of patellar resurfacing on the success of TKA, especially if a secondary operation was required to resurface the patella to manage subsequent patellofemoral symptoms.
Fig. 7
The yearly distribution (%) of the use of patella button in total knee arthroplasty (TKA). (Reproduced with permission from SKAR Annual report 2014).
Table III
The percentage of patients that had their patella resurfaced as part of their total knee arthroplasties performed in various countries
| Swedish Knee Arthroplasty Register annual report 201429 | Norwegian Arthroplasty Register annual report 201056 | Danish Knee Arthroplasty Register annual report 201357 | Australia Orthopaedic Association NJR annual report 201427 | New Zealand joint Registry 15-year report27 |
|---|---|---|---|---|
| 2.3 | 10.4 | 77 | 56.8 | 32 |
| NJR, National Joint Registry |
Almost one quarter of TKAs undertaken in Australia are computer navigated.27 At ten years post-operatively, there is no overall difference in CRR between navigated and non-navigated TKAs. However, navigated TKAs performed in patients aged < 65 years are less likely to require revision for loosening or osteolysis.27 Finally, the outcome is better if the operation is undertaken by a surgeon who performs a high volume of operations. This has been reported several times.58,59
Implant design
In Australia an unconstrained (cruciate retaining) prosthesis is most commonly used (79.2%) followed by a posterior stabilised (cruciate sacrificing) TKA (28%).27 This is similar in England where an unconstrained prosthesis is the most popular (65.6%) followed by a posterior stabilised TKA (22.8%). Overall, the most commonly used TKA in England is a cemented unconstrained fixed bearing design, comprising 55.9%.26 This design has the lowest rate of revision with a CRR ten years post-operatively of 3%. Posterior stabilised designs have the highest rates of revision.26,28 For cemented TKA, the mobile-bearing posterior stabilised prosthesis has CRR ten years post-operatively of 5%; and for uncemented TKA, the fixed-bearing posterior stabilised prosthesis has a CRR at ten years post-operatively of 9%.26 In contrast, the AJRR’s first annual report (2013) clearly shows that the most popular design of TKA is posterior stabilised (59.1%) followed by unconstrained (40.1%).30 The impact of this on revision has not been estimated, perhaps as the registry is in its infancy.
The Australian Registry showed that fixed tibial bearings had a lower rate of revision compared with rotating or slide-rotating mobile tibial bearings.27 There were no differences in the rates of revision between the modular fixed tibial component and the non-modular all-polyethylene tibial component.27 However, the all-polyethylene tibial components have higher rates of revision compared with the polyethylene moulded tibial components.27 A large single centre retrospective study recently reported that all-polyethylene tibial components had a significantly lower risk of revision compared with metal-backed modular implants, (hazard ratio, 0.3; 95% CI 0.2 to 0.5; p < 0.0001) and therefore were associated with a better outcome.60 The authors also reported that in metal-backed modular tibial designs, unconstrained TKA performed better than the posterior-stabilised TKA (p = 0.002), which is consistent with registry findings.
