Introduction. The revision rate of unicompartmental knee replacement (UKR) in national joint registries is much higher than that of total knee replacements and that of UKR in cohort studies from multiple high-volume centres. The reasons for this are unclear but may be due to incorrect patient selection, inadequate surgical technique, and inappropriate indications for revision. Meniscal bearing UKR has well defined evidence based indications based on preoperative radiographs, the surgical technique can be assessed from post-operative radiographs and the reason for revision from pre-revision radiographs. However, for an accurate assessment aligned radiographs are required. The aim of the study was to determine why the revision rate of UKR in registries is so high by undertaking a radiographic review of revised UKR identified by the United Kingdom's (UK) National Joint Registry (NJR). Methods. A novel cross-sectional study was designed. Revised medial meniscal bearing UKR with primary operation registered with the NJR between 2006 and 2010 were identified. Participating centres from all over the country provided blinded pre-operative, post-operative, and pre-revision radiographs. Two observers reviewed the radiographs. Results. Radiographs were provided for 107 revised UKR from multiple centres. The recommended indications were not satisfied in 30%. The most common reason was the absence of bone-on-bone arthritis, and in 16 (19%) the medial joint space was normal or nearly normal. Post-operative films were mal-aligned in 50%. Significant surgical errors were seen in 50%, with most errors attributable to tibial component placement and orientation. No definite reason for revision was identified in 67%. Reasons for revision included disease progression (10%), tibial component loosening (7%), dislocation of the bearing (7%), infection (6%) femoral component loosening (3%), and peri-prosthetic fracture (2% - one femur, one tibia). Discussion and Conclusion. This study found that improper patient selection, inadequate surgical technique, inappropriate revisions and poorly taken radiographs all contributed to the high revision rate. There is a misconception that UKR should be used for early OA. Bone-on-bone arthritis is a requirement and was definitely not present in about 20%. There were many surgical errors, particularly related to the tibial cut: The new instrumentation should reduce this. There was a high prevalence of mal-aligned radiographs. Revisions should be avoided unless there is a definite problem, as the outcome of revision is usually poor in this situation. 80% of UKR revisions could potentially be avoided if surgeons adhered to the recommended indications for primary and revision surgery, and used the recommended surgical techniques. This study therefore suggests that if UKR was used appropriately the revision rate would be substantially lower and probably similar to that of TKR

Orthopaedic practice is increasingly guided by conclusions drawn from analysis of Joint Registry Data. Analysis of the England and Wales National Joint Registry (NJR) led Sibanda et al to conclude that UKR should be reserved for more elderly patients due to higher revision rates in younger patients. To determine our UKR revision rates at the Great Western Hospital we requested knee arthroplasty data from the NJR, Hospital Episode Statistics (HES) data submitted by our centre to the Primary Care Trust, and interrogated our internal theatre implant database. This revealed significant discrepancies between different data sources. We collected data from each source for 2005, 2006, and 2007. Operations were classified as TKR, UKR, Other or Unspecified. Results are illustrated in the attached table:. Key findings:. Our theatre implant database appears most accurate and includes a greater number of joint replacement operations than NJR or HES data and fewer ‘unspecified’ procedures. On average 15% NJR, 0% HES and 0.3% theatre data procedures were ‘unspecified’. NJR data comprises an average 17 fewer, and HES data an average 36 fewer procedures each year compared with our theatre data. Up to 80% UKRs performed are recorded as TKR in HES data. In summary there is significant inaccuracy in our NJR data which may affect the validity of conclusions drawn from NJR data analysis. HES data is even less accurate with implications for hospital funding. We strongly advise other centres to continue to maintain accurate implant data and to perform a similar audit to calculate error rates for NJR and HES data. Further analysis is required to identify at which stage of data collection inaccuracies occur so that solutions can be devised. We are currently analysing data from 2008 and 2009

