www.njrcentre.org.uk www.njrcentre.org.uk
National Joint Registry 5th Annual Report for England and Wales
ISSN1753-9374 Prepared by
The NJR Centre, Hemel Hempstead
David Emsley
John Martin
Claire Newell
Dr Martin Pickford
Martin Royall
Mike Swanson
M&M Communications
Royal College of Surgeons Clinical Effectiveness Unit
Jan van der Meulen
Susan Charman
The Department of Health
The NJR Steering Committee
Professor Paul Gregg
Mr Martyn Porter
Mr Keith Tucker
NJR Regional Clinical Coordinators’ Network
Mr Peter Howard
This document is available in PDF format for download from the NJR website. www.njrcentre.org.uk National Joint Registry 5th Annual Report for England and Wales
Contents
Chairman’s Introduction 08 Vice-Chairman’s Foreword 10 Executive Summary 11
Part 1 1. Introduction 16 1.1 Annual Report 17 1.2 The National Joint Registry 17
2. Progress 19 2.1 Challenges 20 2.2 Progress 20 2.3 Key Figures 25 2.4 Developments 31 2.5 NJR Stakeholders, Governance and Support 32 2.6 Information, Communication and Monitoring 33
3. Finance 34 3.1 Income and Expenditure 35
4. Taking The NJR Forward 36 4.1 NJR Service Developments 37 4.2 The Future 37 4.3 Information to be published on the NJR website 37 4.4 List of Appendices to Part 1 37 4.5 NJR Centre contact details 37
Appendices 38 Appendix 1 NJR Steering Committee 2007/08 39 Appendix 2 NJR Steering Committee Terms of Reference 41 Appendix 3 NJR Regional Clinical Coordinators 2007/08 42 Appendix 4 Regional Clinical Coordinators’ Network Terms of Reference 43
National Joint Registry 3 Figures Figure 1 NJR Compliance: 2003 – 2008, based on levies from implant sales Figure 2 NJR Consent: quarterly analysis of total records received and those with patient consent, 2003/04 – 2007/08 Figure 3 NJR Linkability: analysis of total records received and those for which NHS numbers have been traced, 2003/04 – 2007/08 Figure 4 NJR Linkability: analysis of actual linkability rate to new linkability rate achieved through linkage to Hospital Episode Statistics (HES) Figure 5 Total hip and knee joint replacement operations entered on the NJR, 2003/04 to 2007/08, recorded by country in which the operation took place Figure 6 Type: hip and knee joint replacement operations entered on the NJR, 2003/04 to 2007/08, recorded by type of operation Figure 7 Proportion of reported operations by provider type Figure 8 Proportion of hip replacement operations by provider type Figure 9 Proportion of knee replacement operations by provider type Part 2
1. Introduction 44
2. Compliance and Linkage to the HES Database 46 2.1 Compliance for Procedures Undertaken in 2007 47 2.2 Linkage of NJR Procedures to HES records 47
3. Overview of Hip and Knee Replacement Procedures 53 3.1 Hospitals and Treatment Centres Participating in the NJR 54 3.2 Hip Replacement Procedures 56 3.3 Knee Replacement Procedures 58
4. Hip Replacement Procedures 61 4.1 Hip Replacement Procedures, 2007 62 4.2 Outcomes Following Primary Hip Replacement, 2003 – 2007 74
5. Knee Replacement Procedures 93 5.1 Knee Replacement Procedures, 2007 94 5.2 Outcomes Following Primary Knee Replacement, 2003 – 2007 95
4 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Tables Table 2.1 Hip or knee joint replacement procedures carried out in England between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode, by provider type Table 2.2 Comparison of patient, procedure and provider type for HES/Patient Episode Database Wales (PEDW)-linked and non-HES/PEDW-linked hip replacement procedures, 1 st April 2003 to 30 th September 2007 Table 2.3 Comparison of patient, procedure and provider type for HES/PEDW-linked and non-HES-linked knee replacement procedures, 1 st April 2003 to 30 th September 2007 Table 3.1 Total number of hospitals and Treatment Centres in England and Wales able to participate in the NJR and proportion actually participating in 2007 Table 3.2 Number of participating hospitals according to number of procedures performed in 2007 Table 3.3 Patient characteristics and procedure details according to type of provider for hip procedures in 2007 Table 3.4 Patient characteristics and procedure details according to type of provider for knee procedures in 2007 Table 4.1 Patient characteristics for primary hip replacement procedures in 2007, according to type of procedure Table 4.2 Age and gender for primary hip replacement in 2007 Table 4.3 Characteristics of surgical practice for primary hip replacement procedures in 2007, according to type of procedure Table 4.4 Thromboprophylaxis regime for primary hip replacement patients prescribed at time of operation, according to procedure Table 4.5 Reported untoward intra-operative events for primary hip replacement patients in 2007, according to procedure type Table 4.6 Patient characteristics for hip revision procedures in 2007, according to procedure type Table 4.7 Indication for surgery for hip revision procedures in 2005 to 2007 Table 4.8 Components removed during hip revision procedures in 2007 Table 4.9 Components used during single stage hip revision procedures in 2007 Table 4.10 Types of hip re-operation other than revisions in 2007 Table 4.11 Characteristics of patients undergoing hip re-operations other than revision Table 4.12 Revision rate at 3 years for primary hip replacement procedures, 1 st April 2003 to 30 th September 2007 that were linked to a HES episode Table 4.13 Hip revision rates (95% confidence intervals) in other national registries Table 4.14 Revision rate at 3 years according to stem brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.15 Revision rate at 3 years according to cup brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.16 Revision rate at 3 years according to patient characteristics and provider for patients who received a primary hip resurfacing between 1st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.17 Reasons for revision according to primary prosthesis type for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007. These results are based on linkage of primary replacements and revisions within the NJR only Table 4.18 Revision rate at 3 years according to resurfacing cup brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.19 Revision rate at 3 years according to bearing surface for primary hip replacement procedures (excluding all resurfacing procedures) undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.20 Revision rate at 3 years according to bearing surface and age for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.21 Revision rate at 3 years according to bearing surface and patient physical status for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode
National Joint Registry 5 Table 4.22 Revision rate at 3 years according to head size for patients who received a primary hip replacement with metal on polyethylene articulation undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.23 Mortality within 3 years according to patient characteristics, prosthesis type and provider for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 4.24 Patient characteristics according to thromboprophylaxis Table 4.25 Patient characteristics according to chemical thromboprophylaxis Table 4.26 Pulmonary embolism and mortality at 3 months according to method of thromboprophylaxis for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 5.1 Patient characteristics for primary knee replacement procedures in 2007, according to type of procedure Table 5.2 Age and gender for primary knee replacements in 2007 Table 5.3 Characteristics of surgical practice for primary knee replacement procedures in 2007, according to type of procedure Table 5.4 Thromboprophylaxis characteristics for primary knee replacement patients prescribed Table 5.5 Reported untoward intra-operative events for primary knee replacement patients in 2007 according to procedure type Table 5.6 Patient characteristics for knee revision procedures in 2007, according to procedure type Table 5.7 Combinations of implants removed during Single and Stage 1 of 2 Knee Revision procedures in 2007 Table 5.8 Revision rate at 3 years for primary knee (excluding patello femoral) replacement procedures, 1 st April 2003 and 30 th September 2007 and linked to a HES episode Table 5.9 Knee revision rates (95% confidence intervals) in other national registries Table 5.10 Revision rate at 3 years according to brands for primary total condylar knee replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 5.11 Mortality within 3 years according to patient characteristics, prosthesis type and provider for patients who received a primary knee replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Table 5.12 Patient characteristics according to thromboprophylaxis Table 5.13 Patient characteristics according to chemical thromboprophylaxis Table 5.14 Pulmonary embolism and mortality rates at 3 months according to method of thromboprophylaxis for patients who received a primary knee replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode
6 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Figures Figure 2.1 Flow chart illustrating linkage of the NJR with HES and PEDW Figure 3.1 Number of participating hospitals by number of procedures, 2004 – 2007 Figure 3.2 Primary hip procedures by provider type Figure 3.3 Primary knee procedures by provider type Figure 4.1 Type of primary hip replacement procedure undertaken between 2005 and 2007 Figure 4.2 Age and gender for primary hip replacement patients in 2007 Figure 4.3 Implant survival of HES-linked primary hip replacement prosthesis according to prosthesis type, 2003 – 2007 Figure 4.4 Implant survival to revision of HES-linked primary hip replacement prosthesis according to age and prosthesis type, 2003 – 2007 Figure 4.5 Implant survival to revision of HES-linked primary hip replacement prosthesis according to gender, 2003 – 2007 Figure 4.6 Implant survival to revision of HES-linked primary hip resurfacings according to age and gender Figure 4.7 Implant survival to revision of HES-linked primary conventional total hip replacements according to age and bearing surface Figure 4.8 Implant survival to revision of HES-linked primary conventional total hip replacements according to patient physical status and bearing surface Figure 5.1 Type of primary knee replacement procedure undertaken between 2005 and 2007 Figure 5.2 Age and gender for primary knee replacement patients in 2007 Figure 5.3 Implant survival of HES-linked primary knee replacement prosthesis according to prosthesis type, 2003 – 2007 Figure 5.4 Survival to revision of HES-linked primary knee replacement prosthesis according to age and prosthesis type, 2003 – 2007 Figure 5.5 Implant survival to revision of HES-linked primary knee replacement prosthesis according to gender and prosthesis type, 2003 – 2007
Glossary 113
National Joint Registry 7 Chairman’s Introduction Bill Darling C.B.E. D.L. F.R.Pharm.S
It is my pleasure as Chairman of the National Joint • the re-design and development of the bulk upload Registry (NJR) Steering Committee to introduce our facility to accommodate Minimum Data Set Version 5th Annual Report, the format of which differs from its 3, and to enable access to the NJR component predecessors. The report will focus on the following database by hospital computer systems, thus ensuring three topics. that the complete record can be uploaded in a single transfer avoiding the need for manual data input 1. An examination of current practice and outcomes • the re-design and re-development of the NJR for hip replacement procedures, including an website in order to bring it up to date and make it update on the analysis of hip resurfacings published easier for stakeholders to find information relevant in the 4 th Annual Report and an analysis of the to their needs. NJR StatsOnline has also been outcomes of the use of different bearing re-developed so that it has the same design as combinations. the website, and will improve the timeliness of 2. An analysis of current practice and outcomes for information provided total and unicondylar knee replacement procedures. • the development of a methodology to identify 3. An analysis of mortality and pulmonary embolism potential outlier prostheses, outlier surgeons and following hip and knee replacement surgery in outlier units, and a protocol to be followed when relation to thromboprophylaxis. a possible outlier scenario is identified. This is a The data for hip and knee procedures which took particularly sensitive subject, and I would like to place between 1 st January 2007 and 31 st December record my thanks to the President of the British 2007 will also be available on the NJR website at Orthopaedic Association (BOA) for his support www.njrcentre.org.uk, as will the detailed analysis of and collaboration in taking this forward. all procedures entered on the Register since data In the light of justifiable concerns about missing discs, collection began in 2003. the NJR Steering Committee assured itself that the During the year under review significant progress was NJR contractors’ policy on data protection was made. At the end of March 2008, the total number of entirely satisfactory, and that all reasonable steps are procedures recorded on the Register was 583,724, taken in relation to its storage and transfer between and compliance had reached 95%, patient consent authorised parties. Personal details such as name, 84% and linkability 83%. date of birth and NHS number are, as we all know, held in the NJR database. I can confirm that this As a result of the increased volume of data, it is now information is encrypted and held in an unreadable possible to make more useful and accurate analysis format. available to all stakeholders and therefore contribute significantly to good practice. Key achievements As I said at the British Hip Society conference in during the year included: February, I believe that the NJR is now entering a new phase of its work. Increasingly, hospitals are using • the introduction of Minimum Data Set Version 3 NJR data for clinical audit purposes and for reviewing which improved the value of the Register their own hospital practice against the national considerably, and simplified data submission further statistics, demonstrating that the Register is by adding the option of ‘not known, not available’ to a tool for excellence to enable improvement in best the three fields of ‘knee flexion’, ‘component brand practice and the quality of patient care. removal’ and ‘location of primary hospital/primary operation date’
8 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Currently the NJR Steering Committee is actively considering the following objectives:
• methods of enhancing data feedback to clinicians • improving data quality • making data collection mandatory with a consequent improvement in accuracy • extending the Register to Northern Ireland • the use of NJR data to facilitate research and audit.
Once again I record my thanks to: Professor Paul Gregg, my Vice Chairman, for his support and commitment during the year; to the other members of the NJR Steering Committee for their valuable contribution; Elaine Young, NJR Project Manager, who worked tirelessly to move the project forward; the Regional Clinical Coordinators who gave so much of their time and brought so much enthusiasm to the project and, last but not least, the NJR contractor, Northgate Information Solutions (UK) Ltd, and all those who have contributed to make this another successful year.