Bearing surfaces have an important but undetermined role in TKA. Prostheses using high cross-linked polyethylene (XLPE) have a lower rate of revision compared with those using non cross-linked polyethylene.27 The Australian Registry has shown a cumulative incidence of loosening or lysis of 0.8%, ten years post-operatively, for TKA using XLPE and 1.7% for TKA using non-XLPE. The overall use of XLPE in Australia, however, is much less than non-XLPE.27 In the United States, XLPE is used in 75% of TKAs and non-XLPE in 25%.30
Implant model
Registry analysis of TKAs by brand has shown a risk of revision ten years post-operatively of < 5% with overlapping CIs indicating similar performances in terms of the risk of revision.26 In England, the highest rates of revision related to the Rotaglide + mobile knee TKA which had a revision rate of 7.14% at ten years, and the Preservation UKA which had a revision rate of 20.14% at ten years.26 Using Cox regression analysis and adjustments for differences in gender, age and year of operation, the SKAR also found that the Preservation UKA had the highest risk ratio for revision (RR 1.57, p = 0.04; 95% CI 1.02 to 2.40).29 The TKAs performed in Sweden with the highest risk of revision are the F/S MIII (RR 1.45, p < 0.01; 95% CI 1.21 to 1.74) and the PFC RP (RR 1.67, p < 0.01; 95% CI 1.27 to 2.19).29 The New Zealand Registry report that the Insall/Burstein (1.8 per 100 observed component-years; ocys) and Optetrak (0.93 ocys) prostheses have a significantly higher rate of revision than the overall rate of 0.5 per 100 ocys at the 95% CI.28
The Australian Registry reports that the Nexgen CR/NExgen TKA has the overall lowest CRR ten years post-operatively of 3.2% (2.9% if uncemented and 2.7% if cemented).27 Using a standardised approach, they identified six types of TKA as outliers with respect to CRR. These include the Genesis II CR (cementless) / Profix Mobile (cemented) combination (ten year CRR, 13%); Genesis II Oxinium PS (cemented) / Genesis II (Cementless) combination (five year CRR, 31%); GMK Primary (cementless)/ GMK Primary (Cementless) combination (one year CRR, 3%); Optetrak CR (cemented) / Optetrak (cemented) combination (ten year CRR, 11.9%); Scorpio NRG PS (cementless) / series 7000 (cementless) combination (five year CRR, 8.4%); Vanguard PS / Maxim combination (five year CRR, 5.8%); and the LCS PS femoral component (three year CRR, 7.5%). The comparator group includes all other types within the same class, regardless of their rate of revision.
In conclusion, registries are an indispensable tool in providing comparative data from large populations on the outcomes of individual TKAs in a community. Such data are critical to an evidence-based approach to the choice of prosthesis. Registries also serve as a system of quality-assurance surveillance to monitor the outcomes of new devices, generate observational evidence, and cost savings. However, there remains an important role for well conducted large-scale randomised controlled trials to assess individual types of prostheses to a depth which cannot be achieved by registries.20 This will also minimise the introduction of new technologies directly from industry into health care. The National Institute for Health and Clinical Excellence (NICE) in the United Kingdom has set the benchmark for approving new prostheses at 10% revision ten years post-operatively.
The overall percentage of patients requiring revision TKA is low. However, with the large number of TKAs performed annually, this translates into a significant absolute number with considerable financial and healthcare implications for patients and healthcare systems. The survivorship of TKAs is determined by the changing demographics of patients, surgical technique and implant-related factors. The most common modes of failure include aseptic loosening, infection, pain and instability. Future research, education, and training should consider regenerative techniques that halt or reverse degenerative joint disease in younger patients, improving the selection of patients for surgery and the appropriate early evaluation and long-term monitoring of outcomes through prospective studies and NJRs.