Introduction. Unicompartmental knee replacement (UKR) offers advantages over total knee replacement but has higher revision rates particularly for aseptic loosening. Cementless UKR was introduced in an attempt to address this. We used National Joint Registry (NJR) data to compare the 10-year results of cemented and cementless mobile bearing UKR whilst matching for important patient, implant and surgical factors. We also explored the influence of caseload on outcome. Methods. We performed a retrospective observational study using NJR data on 30,814 cemented and 9,708 cementless mobile bearing UKR implanted between 2004 and 2016. Logistic regression was utilised to calculate propensity scores allowing for matching of cemented and cementless groups for various patient, implant and surgical confounders, including surgeon's caseload, using a one to one ratio. 14,814 UKRs (7407 cemented and 7407 cementless) were propensity score matched. Outcomes studied were revision, defined as removal, addition or exchange of a component, and reasons for revision. Implant survival was compared using Cox regression models and groups were stratified according to surgeon caseload. Results. Based on raw unmatched data the 10 year survival for cementless and cemented UKR were 89% (95% CI 88%–90%) and 93% (CI 90%–96%), with cementless having a lower revision rate (Hazard ratio (HR)=0.59 (CI 0.52–0.68, p<0.001). However, there were differences between the cohorts in many potential confounding factors particularly surgeons caseload: Surgeons using cementless had a higher caseloads than those using cemented and for both cohorts the revision rate decreased with increasing caseload. Following matching, all potential confounders were well balanced and the 10-year survival for cementless and cemented were 90% (CI 88%–92%) and 93% (95% CI 90–96%) with cementless having a lower revision rate (HR 0.76; CI 0.64–0.91; p=0.003). This was due to rate of revision for aseptic loosening more than halving (p<0.001) in the cementless (n=31, 0.4%) compared to cemented (n=74, 1.0%) and the rate of revision for pain decreasing (p=0.03) in the cementless (n=34, 0.5%) compared to the cemented (n=55, 0.7%). However, the rate of peri-prosthetic fracture increased significantly (p=0.01) in the cementless (n=19, 0.3%) compared to the cemented (n=7, 0.1%). Following matching the decrease in revision rate with the cementless was similar for low (<10 cases/year; HR 0.74), medium (10–30 cases/year; HR 0.79) and high (>10 cases/year; HR 0.79) caseload surgeons. The 10- year survival for cementless and cemented were for low caseload 87% & 82%, medium caseload 94% & 92% and high caseload 98% & 94% respectively. Conclusions. This is the first study to compare the 10-year survival of the cementless and cemented mobile bearing UKR. We have demonstrated that the cementless device has a 24% reduced risk of revision and that this was independent of surgeon caseload and other important patient, surgical and implant confounders. This improvement was due to the rate of revision for aseptic loosening and pain halving. However, there was a small increase in rate of periprosthetic fracture. The results of both cemented and cementless UKR improved with increasing surgeon caseload. Low volume surgeons have poor results with both cemented and cementless UKR so should consider either stopping doing UKR or doing more. Medium and high volume surgeons should consider using the cementless. High volume surgeons using the cementless had particularly good results with a 10-year survival of 98%. For figures, tables, or references, please contact authors directly

Purpose. To establish the reliability of reporting and recording revision hip and knee arthroplasties by comparing data in the National Joint Registry (NJR), Hospital Episode Statistics (HES) and our local theatre records. Methods. The paper theatre registers for all orthopaedic theatres in the Royal Derby Hospitals NHS Trust were examined for details of revision hip and knee replacements carried out in 2007 and 2008. This was then cross-checked and merged with the local electronic theatre data to obtain a definitive local record of all revision hip and knee arthroplasties. Data for the same period was requested from the NJR and HES and these data were checked against our definitive local record for discrepancies. The HES codes used were the same codes used to compile the recent NJR annual reports. Results. The theatre registers and ORMIS identified 271 revision hip and knee arthroplasties in the study period. The NJR had corresponding data for 176 (65%) of these, and HES had 250 (92%). 10 cases (4%) were not recorded by either NJR or HES: 8 secondary resurfacings of patellae and 2 posterior lip augmentations in hips. Of those operations “missed” by HES, most had been assigned a correct “W” code, but had a “Y” or “Z” OPCS code not used in the NJR annual reports. Conclusion. When HES and the NJR data are combined, they are an accurate representation of real practice. More robust methods of reporting revision arthroplasty to the NJR are required. The OPCS codes used to indicate a revision need to be reviewed