Bill Darling Chairman, NJR Steering Committee
National Joint Registry 9 Vice-Chairman’s Foreword Professor P J Gregg
As Vice-Chairman of the NJR Steering Committee, relation to this, it is now extremely important that much I am pleased to introduce our 5 th Annual Report. emphasis is placed on the assessment of data quality and, where required, its improvement. It will also I believe that further significant progress with the become very important to develop a workable system development of the NJR has been made during the for the assessment of case complexity. year, as outlined in the Chairman’s Introduction. It is not necessary for me to repeat these achievements I am particularly pleased that the NJR has been able to here. establish a working relationship with the Department of Health’s National Patient Reported Outcome However, while reading the 5 th Annual Report, I have Measurement Study (PROMS). This should allow us to reflected on the development of the Registry since its carry out further PROMS studies through the NJR, by inception in April 2003. I am well aware that there are linking pre and post-operative Oxford Hip and Knee those who believe that progress with the development scores to large numbers of patients undergoing of the Registry has been slower than had been hoped treatment with different prostheses and with differing for. However, taking into account the following points, surgical practices and providers. I believe we really have made significant progress: Once again I wish to record my thanks to Bill Darling, • the extremely large number of orthopaedic units, Chairman of the NJR Steering Committee, for all his both NHS and private, which needed to be brought hard work, both within and outside the Committee, on board in England and Wales particularly during these difficult times. In addition I • the huge number of total joint replacements and would also like to sincerely thank the other members revisions being performed in these units of the NJR Steering Committee; Elaine Young, NJR • the large number of stakeholder organisations which Project Manager; the Regional Clinical Coordinators we are required to consult with during development and Regional Coordinators and Northgate Information • the difficulties experienced at local Trust level with Solutions (UK) Limited for all their hard work and the provision of adequate resources to support the valuable contributions. input of data to the registry • the lack of mandatory status for the registration of Finally, thank you to all the orthopaedic surgeons for joint replacements entering their data. Hopefully, the increased feedback • the obvious sensitivities, particularly for orthopaedic to surgeons, which is now coming on stream, will be surgeons, with the establishment of the NJR. found extremely useful and encourage them to continue to enter their data! Speaking personally, I believe the arrangements for the identification and management of potential outlier prostheses and surgeons is both fair and practical and fulfils the requirements for present day Clinical Governance. It is now even more important that we work towards registration of total joint replacements becoming mandatory so that all surgeons and hospitals are subject to the same scrutiny. Also, in Professor P J Gregg Vice-Chairman, NJR Steering Committee
10 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Executive Summary
This, the National Joint Registry (NJR) 5 th Annual Progress during the year 2007/08 included: Report, is the formal, public report of the Registry for the period 1 st April 2007 to 31 st March 2008. The • recording 151,496 hip and knee joint replacement report also includes detailed analyses of the data held operations, representing 95% 1 of the total carried by the Registry. out in England and Wales in both the NHS and independent healthcare sector. This is the highest The NJR began collecting data in April 2003 with the ever annual total and brings the total number of aim of improving care for all patients undergoing hip procedures recorded by the NJR, as at 31 st March and knee replacement surgery. This is achieved 2008, to 583,724 through the collection of data that is used to monitor • continuing to improve the number of records and measure the long term effectiveness of implants submitted to the NJR with patient consent to 86% used in hip and knee joint replacement surgery and by the end of Quarter 4 2008 the procedures involved. • increasing the number of the records submitted with both patient consent and the patient’s NHS number The work of the NJR in England and Wales is from 69% on 31 st March 2007, as recorded in the 4 th overseen by the NJR Steering Committee, working Annual Report, to 83% by 31 st March 2008 alongside the Department of Health as an advisory, • increasing the overall quality of the data held in the non-departmental public body. database by linking NJR records to data held by the Hospital Episode Statistics (HES) service to find In January 2008 it was announced that the patient NHS numbers. responsibility for the NJR would transfer, on 1 st April 2008, from the Department of Health to the Healthcare Data quality has been improved by: Quality Improvement Partnership (HQIP) as part of the new host arrangements for the National Clinical Audit • continuing to maintain direct, on-site support and and Patient Outcomes Programme. training for individual units and the development of a benchmarking tool to compare the performance Part 1 of units with regard to the submission of data to the NJR Part 1 of the report focuses on the performance and • the implementation of a new dataset on 1 st April progress of the Registry for the financial year 2007/08, 2007 highlighting key developments. It provides an account • obtaining permission to use a shortened form of of the work undertaken by the NJR, its funding and consent for inclusion on local hospital consent forms governance arrangements, planned developments, • increasing awareness of the NJR through and the composition and terms of reference for both attendance at conferences and exhibitions, the NJR Steering Committee and the NJR Regional publishing a quarterly newsletter (Joint Approach), Clinical Coordinators’ Network. and organising and hosting regional events for unit staff responsible for NJR.
1. Number of operations recorded on the NJR compared to the sales of implants between 1 st April 2007 and 31 st March 2008 and reported on in July 2008. Many procedures carried out in the last quarter of the financial year are not reported to the NJR until the first quarter of the new financial year.
National Joint Registry 11 Key developments and achievements in the year • the re-development and re-launch of the NJR 2007/08 included: website to ensure that NJR stakeholders can find relevant content more easily • publishing the 4 th Annual Report which provided a • the provision of support into a study on metal on detailed analysis of the data held by the NJR and metal hip articulations. included, for the first time, survivorship analysis for certain types of procedure Consideration is also being given to: • continuing to develop links to data held by HES to improve the type of analysis of hip and knee joint • extending the NJR to include ankle, elbow, shoulder, surgery that can be undertaken and wrist joint replacement surgery • the launch of version 3 of the Minimum Dataset • extending the NJR to Northern Ireland. (MDSv3) accompanied by changes to the application which have improved data quality and reduced the Part 2 amount of time needed to enter operation details • the development of a statistical methodology for the Part 2 of the NJR 5 th Annual Report summarises the identification of higher than expected revision rates. data for hip and knee procedures carried out between Agreement on how to investigate higher than 1st January 2007 and 31 st December 2007 in England expected failure rates have been agreed by the NJR and Wales, and entered into the NJR by 29 th February Steering Committee following consultation with the 2008. It also includes a description of outcomes after British Orthopaedic Association (BOA), the hip and knee replacements that were entered into the Medicines and Healthcare products Regulatory NJR since the start of data collection in April 2003. Authority (MHRA), and the suppliers of orthopaedic Revision rates, mortality and pulmonary embolism are implants also included. The analysis of these outcomes is • providing data to the British Hip Society as part of based on linkage of procedures entered into the NJR a Parliamentary Sub-Committee review into the use with records from HES and the Patient Episode of chemical thromboprophylaxis following surgery Database Wales (PEDW). This is the first time that • working with the MHRA, units, and the orthopaedic PEDW data has been used. suppliers to identify quickly those patients who had received implants which were subsequently the NJR Compliance in 2007 subject of a device alert or product recall, significantly reducing the period of time between • The total number of levies collected in England and the issuing of the notice and clinical review Wales from 1 st January to 31 st December 2007 was • improving the Bulk Upload process in order to make 164,053 and the total number of hip and knee it easier for units to submit data in bulk from other procedures reported to the NJR was 141,394. computer systems and avoid the need for entering This suggests that the compliance rate for 86% for the same data into different systems. procedures undertaken during the 2007 calendar year. 2 This is an improvement over previous years Future, additional developments include: with compliance at 79% in 2006.
• the development of a NJR Clinician Feedback Overview of Hip and Knee Procedures system that will enable surgeons to assess their performance against defined comparators at • In 2007, 409 hospitals and treatment centres were hospital, sector (NHS or independent healthcare), open and, of these, 383 (94%) submitted data to the and national levels NJR. On average, 181 hip replacements and 190 knee replacements were recorded per participating unit.
2. This figure is different to that quoted earlier (95%) because it relates to the 2007 calendar year, not the 2007/08 financial year.
12 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
• The NJR recorded 68,950 hip procedures, of which 29 were used in revision procedures. The total of 65 9.7% were revisions or re-operations, and 72,480 brands recorded by the NJR and available for use, knee replacement procedures, of which 5.6% were represents a 20% increase from 2006. either revision or re-operation procedures undertaken between 1 st January and 31 st December Outcome of hip and knee procedures 2007. carried out between 1 st April 2003 and 30 st • Of all the hip procedures, 66% were carried out in September 2007 NHS hospitals, 24% in independent hospitals, 5% in NHS treatment centres, and 5% in independent • Revision rate, mortality, and pulmonary embolism sector treatment centres. For all knee operations, rate analyses were based on procedures in the NJR 67% were undertaken in NHS hospitals, 21% in linked to data in HES. Focus has been placed on hip independent hospitals, 5% in NHS treatment resurfacing, unicondylar knee replacement and centres, and 7% in independent sector treatment outcomes of various hip bearing surface centres (ISTCs). combinations. The HES database records only those • 93% of patients at independent hospitals and ISTCs, procedures funded by the NHS in England and compared with 82% at NHS units, were graded as fit includes NHS patients being treated in independent and healthy or with mild disease, according to the sector hospitals. ASA grade system. 98% of procedures undertaken • The overall revision rate following primary hip at ISTCs were primary procedures. Revision replacement was 0.6% (95% CI 0.5% – 0.6%) at 1 procedures undertaken at NHS hospitals account for year and 1.2% (95% CI 1.1% – 1.3%) at 3 years. 81% of all revision procedures. The 3 year revision rates were lowest in patients who • The trend towards more cementless total hip received cemented prosthesis (0.7%, 95% CI 0.6% replacement procedures has continued, with – 0.8%) and highest after hip resurfacing (2.8%, 95% cementless procedures now accounting for 33% of CI 2.3% – 3.5%). Overall revision rates were lower in all total hip replacements. The reduction in the women (1.1%) than in men (1.4%). However, women proportion of cemented procedures has also have higher revisions rates for resurfacing (3.6%) but continued, down from 48% in 2006 to 43% in 2007. lower revision rates for all other prostheses. Hip resurfacing accounts for 8% of total hip • The calculated hazard ratios suggest that prosthesis replacements. type is the most influential factor on implant survival • In 2007, 129 different brands of acetabular cups and following primary hip replacement. Hip resurfacing 144 different brands of femoral stems were recorded procedures have statistically significantly higher in the NJR. The total number of acetabular cups and revision rates than cemented total hip procedures femoral stems recorded in the NJR database is 165 after adjusting for age, sex, ASA grade and provider and 187, respectively. type. • 20% mixed and matched cup-stem combinations • Data on bearing surfaces was available from 99,397 were used, a decrease of 2% from 2006. primary hip replacement procedures. The 3 year • The types of primary knee replacements procedures revision rates for the four most frequently entered have, largely, remained the same, with 83% being articulation combinations (excluding all hip total knee replacements and 8% being unicondylar resurfacing procedures and implants with heads knee replacements. larger than 32mm) show that the highest revision • 48 brands of condylar knee prostheses were rates are for ceramic on polyethylene and ceramic recorded in the NJR, along with 15 brands of on ceramic (p<0.001). Multivariable analysis unicondylar prostheses, 9 brands of patello-femoral adjusting for age, sex, physical status and prosthesis replacement prostheses and 14 brands of hinged type did not confirm this observation (p=0.9). prostheses. Of the 48 brands of condylar knee • The overall revision rate following primary knee prostheses, all were used in primary procedures and replacement was 0.3% (95%CI 0.3% – 0.4%) at
National Joint Registry 13 1 year and 1.2% (95%CI 1.1% – 1.3%) at 3 years after • Of the 55,210 patients prescribed LMWH after hip surgery. The 3 year revision rates were lowest in replacement surgery, the 3 month pulmonary patients who receive a cemented prosthesis (1.2%, embolism rate was 0.6% (95%CI 0.5% – 0.7%). For 95%CI 1.1% – 1.3%) and highest in those who the 16,662 patients prescribed aspirin, the received a unicondylar knee replacement (2.0%, 95%CI corresponding rate was 0.5% (95%CI 0.4% – 0.6%) 1.5% – 2.7%). Overall, revision rates were lower in (p=0.01). The multivariable analysis suggests that the women 1.1% 95% CI 1.0% – 1.3%) than in men (1.3% pulmonary embolism rate with LMWH is about 1.2 95% CI 1.2% – 1.5%) (p=0.003). The statistical (95%CI 0.9 – 1.6) times that for those patients evidence for an interaction between prosthesis type prescribed aspirin. However, the effect was no longer and sex is weak (p=0.3). statistically significant (p=0.1). • Of the 3,377 knee revision procedures reported, 74% • For patients prescribed LMWH, the 3 month mortality were single stage procedures, 11% were stage 1 of a rate following hip replacement surgery was 0.6% 2 stage process, and 14% were stage 2 of a 2 stage (95%CI 0.5% – 0.7%) with the corresponding rate for process. The remainder (less than 1%) were aspirin being 0.7% (95%CI 0.6% – 0.9%). After conversions to arthrodesis and amputations. applying the multivariable analysis, the hazard ratio • The multivariable analysis to investigate the comparing LMWH to aspirin is 0.8 (95%CI 0.6 – 0.9) association between prosthesis type and revision in (p=0.01). This indicates a statistically significantly the first three years after surgery, adjusted for age, lower mortality rate in those that receive LMWH, but sex, physical status and provider type, showed that should be interpreted with caution. prosthesis type seems to be an important factor • The 3 month pulmonary embolism rate after knee determining revision rates. The risk of revision in the surgery was 0.4% (95%CI 0.4% – 0.5%) in the first 3 years after surgery was approximately 1.4 times 58,871 patients prescribed LMWH, and 0.5% (95%CI higher in patients who received a unicondylar knee 0.4% – 0.6%) in the 21,343 patients prescribed replacement than in patients who received a aspirin. The multivariable analysis suggests that the cemented prosthesis. pulmonary embolism rate with LMWH is about 0.9 • The mortality rate for all 102,179 first primary hip (95%CI 0.7 – 1.1) times that with aspirin (p=0.3). This replacements carried out between 1 st April 2003 and is not statistically significant. 30 th September 2007 that could be linked to a HES • The 3 month mortality rate following knee surgery was record was 0.7% (95%CI 0.6% – 0.7%) at 3 months 0.6% (95%CI 0.5% – 0.6%) in patients prescribed after surgery and 6.2% (95%CI 5.9% – 6.4%) at 3 LMWH, with the corresponding rate for those patients years after surgery. The mortality rate for all 111,723 prescribed aspirin 0.5% (95%CI 0.4% – 0.6%). The first primary knee replacements (excluding patello- multivariable analysis suggests that the mortality rate femoral replacements) for the same period was 0.5% for those prescribed LMWH is about 1.2 (95%CI 1.0 – (95%CI 0.5% – 0.6%) at 3 months after surgery and 1.5 ) times higher than in those prescribed with 6.1% (95%CI 6.0% – 6.4%) at 3 years after surgery. aspirin (p=0.08). This is not statistically significant. • The mortality rates at 3 months following hip and Mortality and Pulmonary Embolism after knee replacement for patients receiving mechanical Hip and Knee Replacement prophylaxis only were 0.7% (95% CI 0.6 – 0.9%) and 0.4% (95% CI 0.4% – 0.6%) respectively. For Mortality and pulmonary embolism rates (for both fatal Pulmonary Embolism at 3 months the rates were and non-fatal events) after hip and knee replacement 0.7% (95% CI 0.5% – 0.8%) and 0.4% (95% CI 0.4% were examined more closely for patients receiving – 0.5%) respectively. aspirin and low molecular weight heparin (LMWH). The • In conclusion, there were very small differences in crude rates were then subjected to a multivariable mortality according to the chemical analysis adjusting for age, sex, physical status, and the thromboprophylaxis treatment that patients received. use of mechanical prophylaxis. Patients receiving aspirin had statistically significantly higher mortality after hip surgery and lower mortality
14 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
after knee surgery than those receiving LMWH. These differences should be interpreted with caution, as they are in opposite directions for hip and knee replacements. Also, the risk adjustment that could be applied is likely to be incomplete and therefore ‘confounding by indication’ cannot be excluded as a contributing factor.