1 Ethgen O , BruyèreO, RichyF, DardennesC, ReginsterJY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg [Am]2004;86-A:963–974. Google Scholar
2 Jenkins PJ , ClementND, HamiltonDF, et al.Predicting the cost-effectiveness of total hip and knee replacement: a health economic analysis. Bone Joint J2013;95-B:115–121. Google Scholar
3 Kurtz S , OngK, LauE, MowatF, HalpernM. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg [Am]2007;89-A:780–785. Google Scholar
4 Singh JA . Epidemiology of knee and hip arthroplasty: a systematic review. Open Orthop J2011;5:80–85. Google Scholar
5 McWilliams AB , DouglasSL, RedmondAC, et al.Litigation after hip and knee replacement in the National Health Service. Bone Joint J2013;95-B:122–126. Google Scholar
6 Price AJ , LonginoD, ReesJ, et al.Are pain and function better measures of outcome than revision rates after TKR in the younger patient?Knee2010;17:196–199. Google Scholar
7 Bourne RB , ChesworthBM, DavisAM, MahomedNN, CharronKD. Patient satisfaction after total knee arthroplasty: who is satisfied and who is not?Clin Orthop Relat Res2010;468:57–63. Google Scholar
8 Brander VA , StulbergSD, AdamsAD, et al.Predicting total knee replacement pain: a prospective, observational study. Clin Orthop Relat Res2003;416:27–36. Google Scholar
9 Fisher DA , DierckmanB, WattsMR, DavisK. Looks good but feels bad: factors that contribute to poor results after total knee arthroplasty. J Arthroplasty2007;22(suppl2):39–42. Google Scholar
10 Scott CE , HowieCR, MacDonaldD, BiantLC. Predicting dissatisfaction following total knee replacement: a prospective study of 1217 patients. J Bone Joint Surg [Br]2010;92-B:1253–1258. Google Scholar
11 Kim TK , ChangCB, KangYG, KimSJ, SeongSC. Causes and predictors of patient’s dissatisfaction after uncomplicated total knee arthroplasty. J Arthroplasty2009;24:263–271. Google Scholar
12 Jacobs CA , ChristensenCP, KarthikeyanT. Patient and intraoperative factors influencing satisfaction two to five years after primary total knee arthroplasty. J Arthroplasty2014;29:1576–1579. Google Scholar
13 Von Keudell A , SodhaS, CollinsJ, et al.Patient satisfaction after primary total and unicompartmental knee arthroplasty: an age-dependent analysis. Knee2014;21:180–184. Google Scholar
14 Ali A , SundbergM, RobertssonO, et al.Dissatisfied patients after total knee arthroplasty: a registry study involving 114 patients with 8-13 years of followup. Acta Orthop2014;85:229–233. Google Scholar
15 Singh JA , LewallenDG. Ninety-day mortality in patients undergoing elective total hip or total knee arthroplasty. J Arthroplasty2012;27:1417–1422. Google Scholar
16 Zmistowski B , RestrepoC, KahlLK, ParviziJ, SharkeyPF. Incidence and reasons for nonrevision reoperation after total knee arthroplasty. Clin Orthop Relat Res2011;469:138–145. Google Scholar
17 No authors listed. Health and Social Care Information Centre. Provisional Monthly Patient Reported Outcome Measures (PROMs) in England – April 2014 to October 2014. http://www.hscic.gov.uk/catalogue/PUB17024 (date last accessed 25 June 2015). Google Scholar
18 Lizaur-Utrilla A , Gonzalez-ParreñoS, Miralles-MuñozFA, Lopez-PratsFA, Gil-GuillenV. Patient-related predictors of treatment failure after primary total knee arthroplasty for osteoarthritis. J Arthroplasty2014;29:2095–2099. Google Scholar
19 No authors listed. HSCIC: Patient Reported Outcome Measures (PROMs) in England – April 2014 to October 2014. http://www.hscic.gov.uk/proms (date last accessed 16 June 2015). Google Scholar
20 Delaunay C . Registries in orthopaedics. Orthop Traumatol Surg Res2015;101(suppl):S69–S75. Google Scholar
21 Dreyer NA , GarnerS. Registries for robust evidence. JAMA2009;302:790–791. Google Scholar
22 Robertsson O . Knee arthroplasty registers. J Bone Joint Surg [Br]2007;89-B:1–4. Google Scholar