Abstract

Abstract

Aims. The aim of this study was to evaluate medium- to long-term outcomes and complications of the Stanmore Modular Individualised Lower Extremity System (SMILES) rotating hinge implant in revision total knee arthroplasty (rTKA) at a tertiary unit. It is hypothesized that this fully cemented construct leads to satisfactory clinical outcomes. Methods. A retrospective consecutive study of all patients who underwent a rTKA using the fully cemented SMILES rotating hinge prosthesis between 2005 to 2018. Outcome measures included aseptic loosening, reoperations, revision for any cause, complications, and survivorship. Patients and implant survivorship data were identified through both prospectively collected local hospital electronic databases and linked data from the National Joint Registry/NHS Personal Demographic Service. Kaplan-Meier survival analysis was used at ten years. Results. Overall, 69 consecutive patients (69 knees) were included with a median age of 78 years (interquartile range 69 to 84), and there were 46 females (66.7%). Indications were septic revisions in 26 (37.7%), and aseptic aetiology in the remining 43 (62.3%). The mean follow-up was 9.7 years (4 to 18), and the overall complication was rate was 7.24%, all with patellofemoral complications. Failure rate with ‘any cause revision’ was 5.8%. There was one case of aseptic loosening of the femoral component. At ten years, 17/69 patients (24.63%) had died, and implant survivorship was 92.2%. Conclusion. In our experience, the SMILES rotating hinge prosthesis achieves satisfactory long-term outcomes with ten-year implant survivorship of 92.2% and a patellofemoral complication rate of 7.24%. Cite this article: Bone Jt Open 2023;4(10):776–781

Aims. Aseptic loosening is the most common cause of failure following cemented total knee arthroplasty (TKA), and has been linked to poor cementation technique. We aimed to develop a consensus on the optimal technique for component cementation in TKA. Methods. A UK-based, three-round, online modified Delphi Expert Consensus Study was completed focusing on cementation technique in TKA. Experts were identified as having a minimum of five years’ consultant experience in the NHS and fulfilling any one of the following criteria: a ‘high volume’ knee arthroplasty practice (> 150 TKAs per annum) as identified from the National joint Registry of England, Wales, Northern Ireland and the Isle of Man; a senior author of at least five peer reviewed articles related to TKA in the previous five years; a surgeon who is named trainer for a post-certificate of comletion of training fellowship in TKA. Results. In total, 81 experts (round 1) and 80 experts (round 2 and 3) completed the Delphi Study. Four domains with a total of 24 statements were identified. 100% consensus was reached within the cement preparation, pressurization, and cement curing domains. 90% consensus was reached within the cement application domain. Consensus was not reached with only one statement regarding the handling of cement during initial application to the tibial and/or femoral bone surfaces. Conclusion. The Cementing Techniques In Knee Surgery (CeTIKS) Delphi consensus study presents comprehensive recommendations on the optimal technique for component cementing in TKA. Expert opinion has a place in the hierarchy of evidence and, until better evidence is available these recommendations should be considered when cementing a TKA. Cite this article: Bone Jt Open 2023;4(9):682–688