National Joint Registry 15 part 1
1. introduction
16 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
1.1 Annual Report Data from the Hospital Episodes Statistics database (HES) has been used for some of the analyses. This is the fifth annual report of the National Joint Registry (NJR), which provides information about hip 1.2 The National Joint Registry and knee replacement operations in England and Wales within both the NHS and independent The NJR was established in October 2002 and began healthcare sector. This information is valuable to collecting and studying data on hip and knee surgeons, patients, members of the public and the replacement operations in April 2003. manufacturers of hip and knee implants (artificial joints). The NJR collects information that can be used The NJR was set up to improve patient care by finding to help improve care and treatment by providing out more about hip and knee joint replacement information on surgical practice and on the hip and implants and surgery. It is doing this by building up a knee implants used in joint replacement surgery. database of information. The aim is for the Registry to be able to provide information for patients, surgeons, The report has been divided into two parts for ease of hospitals, manufacturers and healthcare agencies on reference: implant performance, joint replacement surgery and best surgical practice. • Part One – a general outline of the work of the NJR • Part Two – an analysis of the operations recorded by It is essential that, if the NJR is to achieve its aims, it the Registry receives a timely and continuous supply of high quality information about all operations undertaken. This Some of the appendices, tables and charts that have information makes it possible to monitor the been included in previous reports will this year be performance and long-term effectiveness of hip and published only on the NJR website. The information to knee replacement surgery and the implants (also be found on the website is detailed in Part 1 Section known as prostheses) used. Information from the NJR 4.3. is made available to surgeons, patients, the public and the manufacturers of replacement joints to help • Part One is a review of progress made by the NJR improve care and treatment in the future. It is also during the financial year 1 st April 2007 to 31 st March used to assist in monitoring failure rates. This is 2008 and plans to increase still further the value of explained more fully in Section 2.4. information in the future. • Part Two is an analysis of the data recorded for the Data is provided to the NJR by NHS and independent calendar year, 1 st January to 31 st December 2007 healthcare providers throughout England and Wales. st and, using information recorded on NJR since April By 31 March 2008 the NJR had received 583,724 2003, focuses on three main topics: operation records. Although other international – an analysis of the current practice and outcomes arthroplasty registers have existed for longer than the for hip replacement surgery, including hip NJR, the NJR now holds more records about certain resurfacing and the use of different bearing procedure types, for example, hip re-surfacing, than combinations any other registry in the world. – an analysis of the current practice and outcomes for total and unicondylar knee replacement surgery The quality of the data received is very important to – mortality and pulmonary embolism following hip the NJR as it affects the type of information that can and knee replacement surgery in relation to be provided to stakeholders (a list of key NJR thromboprophylaxis. stakeholders is available on the NJR website and included at Appendix 1 to Part 1 of this report).
National Joint Registry 17 5th Annual Report National Joint Registry for England and Wales
Management and funding
The NJR has been managed by Northgate Information Solutions (UK) Ltd since April 2006, under a contract with the Department of Health, and is funded through a levy raised on the sale of hip and knee replacement implants. Further information regarding funding is available in Section 3.
In January 2008 it was announced that responsibility for the NJR was to transfer from the Department of Health to the Healthcare Quality Improvement Partnership (HQIP), a consortium comprising the Academy of Medical Royal Colleges, the Royal College of Nursing and the Long Term Conditions Alliance, as part of the new host arrangements for the National Clinical Audit and Patient Outcomes Programme (NCAPOP), which had been expanded to incorporate the NJR. This change was effective on 1 st April 2008.
The NJR’s Steering Committee (NJRSC), which was set up in October 2002, is an advisory non-departmental public body. A list of current NJRSC members and their declarations of interest is published on the NJR website.
18 National Joint Registry part 1
2. progress
National Joint Registry 19 2.1 Challenges In addition to addressing these targets, the work of the NJR Centre (NJRC) during 2007/08 also focused on:
The success of the NJR continues to be reliant on • improving the data entry application to make data meeting three main challenges: entry easier and to ensure the submission of more complete and accurate records • improving the completeness and quality of the • continuing to share best practice amongst units information held on the Registry through the use of regional workshops and providing • reducing the burden of data entry on hospitals and targeted, on-site support for hospital staff units supplying the information • improving the quality of the data held by the NJR • making information held on NJR available for better and the type of analyses being undertaken. This has patient care. been achieved by linking patient records held in the NJR to corresponding records in a major NHS The continuing challenges on data completeness and system, the Hospital Episode Statistics service quality are to ensure that: (HES )4 • improving the provision of information through the • all hip and knee joint replacement operations carried complete re-design and re-build of the NJR website. out in England and Wales are recorded on the NJR This will ensure that information is structured to meet • the recorded information is of good quality the needs of individual stakeholder groups and can • patient consent is given for data to be stored and be easily accessed used to achieve the aims of the NJR. • undertaking a project that will enable units to transfer data from other hospital computer systems, The last year has seen continued success in meeting reducing the need for entering the same data into these challenges, aided by changes to the way in multiple systems and improving the quality of the which consent is recorded and a change to the data data submitted to the NJR collected by the NJR. These are discussed in more • improving performance monitoring. detail in Section 2.4 below. These initiatives will continue throughout the next year There are three ways of measuring these areas of data in order to ensure that the quality of the data continues quality and completeness which has provided a main to increase. The linkage to HES will be expanded in its focus for the work of the NJR Centre during 2007/08: scope and linkage to its Welsh equivalent, Patient Episode Database Wales (PEDW), will also be • compliance – the percentage of records submitted undertaken. The major focus for the year, however, will to the NJR compared to the total number of hip and be the provision of information to surgeons that will knee replacement operations carried out in England enable them to compare the success of their work to and Wales, also referred to as ‘case ascertainment’ that of other surgeons at a sector level (NHS and – NJR target set at 95% by June 2008. independent health sector) and national level. • consent – the percentage of records submitted to the NJR with consent given by patients for use of 2.2 Progress their personal information 3 – NJR target set at 90% by June 2008. Progress against the three measures of compliance, consent and linkability for the financial year 2007/2008 • linkability – the percentage of records that have is detailed in this section. patient consent and the patient’s NHS number so that all operations performed on the same patient can be linked. The goals of the NJR will only be Compliance achieved if operations on the same patient can be successfully linked. – NJR target set at 90% by All NHS units in England and Wales are expected to June 2008. submit details of all hip and knee replacement
3. Personal information includes NHS number, surname, date of birth and postcode. 4. HES records information about all patient activity in NHS hospitals in England.
20 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
operations that they carry out. For independent Figure 1 (below) illustrates the compliance rate (case healthcare sector units and Independent Sector ascertainment) across all units over the last five years Treatment Centres (ISTCs), the submission of by comparing the number of operations recorded on operations to the NJR is mandatory, and included in the NJR to the number of implants sold. Apart from a their contracts with the Department of Health. The slight decrease during 2006/07, the compliance rate method used to monitor compliance with the NJR is to has increased from approximately 40% in the first year compare the number of levies raised from the sale of of the NJR to 95% in 2007/08. implants to the number of submissions from units. Compliance of NHS units can also be measured by The following hospitals failed to submit any data to the comparison to the number of operations reported to NJR in the financial year from 1 st April 2007 to 31 st HES and PEDW. March 2008:
Trust Hospital Bolton Hospitals NHS Trust The Royal Bolton Hospital Bromley Hospitals NHS Trust Orpington Treatment Centre Bromley Hospitals NHS Trust Princess Royal University Hospital Cambridge University Hospitals NHS Foundation Trust Addenbrooke’s Hospital Central Manchester & Manchester Children’s University Manchester Royal Infirmary Hospitals NHS Trust Guy’s & St Thomas’ NHS Foundation Trust Guy’s Hospital Guy’s & St Thomas’ NHS Foundation Trust Guy’s Nuffield House Homerton University Hospital NHS Foundation Trust Homerton University Hospital Orthopaedic Spine & Specialist Hospital Orthopaedic Spine & Specialist Hospital (Independent) Ramsay Healthcare Bromley Private Patient Unit (Independent) University College London Hospitals NHS Foundation Trust University College Hospital
180,000 100% Figure 1 160,000 90% NJR Compliance: 2003 – 140,000 80% 2008, based on levies from
) implant sales
70% % 120,000 ( e s t
d Source: Operations entered a r 60%
R st o on NJR 1 April 2003 – c 100,000 e e
c st R 50% n 31 March 2008 and levy a i R l submissions to NJR by
J 80,000 p N 40% m implant suppliers and o
60,000 C manufacturers. 30% 40,000 20%
20,000 10% NJR Records
0 0% Levy 2003/04 2004/05 2005/06 2006/07 2007/08 Compliance (%) Year
National Joint Registry 21 Consent Recording a patient’s personal details, including their NHS number, enables the NJR to link operations for Why is it important for patients to give consent? that patient that have either been carried out already, By consenting to the inclusion of their personal details or may be carried out in the future. This will enable on the NJR database, patients enable more accurate problems either with implants or surgery to be study of the outcomes of hip and knee replacement identified at an early stage and the appropriate action operations. Without patient consent, the NJR will not taken. This NJR information will also enable hospitals achieve its aims. Gaining patient consent means that to be identified and assisted when they need to recall the NJR is able to provide information that can be patients who have received an implant that is used to: suspected of failing to perform as expected.
• improve awareness of the outcomes of replacement The rates of patient consent are published monthly on surgery the NJR website in NJR StatsOnline • help patients obtain the best clinical care for joint (www.njrcentre.org.uk). The percentage of records replacement surgery submitted to the NJR with the necessary consent rose • look at the performance of implants and help identify from 78% (37,416 NJR records) in 2005/ 06 Quarter 4 , any brand of implant showing high failure rates to 83% (39,753 NJR records) in 2007/ 08 Quarter 4. • improve surgical practice through the monitoring of clinical performance Figure 2 below shows the steady increase in the • help NHS and other healthcare organisations to recording of patient consent over the five years of the make best use of their resources . NJR.
Patients must give consent for their personal details to be used on the NJR database and, once they understand the purpose of the NJR, they rarely refuse to give consent when asked.