23 Graves SE . The value of arthroplasty registry data. Acta Orthop2010;81:8–9. Google Scholar
24 Knutson K , RobertssonO. The Swedish Knee Arthroplasty Register (www.knee.se). Acta Orthop2010;81:5–7. Google Scholar
25 Robertsson O , RanstamJ, SundbergM, W-DahlA, LidgrenL. The Swedish Knee Arthroplasty Register: a review. Bone Joint Res2014;3:217–222. Google Scholar
26 No authors listed. National Joint Registry for England, Wales and Northern Ireland. 11th annual report, 2014. http://www.njrreports.org.uk/Portals/0/PDFdownloads/NJR%2011th%20Annual%20Report%202014.pdf (date last accessed 25 June 2015). Google Scholar
27 No authors listed. Australian Orthopaedic Association National Joint Replacement Registry: Annual Report, 2014. https://aoanjrr.dmac.adelaide.edu.au/annual-reports-2014 (date last accessed 16 June 2015). Google Scholar
28 No authors listed. The New Zealand Joint Registry. Fifteen Year Report, 2014. http://www.nzoa.org.nz/ (date last accessed 16 June 2015). Google Scholar
29 No authors listed. Swedish Knee Arthroplasty Register. Annual Report 2014. Edited, http://www.myknee.se/en/ (date last accessed 16 June 2015). Google Scholar
30 No authors listed. American Joint Replacement Registry. First Annual Report on Hip and Knee Arthroplasty Data, 2013. https://teamwork.aaos.org/ajrr/AJRR%20Documents/AJRR_2013_Annual_Report.pdf. (date last accessed 16 June 2015). Google Scholar
31 Sadoghi P , LiebensteinerM, AgreiterM, et al.Revision surgery after total joint arthroplasty: a complication-based analysis using worldwide arthroplasty registers. J Arthroplasty2013;28:1329–1332. Google Scholar
32 No authors listed. National Joint Registry for England and Wales. 1st Annual Report, 2004. http://www.njrcentre.org.uk/NjrCentre/Portals/0/Documents/England/Reports/NJR_AR_1.pdf (date last accessed 25 June 2015). Google Scholar
33 No authors listed. Swedish Knee Arthroplasty Register. Annual report, 1999. http://myknee.se/pdf/97_report1999.pdf (date last accessed 25 June 2015). Google Scholar
34 No authors listed. Australian Orthopaedic Association: National Joint Replacement Registry, 1st Annual Report, 2000. https://aoanjrr.dmac.adelaide.edu.au/documents/10180/75186/Annual%20Report%202000?version=1.0& t=1349407109407. (date last accessed 16 June 2015). Google Scholar
35 Sharkey PF , HozackWJ, RothmanRH, ShastriS, JacobySM. Insall Award paper. Why are total knee arthroplasties failing today?Clin Orthop Relat Res2002;404:7–13. Google Scholar
36 Sharkey PF , LichsteinPM, ShenC, TokarskiAT, ParviziJ. Why are total knee arthroplasties failing today--has anything changed after 10 years?J Arthroplasty2014;29:1774–1778. Google Scholar
37 Le DH , GoodmanSB, MaloneyWJ, HuddlestonJI. Current modes of failure in TKA: infection, instability, and stiffness predominate. Clin Orthop Relat Res2014;472:2197–2200. Google Scholar
38 Koh IJ , ChoWS, ChoiNY, KimTK. , Kleos Korea Research Group. Causes, risk factors, and trends in failures after TKA in Korea over the past 5 years: a multicenter study. Clin Orthop Relat Res2014;472:316–326. Google Scholar
39 Kasahara Y , MajimaT, KimuraS, NishiikeO, UchidaJ. What are the causes of revision total knee arthroplasty in Japan?Clin Orthop Relat Res2013;471:1533–1538. Google Scholar
40 Dalury DF , PomeroyDL, GorabRS, AdamsMJ. Why are total knee arthroplasties being revised?J Arthroplasty2013;28(suppl):120–121. Google Scholar
41 Schroer WC , BerendKR, LombardiAV, et al.Why are total knees failing today? Etiology of total knee revision in 2010 and 2011. J Arthroplasty2013;28(suppl):116–119. Google Scholar
42 Bozic KJ , KurtzSM, LauE, et al.The epidemiology of revision total knee arthroplasty in the United States. Clin Orthop Relat Res2010;468:45–51. Google Scholar
43 Hossain F , PatelS, HaddadFS. Midterm assessment of causes and results of revision total knee arthroplasty. Clin Orthop Relat Res2010;468:1221–1228. Google Scholar