Aims. The aim of this study was to identify variables associated with time to revision, demographic details associated with revision indication, and type of prosthesis employed, and to describe the survival of hinge knee arthroplasty (HKA) when used for first-time knee revision surgery and factors that were associated with re-revision. Methods. Patient demographic details, BMI, American Society of Anesthesiologists (ASA) grade, indication for revision, surgical approach, surgeon grade, implant type (fixed and rotating), time of revision from primary implantation, and re-revision if undertaken were obtained from the National Joint Registry data for England, Wales, Northern Ireland, and the Isle of Man over an 18-year period (2003 to 2021). Results. There were 3,855 patient episodes analyzed with a median age of 73 years (interquartile range (IQR) 66 to 80), and the majority were female (n = 2,480, 64.3%). The median time to revision from primary knee arthroplasty was 1,219 days (IQR 579 to 2,422). Younger age (p < 0.001), decreasing ASA grade (p < 0.001), and indications for revision of sepsis (p < 0.001), unexplained pain (p < 0.001), non-polyethylene wear (p < 0.001), and malalignment (p < 0.001) were all associated with an earlier time to revision from primary implantation. The median follow-up was 4.56 years (range 0.00 to 17.52), during which there were 410 re-revisions. The overall unadjusted probability of re-revision for all revision HKAs at one, five, and ten years after surgery were 2.7% (95% confidence interval (CI) 2.2 to 3.3), 10.7% (95% CI 9.6 to 11.9), and 16.2% (95% CI 14.5 to 17.9), respectively. Male sex (p < 0.001), younger age (p < 0.001), revision for septic indications (p < 0.001) or implant fracture (p = 0.010), a fixed hinge (p < 0.001), or surgery performed by a non-consultant grade (p = 0.023) were independently associated with an increased risk of re-revision. Conclusion. There were several factors associated with time to first revision. The re-revision rate was 16.2% at ten years; however, the risk factors associated with an increased risk of re-revision could be used to counsel patients regarding their outcome. Cite this article: Bone Joint J 2023;105-B(1):47–55

Unicompartmental knee arthroplasty (UKA) has advantages over total knee arthroplasty but national joint registries report a significantly higher revision rate for UKA. As a result, most surgeons are highly selective, offering UKA only to a small proportion (up to 5%) of patients requiring arthroplasty of the knee, and consequently performing few each year. However, surgeons with large UKA practices have the lowest rates of revision. The overall size of the practice is often beyond the surgeon’s control, therefore case volume may only be increased by broadening the indications for surgery, and offering UKA to a greater proportion of patients requiring arthroplasty of the knee. . The aim of this study was to determine the optimal UKA usage (defined as the percentage of knee arthroplasty practice comprised by UKA) to minimise the rate of revision in a sample of 41 986 records from the for National Joint Registry for England and Wales (NJR). UKA usage has a complex, non-linear relationship with the rate of revision. Acceptable results are achieved with the use of 20% or more. Optimal results are achieved with usage between 40% and 60%. Surgeons with the lowest usage (up to 5%) have the highest rates of revision. With optimal usage, using the most commonly used implant, five-year survival is 96% (95% confidence interval (CI) 94.9 to 96.0), compared with 90% (95% CI 88.4 to 91.6) with low usage (5%) previously considered ideal. . The rate of revision of UKA is highest with low usage, implying the use of narrow, and perhaps inappropriate, indications. The widespread use of broad indications, using appropriate implants, would give patients the advantages of UKA, without the high rate of revision. Cite this article: Bone Joint J 2015;97-B:1506–11

Aims

The mean age of patients undergoing total knee arthroplasty (TKA) has reduced with time. Younger patients have increased expectations following TKA. Aseptic loosening of the tibial component is the most common cause of failure of TKA in the UK. Interest in cementless TKA has re-emerged due to its encouraging results in the younger patient population. We review a large series of tantalum trabecular metal cementless implants in patients who are at the highest risk of revision surgery.

Aims

The aim of this study was to investigate whether wear and backside deformation of polyethylene (PE) tibial inserts may influence the cement cover of tibial trays of explanted total knee arthroplasties (TKAs).