45,000 100% Figure 2 40,000 90% NJR Consent: quarterly 35,000 80% analysis of total records 70% received and those with 30,000 patient consent, 2003/04 – 60% 2007/08 25,000 50% 20,000 Source: Operations entered 40% on NJR 1 st April 2003 – 31 st 15,000 March 2008. 30% 10,000 20% Number of operations 5,000 10% on NJR Number of operations 0 0%
N on NJR with consent Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 u m Consent rate (%) b 2003/04 2004/05 2005/06 2006/07 2007/08 e r o f o p e r a t i o n s o
22 n National Joint Registry N R J a n d n u m b e r w i t h c o n s e n t National Joint Registry 5th Annual Report for England and Wales
The consent rate has continued to rise every quarter “I have read and comprehended the Patient since the second quarter of 2005/06. It reached its Information Leaflet and consent to my personal details highest ever recorded rate of 86% in Quarter 4 of being submitted to the National Joint Registry in the 2007/08, following two quarters of significant knowledge that they will only be disclosed in the public improvement. Progress is still required over the next interest or in other circumstances permitted by law. I year to reach the target consent level of 90%. have been assured and understand that by declining Initiatives with ‘Section 60’ support and new NJR my consent my care and treatment will not be affected consent wording, described below, are expected to in any way. ” improve performance further. Linkability When submitting records to the NJR, there are three options that can be entered for consent; ‘Yes’, ‘No’, Linkability refers to the ability to link all operations ‘Don’t Know’. For 2007/2008, 6% of records were relating to a single patient, for example linking a submitted as ‘No’, whilst 8% were submitted as ‘Don’t revision operation to the patient’s first, or primary, Know’. Very few patients actually decline consent and replacement operation. The ability to link operations is the relatively high numbers of ‘No’ and ‘Don’t Know’ key to monitoring the performance of implants and submissions are generally linked either to the patient surgery ; the NJR will fail in its aims if records cannot not being asked for their consent or the completed be linked . consent form not being available to the person submitting the data. In order to create a link between operations, personal information (normally a patient’s NHS number) is Section 60 Support 5 required. This requires patient consent. Records must include either the patient’s NHS number or their The Patient Information Advisory Group (PIAG) granted surname, date of birth and postcode so that the ‘Section 60’ support to the NJR, allowing the Registry patient’s NHS number can subsequently be traced to record patient identifiable data where ‘Don’t Know’ through the National Strategic Tracing Service (NSTS). was indicated for patient consent. This was sought to improve the linkability of the records submitted to the The percentage of linkable records held on the NJR NJR and implemented on the data entry system on 1 st from 2003/04 to 2007/08 is shown in Figure 3 below. April 2007. The improvements to record linkability can The linkability rate has continued to increase and the be seen in Figure 3. rate of increase between 2006/07 and 2007/08, from 72% to 83% is the largest annual increase to date. The exemption is retrospective, enabling the NJR Further improvement is required over the coming year Centre to begin to improve the linkability of previous to meet the linkability target of 90% and there are years’ records. ongoing initiatives to address this.
The support from Section 60 also permits consented Changes to the data entry application implemented on patient records to be linked to HES and PEDW data to 1st November 2007 made it mandatory to submit improve and extend the types of analyses that can be either the patient’s NHS number or surname, date of undertaken. birth, and postcode. This has increased both the number of records submitted with an NHS number Changes to NJR consent and the success rate of using the NSTS to find a patient’s NHS number. A shortened form of consent, designed for inclusion st on local ‘consent to operate’ forms, was approved by The data entry application was changed on 1 April the NJR Steering Committee and the Department of 2007 to implement the support granted under Section Health in March 2008. Gaining consent for the NJR in 60. This enabled the recording of patients’ personal this way will ensure that that consent forms part of a information for those records submitted with ‘Don’t patient’s notes and will always be available to the Know’ recorded against consent. The linkability rate st th person responsible for submitting the data to the NJR. for the month 31 March to 30 April 2007 increased The shortened text is: by 6.3%, from 74.3% to 80.6%. This represents the largest monthly increase in the five years of the NJR.
5. Although Section 60 of the Health and Social Care Act 2001 has been superseded by Section 251 of the NHS Act 2006, Section 60 is still referred to by the Patient Information Advisory Group (PIAG). The support provides relief from the common law duty of confidentiality which requires that patient identifiable data is only used with consent. Section 251 powers can only be used for medical purposes where it can be shown that the use of identifiable patient data is absolutely necessary. National Joint Registry 23 160,000 90% Figure 3
140,000 80% NJR Linkability: analysis of total records received and 70% 120,000 those for which NHS numbers 60% have been traced, 100,000 2003/04 – 2007/08 50% 80,000 40% Source: Operations entered on NJR 1 st April 2003 – 60,000 30% 31 st March 2008. 40,000 20% 20,000 Number of operations
N 10% u m N Number with NHS b H 0 0%
e number traced S r
n 2003/04 2004/05 2005/06 2006/07 2007/08 o u
f Linkage rate (%) m r L e b i
c Year n e o k r r a d a b s f i t l i e s t y u Thr e support granted under Section 60 has also been The linkage to HES has increased an actual linkability b N R m S a used retrospectively to improve the linkability of the rate of 27% in 2003/04 to 55%. The results mean that T t i t e t S
recoe rds held by the NJR. Consented records have far more detailed analysis of the data can be d S e
beena linked to HES in order to confirm a patient’s undertaken, with greater levels of confidence in its a n r d
idc entity and to determine their NHS number. The accuracy and conclusions. It is planned to undertake h w
resui lts of this work are shown in Figure 4 below. similar work using PEDW data. t h
90 Figure 4 80 NJR Linkability: analysis of 70 actual linkability rate to new linkability rate achieved 60 through linkage to HES 50 Source: Operations entered on 40 NJR 1 st April 2003 – 31 st March 2008. 30 L i n k
a 20 b i l i t y
R 10 a t e Linkability
( 0 %
) 2003/04 2004/05 2005/06 2006/07 2007/08 New Linkability Year
24 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
2.3 Key figures 2.3.2 Operation totals
2.3.1 Information collected Since the launch of the NJR in April 2003, a total of 583,724 hip and knee joint replacement operations in England and Wales has been recorded on the NJR. NJR records include the following information about During the first year, the number of operations each operation: reported was 62,189; the figure for this year is 151,496, which is the highest number of submissions in any • patient details (if patient consent is recorded as single year and represents an increase of 15.75% on either ‘Yes’ or ‘Don’t Know’) the previous year. • hospital • surgeon Figure 5 below shows the total number of operations • operation recorded on NJR in England and Wales each year • surgical approach from 2003/04 to 2007/08. In England, there were a • any untoward event which may have occurred total of 69,425 hip replacement operations and 72,867 during the operation knee replacement operations undertaken. The number • details of the implants used. of hip and knee replacement operations undertaken in Wales was 4,207 and 4,322 respectively. The 4 th Annual Report recorded that, for the first time, in 2006/07 the number of knee replacement operations exceeded the number of hip replacement operations. This trend has continued for 2007/08 with an increased difference between the totals in favour of knee replacement operations.
80,000 Figure 5
70,000 Total hip and knee joint replacement operations entered on the NJR, 2003/04 60,000 to 2007/08, recorded by country in which the 50,000 operation took place
40,000 Source: Operations entered on NJR 1 st April 2003 – 30,000 31 st March 2008. N u m 20,000 b England Hip e r o
f 10,000 Wales Hip o p e
r England Knee a
t 0 i o
n 2003/04 2004/05 2005/06 2006/07 2007/08 Wales Knee s Operations by year and country
National Joint Registry 25 Total operations • re-operation other than revision – an operation following either a primary or revision operation that 2003/04 62,189 does not require any joint implants to be removed 2004/05 107,172 or replaced, for example, if an implant needs to be re-aligned or has become loose .6 2005/06 132,577 2006/07 131,378 Primary operations continue to represent the most 2007/08 151,496 reported procedures (93%) with the difference in the number of knee primary operations and hip primary Source: Operations entered on NJR 1 st April 2003 – operations continuing to increase in favour of the 31 st March 2008. former (2.5% in 2006/07 and 4.4% in 2007/08).
Operation types Figure 6 (below) shows the number of operations Hip and knee joint replacement operations take three reported by type from 1 st April 2003 to 31 st March forms: 2008.
• primary – the first time a joint is replaced • revision – an operation that involves the removal and replacement of one or more components of a joint replacement
80,000 Figure 6 70,000 Type: hip and knee joint 60,000 replacement operations 50,000 entered on the NJR, 40,000 2003/04 to 2007/08, recorded by type of 30,000 operation
N 20,000 u
m Source: Operations entered b
e 10,000 on NJR 1 st April 2003 – r
o st f 0 31 March 2008. o p
e 2003/04 2004/05 2005/06 2006/07 2007/08 r a t i Hip Po rimary 30,036 49,168 59,817 58,746 66,839 n Hip Primary s Knee Primary 27,911 49,477 62,423 61,919 73,455 Knee Primary
Revision Hip 3,012 5,168 6,211 6,223 6,757 Revision Hip
Revision Knee 1,232 2,297 3,040 3,312 3,719 Revision Knee
Hip Re-operation 245 340 306 284 Hip Re-operation
Knee Re-operation 185 333 375 442 Knee Re-operation
Operations by year and joint/type
6. Re-operation information was not collected on the first version of the Minimum Dataset (MDSv1) from 1 st April 2003 to 31 st March 2004. It was included on MDSv2 from 1 st April 2004 and removed from MDSv3 which came into use on 1 st December 2007. The figures are included for completeness only.
26 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Total operations: Where did the operations take place? Of the 583,724 operations submitted to the NJR, 2003/04 95.4% were carried out in England and 4.6% in Wales. Hip: 32,935 In 2007/08, 142,292 (94.3%) operations were carried Knee: 29,049 out in England, compared to 8,529 (5.7%) in Wales.
2004/05 There are four types of organisations in England Hip: 54,799 carrying out hip and knee joint replacement surgery: Knee: 51,643 • NHS hospital 2005/06 • NHS treatment centre Hip: 66,138 • independent healthcare hospital Knee: 65,567 • Independent Sector Treatment Centre (ISTC).
2006/07 There are no NHS or treatment centres or ISTCs in Hip: 65,067 Wales. Knee: 65,312 Of the 583,724 operations submitted to the NJR, 2007/08 380,223 (65.1%) were performed in NHS hospitals in Hip: 73,632 England and Wales; 25,646 (4.4%) in NHS treatment Knee: 77,189 centres; 157,881 (27.1%) in independent sector healthcare hospitals in England and Wales; and Source: Operations entered on 19,970 (3.4%) in ISTCs in England. NJR 1 st April 2003 – 31 st March 2008.
National Joint Registry 27 100% Figure 7 90%
80% Proportion of reported operations by provider type 70% Source: Operations entered 60% on NJR 1 st April 2003 – st 50% 31 March 2008.
40%
30% ISTCs
20% NHS Treatment Centres
P 10% Independent Hospitals r o p
o NHS Hospitals
r 0% t i o 2003/04 2004/05 2005/06 2006/07 2007/08 n b o y f Year h p i r p o v a i n d d e r k t
n 2003/04 2004/05 2005/06 2006/07 2007/08 y e p e NHeS Hospitals 65.4% 64.7% 61.5% 66.2% 66.4% o
p 40,671 69,174 81,279 86,644 100,555 e r a t
Indepi endent Hospitals 33.7% 30.6% 29.9& 23.2% 22.6% o
n 20,971 32,784 39,544 30,376 34,206 s NHS Treatment Centres 0.9% 3.4% 5.8% 6.0% 5.2% 551 3,674 7,605 7,852 7,868
ISTCs 0.0% 1.3% 2.8% 4.6% 5.9% 1,358 3,735 6,011 8,866
Total 62,193 106,990 132,163 130,883 151,495
Source: Operations entered on NJR 1 st April 2003 – 31 st March 2008.
The number of hip and knee replacement operations The proportion of operations undertaken at undertaken in ISTCs has continued to increase from independent sector hospitals has continued to drop 1.3% in 2004/05 to 5.9% in 2007/08. By the end of (from 33.7% in 2003/04 to 22.6%) although the rate 2007/08, ISTCs had reported 19,970 operations of decrease has slowed down in the last year. The (3.4% of the total). The number of operations significant drop between 2005/06 and 2006/07 (6.7%) performed in NHS Treatment Centres has dropped was due to the ending of the second round of General slightly in 2007/08 but the total number of operations Supplementary Procurement (GSUPP2) in that year. reported to the NJR is 25,646 (4.4% of the total).
28 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
100% Figure 8 90%
80% Proportion of hip replacement operations by 70% provider type 60% Source: Operations entered 50% on NJR 1 st April 2003 – 31 st March 2008. 40%
30% ISTCs
20% NHS Treatment Centres
10% Independent Hospitals P r o NHS Hospitals p 0% o r
t 2003/04 2004/05 2005/06 2006/07 2007/08 i o n
o Year b f y h i p p r r o e v p i d l
a 2003/04 2004/05 2005/06 2006/07 2007/08 e c r e t m NHy S Hospitals 63.4% 63.6% 61.0% 65.1% 65.7% p e e
n 20,961 34,983 40,469 42,487 48,537 t o
Indepp endent Hospitals 35.8% 32.1% 31.0% 25.2% 24.0% e
r 11,817 17,684 20,595 16,456 17,751 a t i o
NHS Tn reatment Centres 0.8% 3.2% 5.4% 5.5% 5.0% s 269 1,787 3,585 3,587 3,720
ISTCs 0.0% 1.3% 2.8% 4.6% 5.9% 577 1,719 2,747 3,871
Total 33,047 55,031 66,368 65,277 73,879
Source: Operations entered on NJR 1 st April 2003 – 31 st March 2008.
The proportion of hip joint replacement operations (3,871 operations) in 2007/08. The increase in the recorded at NHS hospitals during this period number of operations in NHS hospitals and ISTCs has increased by 4.7% to 65.7% (48,537 operations). The been accompanied by a 7% decrease in the proportion of hip replacement operations undertaken proportion of operations in independent sector by ISTCs has doubled from 2.6% in 2006/07 to 5.2% hospitals.