44 Mulhall KJ , GhomrawiHM, ScullyS, CallaghanJJ, SalehKJ. Current etiologies and modes of failure in total knee arthroplasty revision. Clin Orthop Relat Res2006;446:45–50. Google Scholar
45 Sheng P , LehtoM, KatajaM, et al.Patient outcome following revision total knee arthroplasty: a meta-analysis. Int Orthop2004;28:78–81. Google Scholar
46 Fehring TK , OdumS, GriffinWL, MasonJB, NadaudM. Early failures in total knee arthroplasty. Clin Orthop Relat Res2001;392:315–318. Google Scholar
47 Haasper C , KendoffD, GebauerM, GehrkeT, KlauserW. Revision of unconstrained total knee arthroplasty - a technical analysis. Z Orthop Unfall2012;150:290–295. Google Scholar
48 Julin J , JämsenE, PuolakkaT, KonttinenYT, MoilanenT. Younger age increases the risk of early prosthesis failure following primary total knee replacement for osteoarthritis. A follow-up study of 32,019 total knee replacements in the Finnish Arthroplasty Register. Acta Orthop2010;81:413–419. Google Scholar
49 Lohmander LS , Gerhardsson de VerdierM, et al.Incidence of severe knee and hip osteoarthritis in relation to different measures of body mass: a population-based prospective cohort study. Ann Rheum Dis2009;68:490–496. Google Scholar
50 Wang Y , SimpsonJA, WlukaAE, et al.Relationship between body adiposity measures and risk of primary knee and hip replacement for osteoarthritis: a prospective cohort study. Arthritis Res Ther2009;11:R31. Google Scholar
51 Jämsen E , HuhtalaH, PuolakkaT, MoilanenT. Risk factors for infection after knee arthroplasty. A register-based analysis of 43,149 cases. J Bone Joint Surg [Am]2009;91-A:38–47. Google Scholar
52 Schrama JC , EspehaugB, HallanG, et al.Risk of revision for infection in primary total hip and knee arthroplasty in patients with rheumatoid arthritis compared with osteoarthritis: a prospective, population-based study on 108,786 hip and knee joint arthroplasties from the Norwegian Arthroplasty Register. Arthritis Care Res (Hoboken)2010;62:473–479. Google Scholar
53 Borus T , ThornhillT. Unicompartmental knee arthroplasty. J Am Acad Orthop Surg2008;16:9–18. Google Scholar
54 Furnes O , EspehaugB, LieSA, et al.Failure mechanisms after unicompartmental and tricompartmental primary knee replacement with cement. J Bone Joint Surg [Am]2007;89-A:519–525. Google Scholar
55 Liddle AD , JudgeA, PanditH, MurrayDW. Adverse outcomes after total and unicompartmental knee replacement in 101,330 matched patients: a study of data from the National Joint Registry for England and Wales. Lancet2014;384:1437–1445. Google Scholar
56 No authors listed. The Norwegian Arthroplasty Register. Annual Report, 2010. Edited, http://nrlweb.ihelse.net/eng/Report_2010.pdf (date last accessed 25 June 2015). Google Scholar
57 No authors listed. Danish Knee Arthroplasty Register. Annual Report, 2013. http://www.dkar.dk/ (date last accessed 8 July 2015). Google Scholar
58 Katz JN , BarrettJ, MahomedNN, et al.Association between hospital and surgeon procedure volume and the outcomes of total knee replacement. J Bone Joint Surg [Am]2004;86-A:1909–1916. Google Scholar
59 Shervin N , RubashHE, KatzJN. Orthopaedic procedure volume and patient outcomes: a systematic literature review. Clin Orthop Relat Res2007;457:35–41. Google Scholar
60 Kremers HM , SierraRJ, SchleckCD, et al.Comparative Survivorship of Different Tibial Designs in Primary Total Knee Arthroplasty. J Bone Joint Surg [Am]2014;96-A:121. Google Scholar
Author contributions:
M. Khan: Writing paper, research, data analysis.
K. Osman: Writing paper, research.
G. Green: Writing paper.
F. S. Haddad: Writing paper.
This research / study / project was supported by The National Institute for Health Research University College London Hospitals Biomedical Research Centre.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
This article was primary edited by J. Scott.
This paper is based on a study which was presented at the 2014 Current Concepts in Joint Replacement® - The Journey Continues meeting held in Iguassu Falls, Brazil, 17th-20th September.