Abstract. Introduction. Isolated patellofemoral osteoarthritis accounts for 10% of knee osteoarthritis. Many of these will not require arthroplasty solutions, but for those who are sufficiently symptomatic, patellofemoral joint (PFJ) replacement has been shown to be an effective procedure. The National Joint Registry (NJR) has shown a higher revision rate for this operation, particularly in younger patients (males <55 years 13.3% failure at 5 years, females 9.6%). The aim of this study is to report on the medium-term outcome of the Avon patellofemoral joint arthroplasty in patients under 55 from a non-design centre. There is no other published case series on this young patient cohort. Methodology. 50 Avon PFJ replacements (Stryker, Kalamazoo USA) were undertaken in 46 patients under 55 years old (range 35 – 54, mean 48.8) between 2010 and 2022 for end-stage isolated PFJ arthritis shown on Xray and MRI scan. The outcome measure was all-cause revision rate. This was assessed by review of clinical notes, imaging and NJR data. Results. The mean follow up was 5.8 years (range 6 months to 12.9 years). Only one patient had a revision procedure (for progressive osteoarthritis) which was 3 years after the primary procedure. This patient has had no further surgery. The implant survival rate was 97.2% at 5 years and 97.2% at 10 years. Conclusion. This study shows that patellofemoral joint replacement with the Avon prosthesis can give a satisfactory revision rate in the medium term in patients under 55 if carefully selected

Abstract. Introduction. Aseptic loosening is the most common cause of failure following cemented total knee arthroplasty (TKA) and this has been linked to poor cementation technique. We aimed to develop a consensus on the optimal technique for component cementation in TKA. Methodology. A UK based, three round, online modified Delphi Expert Consensus Study was completed focussing on cementation technique in TKA. Experts were identified as having a minimum of 5 years Consultant experience in the NHS and fulfilling any one of the following three criteria:. · A ‘high volume’ knee arthroplasty practice (>150 TKA per annum) as identified from the National joint registry (NJR). · A senior author of at least 5 peer reviewed articles related to TKA in the previous 5 years. · A named trainer for a post CCT fellowship in TKA. Results. Eighty-one experts (Round 1) and eighty experts (Round 2 and 3) completed the Delphi Study. Four domains with a total of twenty-four statements were identified. 100% consensus was reached within the cement preparation, pressurisation, and cement curing domains. 90% consensus was reached within the cement application domain. Consensus was not reached with only one statement regarding the handling of cement during initial application to the tibial and/or femoral bone surfaces. Conclusion. The CeTIKS expert consensus study presents comprehensive recommendations on the optimal technique for component cementation in TKA. Expert opinion has a place in the hierarchy of evidence and until better evidence is available these recommendations should be considered when cementing a TKA

Aims. The aim of this study was to evaluate medium-term outcomes and complications of the S-ROM NOILES Rotating Hinge Knee System (DePuy, USA) in revision total knee arthroplasty (rTKA) at a tertiary unit. Methods. A retrospective consecutive study of all patients who underwent a rTKA using this implant from January 2005 to December 2018. Outcome measures included reoperations, revision for any cause, complications, and survivorship. Patients and implant survivorship data were identified through both local hospital electronic databases and linked data from the National Joint Registry/NHS Personal Demographic Service. Kaplan-Meier survival analysis was used at ten years. Results. A total of 89 consecutive patients (89 knees) were included with 47 females (52.8%) and a median age of 74 years (interquartile range 66 to 79). The main indications were aseptic loosening with instability (39.4%; n = 35) and infection (37.1%; n = 33) with the majority of patients managed through two-stage approach. The mean follow-up was 7.4 years (2 to 16). The overall rate of reoperation, for any cause, was 10.1% (n = 9) with a rate of implant revision of 6.7% (n = 6). Only two cases required surgery for patellofemoral complications. Kaplan-Meier implant-survivorship analysis was 93.3% at ten years, using revision for any cause as an endpoint. Conclusion. This implant achieved high ten-year survivorship with a low complication rate, particularly patellofemoral complications. These can be avoided by ensuring central patella tracking and appropriate tension of the patellofemoral joint in this posterior hinge design. Cite this article: Bone Jt Open 2022;3(3):205–210