National Joint Registry 29 100% Figure 9 90%
80% Proportion of knee replacement operations by 70% provider type
60% Source: Operations entered 50% on NJR 1 st April 2003 – 31 st March 2008. 40%
30% ISTCs
20% NHS Treatment Centres
10% Independent Hospitals P r
o NHS Hospitals
p 0% o
r 2003/04 2004/05 2005/06 2006/07 2007/08 t i o n Year o f b k y n e p e r o r v e i
p 2003/04 2004/05 2005/06 2006/07 2007/08 d l e a r c t
NHS He ospitals 67.6% 65.8% 62.0% 67.3% 67.0% y m p 19,706 34,191 40,810 44,157 52,018 e e n Indept endent Hospitals 31.4% 29.1% 28.8& 21.2% 21.2% o
p 9,154 15,100 18,949 13,920 16,455 e r a t
NHS Ti reatment Centres 1.0% 3.6% 6.1% 6.5% 5.3% o
n 282 1,887 4,020 4,265 4,148 s ISTCs 0.0% 1.5% 3.1% 5.0% 5.9% 781 2,016 3,264 4,995
Total 29,142 51,959 65,795 65,606 77,616
Source: Operations entered on NJR 1 st April 2003 – 31 st March 2008.
The proportion of knee replacement procedures 2006/07 to 67% (52,018 operations) in 2007/08. carried out in ISTCs has doubled from 3.1% in These increases have been accompanied by a drop 2006/07 to 6.4% (4,995 operations) in 2007/08. The in the proportion of knee replacement procedures proportion of knee replacement operations carried out performed in independent sector hospitals from 28.8% in NHS hospitals has also increased from 62% in in 2006/07 to 21.2% (16,455 operations) in 2007/08.
30 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
2.4 Developments where they have been implanted, the NJR Centre can quickly inform units of affected patients, enabling a swift clinical review to be undertaken. Using NJR data Linkage to the Hospital Episode Statistics in this way can shorten the process from notification to (HES) clinical review by up to six weeks, a significant improvement in patient safety. The HES database contains information about all in- patient, out-patient, and accident and emergency Identification of potential ‘outliers’ ‘episodes’ that a patient has. HES is essentially an administrative system but, by linking consented patient During the year the NJR has developed a statistical records held in NJR to a patient’s corresponding methodology for identifying rates of revision surgery record in HES, the type of information that can be fed that are higher than would be expected. The back to stakeholders is significantly improved. methodology provides an indication that closer scrutiny of the data might be required. Working with A major HES linkage project was undertaken for the the British Orthopaedic Association (BOA ), the NJR NJR’s 4 th Annual Report which enabled an analysis of Steering Committee has agreed a process by which to the length of stay in hospital for different procedure investigate higher than expected failure rates where and patient types, in addition to an analysis of mortality clinical issues appear to predominate. A similar and the use of chemical and physical thrombo- process has also been agreed with the MHRA and the prophylaxis. The linkage work also significantly orthopaedics industry where performance increased the number of linked revisions held in the issues appear to be related to a particular prosthesis, NJR and enabled, for the first time, the production of which will be dealt with under existing arrangements survivorship curves for certain procedure types and for medical device vigilance. Documents detailing the major brands of implants. The linkage exercise statistical method being used and the process by enabled this work to be undertaken much earlier than which higher than expected failure rates are expected. investigated will be communicated via the BOA and made available on the NJR website. These will be Later in the year, the NJR carried out another linkage communicated to stakeholders via the BOA exercise, looking at the incidence of deep vein and Association of British Healthcare Industries (ABHI ). thrombosis and pulmonary embolism following hip replacement surgery. This linked data was then used A new version of the NJR minimum by the British Hip Society when presenting to a dataset Parliamentary Sub-Committee review into the use of chemical thromboprophylaxis which resulted in a Following a review of the dataset collected by the reconsideration of the National Institute for Health and NJR, the Steering Committee approved the adoption Clinical Excellence ( NICE ) guidelines. of the Minimum Dataset Version 3 (MDSv3). The Review of Central Returns (ROCR) subsequently Patient safety: support to the regulatory approved the new dataset in May 2007. MDSv3 is authorities shorter than the previous dataset with fewer selectable options for most fields. The NJR is the single source of information about Implementing MDSv3 required significant changes to where a particular batch or lot of implants has been the web-based data entry application and the project used, and in which patients they have been implanted. was extended in scope to improve the user’s The NJR now has a collaborative working experience of using the application. The new interface arrangement with the Medicines and Healthcare went live on the 1 st December 2007 and has resulted products Regulatory Authority (MHRA) with regard to in the time taken to enter records being reduced. As at device alerts and product recalls. By using data held in 31 st March 2008 more than 75% of submissions were the NJR, the NJR Centre can quickly determine if a made using the new dataset. The deployment of group of implants subject to an alert or recall have MDSv3 has also been accompanied by higher than been used, where they have been used and in whom ever compliance and linkability rates. they have been implanted. Whilst the MHRA and manufacturers are informed of the numbers used, and
National Joint Registry 31 Bulk upload 2.5 NJR Stakeholders,
Many units also have specialist orthopaedic computer Governance, and Support applications which collect the same data as required by the NJR, and it is recognised that the requirement The NJR is involved with a large and diverse number to enter the same data into more than one system is of stakeholders, all of whom benefit from the Registry not an economic use of resources. Enabling units to and its work. A comprehensive list of these transfer data electronically from their own systems to stakeholders can be found on the NJR website. the NJR will ensure maximum use of resources, thereby encouraging compliance, and improve data quality by Steering Committee minimising the risk of erroneous input by an operator. The Steering Committee met four times during Following a review of the current Bulk Upload 2007/08 and the minutes of its meetings can be capability, a proposal to improve it was approved by found on the NJR website. The Committee’s current the NJR Steering Committee in November 2007, and members were appointed by the Appointments work started on the project in January 2008. The Commission on behalf of the Secretary of State for improved capability will be designed to support the Health following a formal recruitment process. For a new minimum data set, MDSv3, and will enable users list of NJR Steering Committee members and their of other specialist systems to access key component declarations of interest, please visit the NJR website. information directly from the NJR database. This will improve the accuracy of information being entered into Regional Clinical Coordinators (RCC) third party systems and increase the number of records that can be transferred directly to the NJR. The NJR Regional Clinical Coordinators’ Network The enhanced capability will also improve a user’s consists of 23 consultant orthopaedic surgeons, experience by removing much of the technical acting as local ‘champions’ for the service and complexity of the process and providing clear reports supporting the work of the NJR Steering Committee, about the progress and success of the data transfer and the Regional Coordinators. The Chair and Vice process. The NJRC is working closely with the Chair are appointed following interview, with the Chair vendors of specialist orthopaedic and patient having ‘observer’ status at NJR Steering Committee administration systems in the development of the new meetings. Further information about the RCC Network facility. and its members can be found on the NJR website.
NJR website Regional Coordinators
In November 2007 the NJR Steering Committee The NJR Centre has six Regional Coordinators (RCs), approved the complete re-development of the NJR each providing direct, on-site support to approximately website. The purpose of the project was to ensure that 70 units. Further information about the role of the RCs stakeholders could easily find and access information is available on the NJR website. appropriate to their needs and to make management of the website less dependent upon technical staff. At Priorities the same time, all the information on the previous website was reviewed and removed, or amended as Work during 2007/08 continued to focus on ensuring appropriate. New content will be continuously added. that performance against the key targets for compliance , consent and linkability improved. Accompanying this Members of the NJR Steering Committee and work were a number of initiatives to determine Regional Clinical Coordinators’ Network were actively stakeholder requirements and to develop the routines involved in the website redevelopment project. The and methods for providing information to stakeholders new website went live in April 2008. once the data was of sufficient quantity and quality.
32 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
2.6 Information, communication Publications and monitoring • National Joint Registry 4 th Annual Report. • Three issues of the NJR newsletter ‘Joint Approach’. Keeping stakeholders informed • A revised starter pack of information explaining the NJR and providing guidance for best practice for It is essential that stakeholders benefit from the data new unit staff was developed and more than 85 that is provided to the NJR and that information is distributed. made widely available. The NJR now has a • 60,000 patient information leaflets printed and considerable volume of good quality and unique data distributed. that will enable stakeholders to deliver real • Updated user manuals for users of the data entry improvements to patient care. As the NJR moves into application. the next stage of its development, utilisation of the • Article in British Orthopaedic News in April 2007. quantity and quality of the data base information it has established since 2003, to support other studies, audit Conferences and exhibitions and research, will be a key priority. Information is provided to stakeholders as outlined here. Representatives of the NJR Centre attended conferences for the British Orthopaedic Association, NJR ReportsOnline the British Hip Society, and the British Association for Surgery of the Knee. They also presented at a number NJR ReportsOnline is a capability that enables of other symposiums and workshops, including the surgeons and hospital data managers to view reports BUPA National Administrator’s workshop, the Special about numbers of procedures, case mix and implant Orthopaedic Alliance meeting at Nuffield Oxford, the usage. It is accessed via the NJR data entry application. BUPA National Theatre Manager’s Day, Biomet Sponsored training days and a study period held by Spire. Performance monitoring In an effort to share best practice among units, to The performance management system, implemented update units on the progress of the NJR, and to in 2006, continues to provide management reports to understand unit issues and requirements, the NJRC stakeholders and to accurately target support to hospitals. instigated a series of regional workshops. This year, they were held at Sheffield Northern General Hospital, NJR StatsOnline for the South and West Yorkshire regions, and Northwick Park Hospital for the London SHAs. Workshops for the East of England, Thames Valley, NJR StatsOnline is accessed through the NJR website and Wales are planned for the forthcoming year. and provides basic information to patients and hospitals about the number of procedures undertaken NJR staff also contributed to the Royal London Hospital at a hospital , trust, and strategic health authority (SHA) Surgeon’s Training Day and the Regional Registrars’ Day level . at Sheffield Northern General Hospital. Staff have also been invited to attend a number of clinical audit sessions Information Bureau which have provided excellent opportunities to present to both senior clinicians and management about the The NJR receives many requests for information benefits and the need for the NJR. These have included throughout the year and operates an Information presentations at Good Hope Hospital in Solihull, Bureau to respond to those requests. Anyone can Solihull Hospital and University College Hospital London. request data from the NJR using the application form on the website. There is a well established process for Training authorising the release of information from the NJR. This is designed to ensure that patient confidentiality is The Regional Coordinators have undertaken 169 visits maintained. The NJR Steering Committee is keen to to hospitals to train new NJR staff. These sessions are make the data available for research and is actively additional to the ‘normal’ support visits and the high considering a number of requests. number reflects the high turnover in hospital staff.
National Joint Registry 33 part 1
3. finance
34 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
3.1 Income and expenditure
The NJR is self financing, funded by a levy raised on the sale of hip and knee implants to NHS and independent healthcare providers in England and Wales. The rate of the levy is set by the Health Minister and is subject to a Memorandum of Understanding between the Department of Health, the National Assembly for Wales, the Independent Healthcare Forum and the Association of British Healthcare Industries Orthopaedic Special Interest Section.
Income to the NJR for financial year 2007/08 was £2,856,823.90 and expenditure for the same period was £2,041,111.69.
Members of the NJRSC and the RCC Network are volunteers and do not receive payment for their work. However, they are reimbursed for any travel and subsistence expense incurred while attending meetings. The total expenditure for members’ expenses during 2007/08 was £20,942.91
The levy was set at £23 per joint from June 2006 and reduced to £20 per joint from June 2007 for 2007/08.
National Joint Registry 35 part 1
4. taking the NJR forward
36 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
4.1 NJR Service Developments 4.3 Information to be published on the NJR Website A number of key developments are summarised below. (www.njrcentre.org.uk) • The development of a Clinician Feedback system was proposed and approved. This will enable Website Documents surgeons to compare their performance with that of their colleagues at a hospital, sector (NHS or 1. Annual Report Part 1. independent healthcare), and national level. 2. Annual Report Part 1 – Welsh language. The system will thus provide surgeons with early 3. Annual Report Part 2. indicators of outlier status. This development is key to the continued success of the NJR and is due to The following information will be published on the NJR be launched in September 2008. website only: • Continued linkage work to data held on HES and PEDW to provide improved NJR data quality and 1. the data supporting the figures for Part 1 extended analysis. 2. the Welsh language version of Part 1 of the • Monitoring developments in the National Programme Annual Report for IT (NPfIT) to identify opportunities for improving 3. prostheses used in hip and knee replacement data collection and quality. procedures. • Continued development of the NJR website to meet the current and future needs of stakeholders as the 4.4 List of Appendices to Part 1 publication of NJR data analysis becomes more dynamic. 1. NJR Steering Committee 2007/08. • Improving the ability to link electronically to other 2. NJRSC Steering Committee Terms of Reference. systems in order to exchange data . 3. Regional Clinical Coordinators. 4.2 The future 4. Regional Clinical Coordinators’ Terms of Reference.
Planned developments are summarised below: 4.5 NJR Centre contact details
• the re-development and re-launch of the NJR National Joint Registry website to ensure that NJR stakeholders can find Northgate Information Solutions (UK) Ltd relevant content more easily Peoplebuilding 2 • implementing the data feedback system to clinicians Peoplebuilding Estate (see 4.1) Maylands Avenue • facilitating the MHRA/BOA led study of Metal on Hemel Hempstead Metal hip articulations. Hertfordshire HP2 4NW The following developments are also being considered:
• extending the NJR to include elbows, shoulders, ankles and wrists following requests to do so from the British Elbow and Shoulder Society and the British Orthopaedic Foot and Ankle Society • extending the NJR to Northern Ireland (NI) following consultation.
National Joint Registry 37 part 1
appendices
38 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Appendix 1 NJR Steering Committee 2007 /08
1.1 NJR Steering Committee – Composition
As an advisory, non-departmental public body, the composition of the NJRSC is:
• Chair 1 • Orthopaedic surgeons 3 • Patient representative groups 2 • Implant manufacturer/supplier industry 3 • Public health/epidemiology 1 • NHS organisation management 1 • Independent healthcare provider 1 • Practitioner with special interest in orthopaedic 1 care who is a GP, nurse or allied health professional (physiotherapist or occupational therapist)
1.2 Membership from 1 st October 2007
With members appointed under new constitution as post became vacant:
Mr Michael Borroff Orthopaedic Device Industry Miss Mary Cowern Patient Group (Arthritis Care) (from 1 st October 2006) Mrs Patricia Cassidy Independent Healthcare Sector (from 19 th April 2007) Mr William Darling Mrs Patricia Durkin Representing Patients (from 1 st March 2007) Professor Paul Gregg Orthopaedic Surgeon Vice Chair (from January 2007) Mr Anthony Lowther Orthopaedic device industry (until November 2007 ) Professor Alex MacGregor Public Health and Epidemiology Miss Carolyn Naisby Practitioner with Special Interest in Orthopaedics Mr Martyn Porter Orthopaedic Surgeon Dean Sleigh Orthopaedic Device Member (from April 08) Mr Keith Tucker Orthopaedic Surgeon replaces Mr Tim Wilton (from 1 st May 2007) Mr Tim Wilton British Society for Surgery of the Knee (until 31 st April 2007) Mr Andrew Woodhead NHS Trust Management (from 4 th January 2007)
National Joint Registry 39 1.3 Observers
The following have regularly attended NJRSC meetings as observers:
Mr Peter Howard NJR Regional Clinical Coordinators’ Network Allan Hiddlerley Medicines and Healthcare products Regulatory Agency (MHRA) Christine Miles Welsh Assembly Government Andy Smallwood NHS Purchasing and Supply Agency (PASA), now NHS Supply Chain Elaine Young Healthcare Quality Improvement Partnership (HQIP)
1.4 NJR Steering Committee Members’ Declarations of Interest
Mr Michael Borroff Chair of ABHI Orthopaedics Special Interest Section and currently employed by DePuy International Ltd, manufacturer of orthopaedic prostheses Mrs Patricia Cassidy Chief Executive, Hospitals, Nuffield Health Miss Mary Cowern None Mr William Darling Managing Director J, M and W. Darling Ltd, Managing Director Galen Pharmacy Ltd. Chair South Tyneside Standards Committee Mrs Patricia Durkin CHKS Health Care Consultancy fee paid work and share holder Professor Paul Gregg Consultant Orthopaedic Surgeon at South Tees Hospitals NHS Trust. Unit receives research/audit funding from DePuy International Ltd, Stryker UK and Smith & Nephew plc Orthopaedic advisor for Capio Healthcare Professor Alexander Professor of Chronic Disease Epidemiology, University of East MacGregor Anglia. Consultant Rheumatologist at Norfolk and Norwich University Hospital NHS Trust Miss Carolyn Naisby Consultant Physiotherapist, City Hospitals Sunderland NHS Foundation Trust Mr Martyn Porter Consultant Orthopaedic Surgeon, Wrightington, Wigan and Leigh NHS Trust. Work at a unit that has received financial support from DePuy International for clinical and RSA studies for Elite Plus femoral stem and C-stem
Dean Sleigh Currently employed by Biomet UK, as National Business Development Manager. Has shareholdings in Biomet, Johnson & Johnson and Tyco Has acted as Consultant to DePuy International for development of hip femoral stem (C-stem AMT)
Mr J Keith Tucker Consultant Orthopaedic Surgeon, Norfolk and Norwich University Hospital NHS Trust. Various sources of financial support for research undertaken by orthopaedic department. Paid royalties by Johnson & Johnson Orthopaedic more than 5 years ago for contribution to design of hip prostheses. Royalties paid to orthopaedic charity Mr Andrew Woodhead CEO Royal National Orthopaedic Hospital
40 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Appendix 2 NJR Steering Committee Terms of Reference 7
The National Joint Registry Steering Committee • providing advice about the performance of (NJRSC) is responsible for: prostheses and best surgical practice to ministers, the Department of Health, the National Assembly for • setting the programme of work for the NJR Centre, Wales, providers of orthopaedic services and, in the in conjunction with the Department of Health future, Northern Ireland Department of Health, Social • ensuring that appropriate stakeholders are involved Services and Public Safety in and consulted on the work of the NJR Centre, as • reviewing the NJR levy on an annual basis with the appropriate Department of Health and making recommendations • recommending any significant changes to the to ministers on the amount at which it should be set contract between the contractor responsible for the • providing an Annual Report to ministers on the work NJR Centre and the Department of Health, including of the NJR and, with ministerial agreement, to make consideration of value for money the report available in both English and Welsh • monitoring the codes of conduct applied to the • facilitating, where appropriate, the use of NJR data contractor in its relationship with other key for research purposes stakeholders (orthopaedic units within NHS and • adhering to the terms of the Code of Practice for independent healthcare organisations and with the National Joint Registry Steering Committee Member s 8 orthopaedic implant industry) 9 • monitoring the delivery and quality of the work of the and the Code of Practice – Register of Interests . NJR against the contract in conjunction with the Department of Health • monitoring the inclusion, closure and performance of orthopaedic units contributing data to the NJR
7. For further information about the NJRSC rules, publication scheme, chair and members duties and roles, code of practice for members and declaration of interests, see the NJR Steering Committee Members’ Handbook, available from the NJR website (www.njrcentre.org.uk) under information on the NJR Steering Committee. 8. See Section 4 of the NJR Steering Committee Members’ Handbook. 9. See Section 5 of the NJR Steering Committee Members’ Handbook.
National Joint Registry 41 Appendix 3 NJR Regional Clinical Coordinators 2007/2008
Regional Clinical Coordinators (RCCs) South East Coast Strategic Health Authority Mr Peter Howard – RCC Chair Mr Hagen Jähnich/Helmut Zahn (shared position) Mr Guy Selmon South West Strategic Health Authority Mr Evert Smith East Midlands Strategic Health Authority Mr Nick Fiddian Mr Colin Esler – Vice Chair Mr Peter Howard West Midlands Strategic Health Authority Mr David Dunlop London Strategic Health Authority Mr Ian D M dos Remedios Vacancy (Mr Mark Rowntree retired February 2008) Mr Gareth Scott North West Strategic Health Authority Mr Glyn Thomas Yorkshire and Humberside Strategic Health Mr David Sochart Authority Mr Ian Stockley North East Strategic Health Authority Mr John Anderson Mid and West Wales Professor Andrew McCaskie Mr David Woodnutt
East of England Strategic Health Authority North Wales Mr Matthew Porteous Mr Glynne Andrew Mr Godfrey Charnley South East Wales South Central Strategic Health Authority Vacant (Mr Robin Rice resigned June 2007) Mr John Britton
42 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Appendix 4 Regional Clinical Coordinators’ Network Terms of Reference
The National Joint Registry Regional Clinical Coordinators’ Network members have a responsibility to: • facilitate and improve NJR compliance and patient consent rates and the quality of data submitted to the NJR by individual NHS trusts and private healthcare providers in their region • establish regular interaction with relevant NJR RCs • assist the RCs to encourage and support the poorly performing NHS trusts and private healthcare providers in their region to submit the required NJR data • provide input to determining appropriate reporting, analysis and interpretation of their region’s data and the overall NJR data • facilitate feedback of NJR data to their region’s NHS trusts and private healthcare providers, orthopaedic surgeons, hospital managers and other appropriate stakeholders on a monthly to quarterly basis • host training roadshows and regional seminars to raise awareness of the NJR and feedback information • communicate between the regionally based hospitals / orthopaedic units / treatment centres and the NJRSC, NJR Centre (contractor) and Department of Health to achieve the aims of NJR • provide continued support to the functioning of the NJR • attend the NJR RCC Network meetings on a regular basis and contribute • provide representation at meetings / committees for NJR at regional and national level as required • provide representation on organising committees for NJR-related events. Note: In considering issues at national level, NJR RCC Network meetings will adopt a UK perspective, which will include taking account of differences between England, Wales and Northern Ireland, and of the status of devolved and non-devolved matters.
National Joint Registry 43 part 2
1. introduction
44 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Part 2 of the NJR 5 th Annual Report summarises the Just as in the previous Annual Report, revision rates data for hip and knee procedures carried out between are described according to prosthesis type, including 1st January 2007 and 31 st December 2007 in England hip resurfacing and unicondylar knee replacement. and Wales and entered into the NJR by 29 th February The outcomes after these types of replacements have 2008. been compared against alternative types. Another special area of interest was the occurrence of This part of the Annual Report also includes a pulmonary embolism in the first three months description of outcomes after hip and knee according to the type of thromboprophylaxis replacements that were entered into the NJR since the described at the time of surgery. As a result of the start of data collection in April 2003, including revision increased number of procedures entered into the NJR rates, mortality and pulmonary embolism. as well as the longer follow-up period, these outcomes can now be presented with greater precision. The analysis of these outcomes is based on linkage of the procedures entered into the NJR with records A number of the comparisons described above were from the Hospital Episode Statistics (HES) database carried out using multivariable analyses in order to take and the Patient Episode Database for Wales (PEDW). into account the fact that patients receiving a HES is an administrative database of all NHS and particular type of prosthesis or thromboprophylaxis privately funded admissions in NHS hospitals in might differ from patients receiving other types. England as well as admissions of NHS funded patients A comparison of revision rates after hip and knee in the independent sector. PEDW is the corresponding replacement against results presented in other national database in Wales and this is the first time that this registries will also be provided. data has been used. This is the second time that a linkage between NJR The NJR provides information about the type and and a hospital administrative database has been brand of prosthesis and the surgical procedure. HES undertaken. Not all procedures entered in the NJR and PEDW have the potential to provide post- could be linked. Therefore, care should be taken in operative information that is not available in the NJR. generalising the results of the analyses to all hip and Furthermore, the availability of data on all hospital knee replacement procedures in England, particularly admissions in HES provides near-complete when comparing provider types. A brief description of information about revision surgery and mortality. The compliance and linkage of the NJR with HES is made linked database therefore offers a unique opportunity in Section 2, with the aim of establishing whether the to study outcomes after joint replacement in a way results presented in this Annual Report can be that would not be possible using either database considered as representative of joint replacement alone. practice in England and Wales.
Data from the mortality registry of the Office for Section 3 gives an overview of the numbers of hip and National Statistics is also included in HES. knee procedures carried out in 2007 in England and Unfortunately, no mortality data appeared to be Wales according to the type of provider (e.g. NHS, available in PEDW. Without information about treatment centre and independent sector) and source mortality, it is impossible to calculate true revision of funding. The succeeding two sections first present rates. Therefore only procedures entered into the NJR the characteristics of patients who underwent a joint linked with HES could be used for the analyses of replacement, the type of replacement and surgical outcomes after joint replacement. It is important to technique used for procedures in 2007. Subsequently, note that the results of these analyses refer only to hip outcomes in terms of revision of the prostheses, and knee replacements carried out in England. mortality, and pulmonary embolism are presented for procedures in the NJR-HES linked database. The Revision rates and mortality are presented as a results are presented separately for hip replacement in function of time after surgery. The impact of patient Section 4 and knee replacement in Section 5. The characteristics, the type of prosthesis as well as the types and brands of prostheses that were used in type of provider, are also explored. Furthermore, 2007 are presented in Part 2, Prostheses used in hip revision rates are presented for a number of frequently and knee replacement procedures 2007/08, which used prostheses brands, bearing surfaces and head can be found on the NJR website. sizes.
National Joint Registry 45 part 2
2. compliance and linkage to the hospital episode statistics database
46 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
The level of completeness of joint procedures reported 2.2 Linkage of NJR procedures to to the NJR is an important factor that determines whether the results presented in the Annual Report HES records can be considered to be representative of joint replacement practice in England and Wales. In this At the time of analysis, final data from HES was section, the rate of compliance is presented. available for hospital admissions up to 31 st March Compliance is defined as the proportion of all joint 2007. Provisional HES data from 1 st April 2007 until procedures undertaken in England and Wales in 2007 30 th September 2007 was also available. Therefore, for th that were entered into the NJR by 29 February 2008, this Annual Report, only those procedures undertaken The outcome analyses presented in the Report were between 1 st April 2003 and 30 th September 2007 based on the procedures, entered in the NJR since could be considered for linkage. the start of data collection in April 2003, that were also in the HES and PEDW-linked database. The Furthermore, HES and PEDW data include NHS and characteristics of these procedures are summarised in privately funded admissions in NHS hospitals in this section and compared to those of the procedures England as well as admissions of NHS funded patients that could not be linked to a HES or PEDW episode. in the independent sector. Consequently, linkage was not possible for independently funded procedures 2.1 Compliance for Procedures undertaken in the independent sector. These Undertaken in 2007 procedures were approximately 15% of the total number of procedures entered into the NJR between 1st April 2003 and 30 th September 2007. In this section the rate of compliance is presented. A levy is collected to cover the costs of the NJR at each sale of a specific hip or knee component. A single leviable component is used for most joint procedures. A levy might not be collected in exceptional cases, such as when custom-made implants are used or when a revision operation does not involve removal and replacement of the leviable component. However, a comparison of the number of joint replacements reported to the NJR with the number of levies collected gives a reasonable approximation of the rate of compliance (case ascertainment).