Abstract. Introduction. In 2020, the National Hip Fracture Database (NHFD) was extended to capture data from patients with periprosthetic femoral fractures (PPFF) with plans to include these patients in Best Practice Tarif. We aimed to describe the epidemiology of PPFF in England and Wales, with a particular focus on fractures occurring around the femoral component of knee prostheses. Methodology. This population-based observational cohort study utilised open-access data available from the NHFD. Patients aged over 60, admitted to an acute hospital in England or Wales with a PPFF, within the period 1st January 2020 to 31st December 2020 were included. The primary outcome was the incidence of PPFF in England and Wales. The secondary outcome was the treatment received. Results. We identified 2606 patients with PPFF from 135 hospitals. Of these, a total of 578 fractures occurred around the femoral component of a knee implant. These were classified as Vancouver A (epicondylar, n=77), B (involving implant/cement, n=166) and C (proximal to implant/cement, n=335). Internal fixation was the most employed treatment, used in 352 cases. Revision arthroplasty was performed in 80 cases, and 100 were managed non-operatively. Only 28% of operated PPFF went to theatre within 36 hours but nearly 90% had orthogeriatrician review within 72 hours. Conclusion. Eighty six percent of patients with PPFF were treated with non-revision surgery and would not be recorded in the National Joint Registry. In response, we support calls for the prioritisation of further research into the prevention and management of PPFF around the knee

Abstract. Background. Since 2012 we have routinely used the cementless Oxford medial unicompartmental knee arthroplasty (mUKA), with microplasty instrumentation, in patients with anteromedial osteoarthritis (AMOA) meeting modern indications. We report the 10-year survival of 1000 mUKA with minimum 4-year follow-up. Methods. National Joint Registry (NJR) surgeon reports were interrogated for each senior author to identify the first 1,000 mUKAs performed for osteoarthritis. A minimum of 4 years follow-up was required. There was no loss to follow-up. The NJR status of each knee was established. For each mUKA revision the indication and mechanism of failure was determined using local patient records. The 10-year implant survival was calculated using life-table analysis. Results. The 1,000 mUKA cohort represented 55% of all primary knee replacements in the period, with an average age of 67.7 years and a 54%/46% male/female split. There were 17 revisions (11 for arthritis progression, 4 infections, 1 dislocation and 1 aseptic loosening). The 10-year survival was 98% (44 at risk in 10th year). One patient sustained a periprosthetic fracture at 3 weeks, treated with buttress plate fixation. Discussion. This is the first detailed series reporting the long-term outcome of the cementless Oxford mUKA implanted using microplasty instrumentation. There was a low failure rate, with only one revision for aseptic loosening. Lateral progression was the commonest cause for revision, with an incidence of 1%. This report provides evidence that the combination of evidence-based indications, well-designed instrumentation and cementless fixation can provide excellent long-term survival for the Oxford mUKA in treating AMOA

Abstract. Aim. End-stage arthropathy is a well-known complication of haemophilia, with recurrent haemarthroses leading to joint destruction, deformity, pain, and stiffness. In the knee, this is often treated with total knee arthroplasty (TKA), which can be more challenging in patients with haemophilia (PwH) and associated with poorer outcomes. We conducted a systematic literature review and meta-analysis to determine implant survivorship, functional outcomes and complication rates. Method. A systematic review was conducted using MEDLINE, EMBASE, and PubMed for studies reporting TKA outcomes with Kaplan-Meier survivorship in PwH (PROSPERO registered). Meta-analysis was performed for survivorship and outcomes, and the results were compared to outcomes from the National Joint Registry (NJR). Results. 19 studies, totalling 1187 TKAs (average age 39 years) were reviewed. In PwH, implant survivorship at 5, 10, and 15 years was 94%, 86%, and 76% respectively, whereas NJR reported survivorship for males <55 years was 94%, 90%, and 86%. Survivorship generally improved over the time period studied (1973–2017), but was inversely correlated with HIV infection (common in PwH). Range of motion improved by 10–20° post-operatively, and there were large improvements in Patient Reported Outcome Measures (PROMS). The prosthetic joint infection rate (PJI) was 6% compared to 0.5-1% in non-PwH, but the reporting of other complications, especially haematological, was inconsistent. Conclusions. TKA in PwH has similar 5-year survivorship to non-PwH, but a six-fold higher infection rate. There were marked improvements in range of motion and PROMS, but complications were poorly reported. There remains a need for larger, long-term studies with standardised reporting