The total number of levies collected in England and Wales from 1 st January to 31 st December 2007 was 164,053 and the total number of hip and knee procedures reported to the NJR was 141,394. This suggests that the compliance was 86% for procedures carried out during the 2007 calendar year. Compared with previous years, compliance rate has improved from 60% in 2004, 77% in 2005 and 79% in 2006.
National Joint Registry 47 Table 2.1 Hip or knee joint replacement procedures carried out in England between 1 st April 2003 and 30 th September 2007 that were linked to a HES/PEDW episode, by provider type % % % % % % % % % % % % % 0 9 4 5 9 0 3 0 8 4 6 4 3 6 4 6 6 6 6 5 6 5 4 4 3 4 - % s W e 2 8 8 8 8 0 9 5 3 7 9 3 8 r s D 2 9 4 8 5 6 4 5 9 9 4 2 7 u e E 2 6 2 2 5 7 7 4 2 8 2 0 9 d , , , , , , , , , , , , p P e / d y 1 3 2 8 5 5 1 5 2 9 5 0 t c e S 6 1 1 2 3 1 3 r k o 1 1 1 E r n n e i p H l d i 8 3 7 2 2 2 9 2 6 9 5 4 3 v 7 6 6 0 6 3 5 1 8 3 5 5 0 o 1 4 0 8 8 9 1 8 2 7 1 2 9 1 r l , , , , , , , , , , , , , p a 2 3 7 4 6 5 5 4 1 8 0 2 2 t l 0 1 1 1 2 2 1 1 7 5 2 l o 1 2 2 1 A T % % % % % % % % % % % % % 4 3 4 9 0 1 6 1 0 7 8 8 0 4 6 2 3 4 1 3 3 2 3 3 1 4 % - s W e 6 2 9 6 8 4 5 6 4 6 9 1 8 s r D 2 9 4 3 1 2 9 8 4 4 2 1 u C E 8 1 8 2 1 7 , , , , d T P e 1 2 3 3 / d t c e S n k o E n r e n i p H l d 0 8 7 5 2 7 2 1 2 2 8 9 5 n 5 5 7 3 6 5 7 3 6 2 5 1 4 e 1 8 2 2 1 8 3 0 1 p 1 , , , , , l e 4 7 8 8 2 a t d o n I T % % % % % % % % % % % % % 3 2 7 1 5 7 0 0 8 6 8 0 3 6 5 6 6 7 6 6 7 4 6 6 6 5 - % s W e 3 5 1 4 8 1 1 2 0 4 9 6 4 r D 8 0 9 8 9 5 1 8 1 2 2 7 4 u E 9 7 5 6 2 1 2 1 2 3 0 3 d , , , , , , P e / d 2 7 1 9 7 3 c e S k o E r n n i H l p s 2 3 0 0 6 9 7 8 3 2 3 0 3 C 4 5 8 0 0 2 7 0 3 7 9 2 8 T 7 3 6 8 3 9 2 3 7 5 5 6 1 l , , , , , , , a 4 1 1 1 0 4 3 S t 1 1 1 H o T N % % % % % % % % % % % % % 7 4 6 8 3 1 9 8 3 0 9 5 9 1 1 1 1 3 2 1 1 1 1 1 s l - % a t s i W e 3 2 3 4 4 7 4 5 6 7 8 5 8 p r D s 1 1 5 4 6 4 5 2 3 0 8 3 9 u E 6 6 3 6 7 9 2 2 3 0 o d , , , , , P h e / d 2 4 4 6 1 c t e S k o n E r n n e i H l p d 0 1 8 4 4 8 1 4 2 6 9 1 2 n 0 3 6 7 2 4 1 8 7 8 4 6 5 e 4 4 5 8 6 5 7 0 6 4 2 0 3 1 p l , , , , , , , , , , e a 5 3 2 2 9 7 2 3 6 1 t d 1 2 3 2 o n T I % % % % % % % % % % % % % 9 6 1 6 3 5 9 1 8 5 4 6 0 6 5 7 7 6 7 6 4 6 4 7 4 5 - s W % e r D u E 0 9 4 3 0 9 7 0 5 6 2 1 8 d 0 8 5 7 3 6 8 9 6 8 9 8 2 P e / d 8 1 5 8 9 4 9 5 8 6 3 3 7 , , , , , , , , , , , , c e S s l 3 7 2 6 7 0 9 4 6 7 4 1 k o E e r n a 5 9 1 1 1 2 i p t H p l 1 1 i n y t p s r 6 1 5 0 5 7 2 7 5 9 1 3 7 o e 8 1 2 4 7 2 8 4 0 0 0 7 2 h d 1 8 7 4 5 2 8 3 7 1 7 9 0 7 i l , , , , , , , , , , , , , v S a 7 0 3 0 8 1 1 1 1 9 0 6 2 t o H 7 1 3 1 6 2 2 6 1 3 1 r o 1 1 1 N T P t t n n t e t e n n m m e t s e e t e o m s c t t n c m n t e a n n e a e l t t l n c e e c p n n p e m a r r l a e e e l e m m e r e e r p m e e p c t h h m m e e e c c e t t e e r a n e e r l c e a a s e o o l l l e c c r g a n p a n p l p p a a t t t i e r d d n l l k m s s e k i e e v p n n n h r n n o p p l e r r r n n c e l e e e e a a t e e a c r a a o o r l m i i a r r t e e f n t t t r m m m r y s s e t i e e g o f a a o p p e e t o e u n n d n i i o - n n n r r t i i j c c s c o o n h h k k e e o i i d d s l o i i e d l l l l l l l o j s s i e p p g g r r g r u i i r e a a a a a a c e i n n o n o v v b b t t t t t t t t i i i b p p - - i i y y n n e e a o o o o o o s s s o y e e u u K T R H T T H H H r P h u n T T T U R r
1. The total number of procedures presented excludes procedures funded privately and undertaken in the independent sector as these procedures were not linkable
48 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
2.2.1 Linkage process An NJR procedure was classified as being HES/PEDW -linked if linkage to a HES/PEDW episode was Figure 2.1 shows a flow chart illustrating the process achieved using at least one of the five criteria listed of linkage of NJR procedures to HES/PEDW episodes. above. In cases where an NJR procedure was linked Linkage of an NJR procedure to a HES/PEDW to more than one patient record in HES/PEDW, the episode was attempted using each of the following five ‘correct’ link was selected hierarchically in the order of hierarchical criteria in the order as listed below, criteria shown in the left hand column. provided the date of operation recorded in the NJR was within the episode start and end dates in HES/PEDW:
1. local hospital number and NHS Trust code 2. NHS number and NHS Trust code 3. NHS number only 4. patient date of birth, gender and NHS Trust code 5. patient date of birth and NHS Trust code.
Figure 2.1 Flow chart illustrating linkage of the NJR with HES and PEDW
NJR HES/PEDW
442,806 hip or knee joint replacements 538,276 hip or knee joint replacements since entered since April 2003 April 2003 (procedure s in NHS and NHS funded in Independent Sector only)
408,050 primary procedures 498,985 primary procedures 34,756 revisions and other re-operations 39,291 and other re-operations
NJR – HES/PEDW LINKAGE
245,938 linked primary procedures
1,658 first revision procedures (excluding re-operations) identified in HES only
Revision rates Mortality
National Joint Registry 49 2.2.2 Coverage of linked procedures Of all 498,985 primary hip or knee replacement procedures recorded in HES between April 2003 and September 2007, 245,938 were included in the linked All analyses of revision rates, mortality and length of database giving a linkage rate of 49%. However, the stay were based on linked procedures. Details of how number of procedures linked differed between HES episodes of revisions for these primary provider types. On average, 69% of hip or knee replacements were identified are given in Sections replacement procedures undertaken at NHS hospitals 4.2.1 and 5.2.1. or treatment centres were linked to a HES episode (Table 2.1). Fewer procedures were linked for An issue of concern is the potential for systematic Independent Sector Treatment Centres (ISTCs) (37%) differences between the patients whose procedures and independent hospitals (19%). Therefore, the HES were HES-linked and those who were not HES-linked. procedures are more representative of procedures Tables 2.1 to 2.3 show a comparison of undertaken in NHS hospitals and treatment centres. characteristics between linked and non-linked procedures for hip and knee joint replacements. NJR procedures are deemed to be linkable if they fall between 1 st April 2003 and 30 th September 2007 and if they are undertaken in the NHS or NHS funded in the independent sector. However, as funding is not a mandatory field and not completely reported, it was not possible to exclude from the non-HES linked procedures all privately funded procedures undertaken in the independent sector.
50 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 2.2 Comparison of patient, procedure and provider type for HES/PEDW-linked and non-HES/PEDW-linked hip replacement procedures, 1 st April 2003 to 30 th September 2007
Linked to a Not linked to a Total number of HES/PEDW HES/PEDW procedures that were episode episode HES/PEDW-linkable 2
125,749 91,190 216,939 Age, years 108,244 39,400 147,644 (consenting patients only) n
Mean (sd) 68 (11) 68 (11) 68 (11) Interquartile range 61 - 76 61 - 76 61 - 76 n % n % n % Gender (consenting patients only) 3 Male 44,182 41% 15,607 40% 59,789 41% Female 64,061 59% 23,791 60% 87,852 59% Patient physical status P1 - Fit and healthy 27,693 22% 23,106 25% 50,799 23%
P2 - Mild disease not incapacitating 77,682 62% 54,478 60% 132,160 61% P3 - Incapacitating systemic disease 19,371 15% 12,816 14% 32,187 13% P4 - Life threatening disease 921 <1% 744 <1% 1,665 <1% P5 - Expected to die within 24hrs 82 <1% 46 <1% 128 <1% Procedure type THR using cement 61,222 49% 40,956 45% 102,178 47% THR not using cement 32,259 26% 17,900 20% 50,159 23% Hybrid or reverse hybrid THR 13,698 11% 13,765 15% 27,463 13% Primary resurfacing 8,547 7% 5,688 6% 14,235 7% Revision 10,023 7% 12,881 14% 22,904 10% Provider type NHS hospital 110,467 88% 57,760 63% 168,227 78% Independent hospital 4,764 4% 24,780 27% 29,544 14% NHS Treatment Centre 7,684 6% 4,222 5% 11,906 5% Independent Treatment Centre 2,834 2% 4,428 5% 7,262 3% Total 125,749 91,190 216,939
2. The total number of procedures presented excludes procedures funded privately and undertaken in the independent sector as these procedures were not linkable 3. 1 unknown
National Joint Registry 51 5th Annual Report National Joint Registry for England and Wales
Table 2.3 Comparison of patient, procedure and provider type for HES/PEDW-linked and non-HES-linked knee replacement procedures, 1 st April 2003 to 30 th September 2007
Linked to a Not Linked to a Total number of HES/PEDW episode HES/PEDW episode procedures that were HES/PEDW-linkable 4
135,460 90,407 225,867 Age, years (consenting patients only) n 116,151 39,526 155,677 Mean (sd) 70 (9) 69 (10) 70 (9) Interquartile range 64 - 77 62 - 76 63 - 76 N % n % n % Gender (consenting patients only) 5 Male 49,551 43% 17,708 45% 67,259 43% Female 66,599 57% 21,817 55% 88,416 57% Patient physical status P1 - Fit and healthy 31,282 23% 30,400 34% 62,682 27% P2 - Mild disease not incapacitating 83,367 62% 50,203 56% 133,570 59% P3 - Incapacitating systemic disease 20,017 15% 9,397 9% 29,414 13% P4 - Life threatening disease 730 <1% 356 <1% 1,086 <1% P5 - Expected to die within 24hrs 64 <1% 51 <1% 115 <1% Procedure type TKR using cement 112,288 83% 66,498 74% 178,786 79% TKR not using cement 9,895 7% 5,217 6% 15,112 7% Hybrid or reverse hybrid TKR 1,758 1% 1,244 1% 3,002 1% Unicondylar knee replacement 5,293 4% 9,569 11% 14,862 7% Patello-femoral replacement 978 <1% 1,275 1% 2,253 1% Revision 5,248 4% 6,604 7% 11,852 5% Provider type NHS hospital 116,728 86% 53,199 59% 169,927 75% Independent hospital 6,224 5% 27,262 30% 33,486 15% NHS Treatment Centre 9,210 7% 4,382 5% 13,592 6% Independent Treatment Centre 3,298 2% 5,564 6% 8,862 5% Total 135,460 90,407 225,867
Of all HES-linked procedures, about 4% of hip or knee and those that were not HES-linked showed that replacements were undertaken at independent differences in the patient characteristics were small hospitals (Tables 2.2 and 2.3). In comparison, of the (Tables 2.2 and 2.3). The differences between HES- non-HES linked procedures 27% of hip replacements linked and non-HES-linked procedures should be and 30% of knee replacements were undertaken at borne in mind when interpreting the results presented independent hospitals. A comparison of patient in the rest of the Report. characteristics between the HES-linked procedures
4. The total number of procedures presented excludes procedures funded privately and undertaken in the independent sector as these procedures were not linkable 5. 1 unknown
52 National Joint Registry part 2
3. overview of hip and knee replacement procedures
National Joint Registry 53 This section summarises the number of hip and knee procedure into the NJR (Table 3.1). The compliance replacement procedures undertaken in England and rate of 86% (determined from the number of leviable Wales between 1 st January 2007 and 31 st December components) was lower than the 94% centre 2007 and entered into the NJR by 29 th February 2008. participation rate, indicating that not all procedures at The procedures are summarised according to the type the participating centres were recorded. On average, of hospital or treatment centre, funding and patient 181 hip replacements and 190 knee replacements characteristics. were recorded per unit over the year, although the numbers varied from 1 procedure to more than 1,300 3.1 Hospitals and Treatment procedures (Table 3.2). Compared with previous years, there was an overall reduction in the number of Centres Participating in the NJR centres reporting fewer than 100 procedures in a year and an increase in centres reporting more than 400 During the 2007 data collection period, 409 hospitals procedures in a year (Figure 3.1). and treatment centres (TCs) were open and of these 383 (94%) submitted at least one hip or knee
Table 3.1 Total number of hospitals and treatment centres in England and Wales able to participate in the NJR and proportion actually participating in 2007
Total no of units No submitting to NJR in 2007
Total 409 383 94% NHS Hospital 224 201 90% England 207 184 89% Wales 17 17 100% Independent hospitals 164 162 99% England 159 157 99% Wales 5 5 100% NHS Treatment Centres 11 10 91% England 11 10 91% Wales 0 0 Independent Treatment Centres 10 10 100% England 10 10 100% Wales 0 0
54 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 3.2 Number of participating hospitals according to number of procedures performed in 2007
2007 Number of Procedures recorded in NJR in 2007
Total number <50 50 100 200 300 400+ Average number Min Max of hospitals - 99 - 199 - 299 - 399 per unit All Operations Hospitals entering hip 381 69 86 98 66 32 30 181 1 1,284 replacements
Hospitals entering knee 381 76 77 93 54 40 41 190 1 1,512 replacements
Primary Operations Hospitals entering primary hip 381 73 90 105 64 30 19 164 1 1,206 replacements
Hospitals entering primary 381 83 75 94 54 42 33 180 1 1,474 knee replacements
2006 Number of Procedures recorded in NJR in 2007
Total number <50 50 100 200 300 400+ Average number Min Max of hospitals - 99 - 199 - 299 - 399 per unit All Operations
Hospitals entering hip 395 82 85 117 57 27 27 166 1 1,282 replacements
Hospitals entering knee 392 87 98 96 45 35 31 166 1 1,394 replacements
Primary Operations
Hospitals entering primary hip 394 87 96 114 59 18 20 150 1 1,195 replacements
Hospitals entering primary 391 92 100 96 47 29 27 157 1 1,357 knee replacements
National Joint Registry 55 Hip Procedures Figure 3.1 140 Number of participating 120 hospitals by number of procedures 2004 – 2007 100
80
60 2004 2005 40 2006 20 2007 0 <50 50-99 100-199 200-299 300-399 400+
Number of Operations
Knee Procedures 140
120
100
80
60 2004 2005 40 2006 20 2007 0 <50 50-99 100-199 200-299 300-399 400+
Number of Operations
3.2 Hip Replacement Procedures more than 64% of these were cemented procedures. The percentage of Primary Hip Resurfacings This section summarises the patient characteristics undertaken in independent hospitals (14%) is nearly and procedure details of hip replacements according double that of NHS hospitals (8%) (Figure 3.2). There to the type of provider (Table 3.3). 93% of patients at were more revision procedures undertaken at NHS independent hospitals and Independent Sector hospitals (11%) than at any other type of provider (9% Treatment Centres, compared with 82% at NHS units, overall). The revision procedures undertaken at NHS were graded as fit and healthy or with mild disease hospitals comprise 81% of all revision procedures. according to the American Society of Anaesthesiologists Explicit comparisons of patient characteristics for hip (ASA) grade system. Nearly all (98%) of procedures replacement procedures carried out in 2007 with undertaken at ISTCs were primary procedures and those in 2005 and 2006 are reported in Section 4.1.