Abstract. Introduction. Our aim was to investigate trends in the incidence rate and main indication for revision knee replacement (rKR) over the past 15 years in the UK. Methodology. Cross-sectional study from 2006 - 2020 using data from the National Joint Registry (NJR). Crude incidence rates were calculated using population statistics from the Office for National Statistics. Results. Annual total counts of rKR increased from 2743 procedures in 2006 to 6819 procedures in 2019 (149% increase). The incidence rate of rKR increased from 6.3 per 100,000 adults in 2006 (95% CI 6.1 to 6.5) to 14 per 100,000 adults in 2019 (95% CI 14 to 14). Annual increases in the incidence rate of rKR became smaller over the study period. The incidence of rKR was highest in patients aged 70–79 years (50 per 100,000 adults [95% CI 48 to 52]). Aseptic loosening was the most frequent indication for rKR overall (20.5% procedures). However, rKR for aseptic loosening peaked in 2012 and subsequently decreased. rKR for infection increased incrementally over the study period to become the most frequent indication for rKR in 2019 (2.7 per 100,000 adults [95% CI 2.6 to 2.9]). Infection accounted for 17.2% first linked rKR, 36.7% second linked rKR and 50.7% third or more linked rKR. Conclusion. Recent trends suggest slowing of the rate of increase in the incidence of rKR. Infection is now the most common indication for rKR, following recent decreases in rKR for aseptic loosening. Infection was prevalent in re-revision KR procedures

Aims. Debate remains whether the patella should be resurfaced during total knee replacement (TKR). For non-resurfaced TKRs, we estimated what the revision rate would have been if the patella had been resurfaced, and examined the risk of re-revision following secondary patellar resurfacing. Methods. A retrospective observational study of the National Joint Registry (NJR) was performed. All primary TKRs for osteoarthritis alone performed between 1 April 2003 and 31 December 2016 were eligible (n = 842,072). Patellar resurfacing during TKR was performed in 36% (n = 305,844). The primary outcome was all-cause revision surgery. Secondary outcomes were the number of excess all-cause revisions associated with using TKRs without (versus with) patellar resurfacing, and the risk of re-revision after secondary patellar resurfacing. Results. The cumulative risk of all-cause revision at ten years was higher (p < 0.001) in primary TKRs without patellar resurfacing (3.54% (95% confidence interval (CI) 3.47 to 3.62)) compared to those with resurfacing (3.00% (95% CI 2.91 to 3.11)). Using flexible parametric survival modelling, we estimated one ‘excess’ revision per 189 cases performed where the patella was not resurfaced by ten years (equivalent to 2,842 excess revisions in our cohort). The risk of all-cause re-revision following secondary patellar resurfacing was 4.6 times higher than the risk of revision after primary TKR with patellar resurfacing (at five years from secondary patellar resurfacing, 8.8% vs 1.9%). Conclusion. Performing TKR without patellar resurfacing was associated with an increased risk of revision. Secondary patellar resurfacing led to a high risk of re-revision. This represents a potential substantial healthcare burden that should be considered when forming treatment guidelines and commissioning services. Cite this article: Bone Joint J 2021;103-B(5):864–871