56 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 3.3 Patient characteristics and procedure details according to type of provider for hip procedures in 2007
Provider Type NHS Ind. NHS Ind. hospitals hospital Treatment Treatment Centres Centre Total
Total 45,106 16,754 3,370 3,683 68,913
Patient Physical Status
P1 - Fit and healthy 7,475 17% 4,814 29% 830 25% 509 14% 13,628 20%
P2 - Mild disease, not incapacitating 29,186 65% 10,656 64% 2,159 64% 2,942 80% 44,943 65%
P3 - Incapacitating systemic disease 8,059 18% 1,230 7% 364 11% 231 6% 9,884 14%
P4 - Life threatening disease 357 1% 42 0% 16 0% 1 0% 416 1%
P5 - Not expected to survive 24 hours 29 0% 12 0% 1 0% 0 0% 42 0%
Procedure Type
Primary Procedure 39,726 88% 15,732 94% 3,185 95% 3,610 98% 62,253 90%
Primary Resurfacing 3,293 8% 2,194 14% 259 8% 95 3% 5,841 9%
Hybrid Total Hip 6,339 16% 2,091 13% 308 10% 299 8% 9,037 15%
Total Hip Replacement not using cement 12,980 33% 5,262 33% 1,537 48% 911 25% 20,690 33%
Total Hip Replacement using cement 17,114 43% 6,185 39% 1,081 34% 2,305 64% 26,685 43%
Revision Procedure 5,113 11% 998 6% 180 5% 62 2% 6,353 9%
Hip Excision Arthroplasty 69 1% 5 1% 0 0% 0 0% 74 1%
Hip Single Stage Revision 4,309 84% 918 92% 158 88% 55 89% 5,440 86%
Hip Stage 1 of 2 Stage Revision 313 6% 37 4% 10 6% 3 5% 363 6%
Hip Stage 2 of 2 Stage Revision 422 8% 38 4% 12 7% 4 6% 476 7%
Hip Re-operation other than Revision 266 1% 24 0% 5 0% 11 0% 306 1%
Bilateral or Unilateral
Bilateral 102 0% 150 1% 0 0% 24 1% 276 0%
Unilateral 45,004 100% 16,604 99% 3,370 100% 3,659 99% 68,637 100%
Primary Complexity Procedure **
Hip Complex Primary 1,948 5% 412 3% 173 5% 66 2% 2,599 4%
Hip Primary 37,779 95% 15,320 97% 3,012 95% 3,544 98% 59,655 96%
Funding **
Independent 823 2% 9,729 58% 14 0% 4 0% 10,570 15%
NHS 36,797 82% 3,711 22% 2,953 93% 2,902 79% 46,363 67%
Not Selected 7,486 17% 3,314 20% 403 13% 777 21% 11,980 17%
** Optional Field
National Joint Registry 57 Hip Procedures 70% Figure 3.2 60% Primary hip procedures by 50% provider type
40%
30%
20%
10% Cemented
0% Cementless NHS hospital NHS Treatment Independent Independent 2003/04 2004/05 2005/06 2006/07 2007/08 Hybrid Centre hospital Treatment Centre Resurfacing
3.3 Knee Replacement Procedures
Table 3.4 summarises the patient characteristics and procedure details of knee replacements according to type of provider. As a percentage of their activity, independent hospitals performed more unicondylar knee replacement procedures (Figure 3.3) compared with any other type of provider. As was observed for hip replacement procedures, there were more revision procedures at NHS hospitals (6%) than at any other type of provider (5% overall). The revision procedures undertaken at NHS hospitals comprised 80% of all revision procedures. Explicit comparisons of patient characteristics for knee replacement procedures carried out in 2007 with those in 2005 and 2006 are reported in Section 5.1.
58 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 3.4 Patient characteristics and procedure details according to type of provider for knee procedures in 2007
Provider Type NHS hospitals Ind. NHS Ind. Total hospital Treatment Treatment Centres Centre Total 48,710 15,206 3,825 4,739 72,480 Patient Physical Status P1 - Fit and healthy 6,225 13% 3,359 22% 552 14% 531 11% 10,667 15% P2 - Mild disease, not incapacitating 33,884 70% 10,552 69% 2,787 73% 3,895 82% 51,118 71% P3 - Incapacitating systemic disease 8,302 17% 1,250 8% 473 12% 309 7% 10,334 14% P4 - Life threatening disease 271 1% 34 0% 13 0% 3 0% 321 0% P5 - Not expected to survive 24 hours 28 0% 11 0% 0 0% 1 0% 40 0% Procedure Type Primary Procedure 45,633 94% 14,664 96% 3,718 97% 4,639 98% 68,654 95% Patello-Femoral Replacement 654 1% 253 2% 30 1% 36 1% 973 1% Hybrid Total Replacement 525 1% 300 2% 44 1% 3 0% 872 1% Total Knee Replacement not using cement 2,656 6% 1,150 8% 623 17% 23 0% 4,452 6% Total Knee Replacement using cement 38,531 84% 11,240 77% 2,808 76% 4,358 94% 56,937 83% Unicondylar Knee Replacement 3,267 7% 1,721 12% 213 6% 219 5% 5,420 8% Revision Procedure 2,709 6% 502 3% 97 3% 69 1% 3,377 5% Knee Amputation 2 0% 0 0% 0 0% 0 0% 2 0% Knee Conversion to Arthrodesis 14 1% 0 0% 0 0% 0 0% 14 0% Knee Single Stage Revision 1,949 72% 411 82% 80 82% 49 71% 2,489 74% Knee Stage 1 of 2 Stage Revision 322 12% 33 7% 7 7% 12 17% 374 11% Knee Stage 2 of 2 Stage Revision 422 16% 58 12% 10 10% 8 12% 498 15% Knee Re-operation other than Revision 368 1% 40 0% 10 0% 31 1% 449 1% Bilateral or Unilateral Bilateral 236 0% 380 2% 4 0% 22 0% 642 1% Unilateral 48,474 100% 14,826 98% 3,821 100% 4,717 100% 71,838 99% Primary Complexity Procedure ** Knee Complex Primary 1,466 3% 368 3% 103 3% 48 1% 1,985 3% Knee Primary 44,167 97% 14,296 97% 3,615 97% 4,591 99% 66,669 97% Funding ** Independent 439 1% 7,372 48% 5 0% 3 0% 7,819 11% NHS 39,615 81% 4,657 31% 3,369 88% 3,705 78% 51,346 71% Not Selected 8,656 18% 3,177 21% 451 12% 1,031 22% 13,315 18%
** Optional Field
National Joint Registry 59 5th Annual Report National Joint Registry for England and Wales
100% Figure 3.3 90% Primary knee procedures by 80% provider type
70%
60%
50%
40%
30%
20% Hybrid 10% Cemented
0% Patello-Femoral NHS hospital NHS Treatment Independent Independent 2003/04 2004/05 2005/06 2006/07 2007/08 Cementless Centre hospital Treatment
Centre Unicondylar
60 National Joint Registry part 2
4. hip replacement procedures
National Joint Registry 61 4.1 Hip Replacement Procedures, 2007
This section summarises the patient characteristics and surgical details of all 62,253 primary, 6,353 revision and 344 re-operation hip replacement procedures undertaken between 1 st January and 31 st December 2007 and entered into the NJR by 29 th February 2008. Fields for which data entry is optional are indicated in the tables of results. All optional fields, apart from “untoward intra-operative events”, were reported in more than 80% of procedures.
62 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.1 Patient characteristics for primary hip replacement procedures in 2007, according to type of procedure
Patient Procedure
Primary Total Primary Total Primary Total Primary Total Prosthetic Prosthetic Prosthetic Resurfacing Replacement Replacement not Replacement not Arthroplasty of using cement using cement classified Joint elsewhere (e.g. HYBRID)
Total hip primaries 26,685 43% 20,690 33% 9,037 15% 5,841 9% 62,253 Total of hip primaries with patient data 21,810 16,406 7,631 4,645 50,492 Average Age 73.0 65.2 68.8 55.3 66.7 Average Age SD 9.1 11.1 10.7 9.0 12.8 Interquartile range 67.8 - 79.2 59.3 - 72.9 62.7 - 76.2 49.9 - 61.4 61.2 - 76.3 Gender Female 14,337 66% 9,398 57% 4,828 63% 1,619 35% 30,182 60% Male 7,473 34% 7,008 43% 2,802 37% 3,026 65% 20,309 40% Unspecified 0% 0% 1 0% 0% 1 0% Patient Physical Status P1 - Fit and healthy 3,512 13% 4,839 23% 1,538 17% 2,821 48% 12,710 20% P2 - Mild disease, not incapacitating 18,617 70% 13,481 65% 6,073 67% 2,833 49% 41,004 66% P3 - Incapacitating systemic disease 4,358 16% 2,266 11% 1,373 15% 179 3% 8,176 13% P4 - Life threatening disease 181 1% 90 0% 49 1% 5 0% 325 1% P5 - Not expected to survive 24 hours 17 0% 14 0% 4 0% 3 0% 38 0% Indications for Surgery Osteoarthritis 25,236 95% 19,270 93% 8,304 92% 5,560 95% 58,370 94% Ankylosing spondylitis 94 0% 78 0% 42 0% 24 0% 238 0% Avasular necrosis 556 2% 584 3% 265 3% 119 2% 1,524 2% Congenital dislocation 128 0% 404 2% 121 1% 147 3% 800 1% Failed hemiarthroplasty 90 0% 24 0% 24 0% 1 0% 139 0% Failed internal fixation 188 1% 163 1% 89 1% 2 0% 442 1% Fractured acetabulum 46 0% 54 0% 31 0% 8 0% 139 0% Fractured neck of femur 432 2% 269 1% 181 2% 4 0% 886 1% Other hip trauma 52 0% 69 0% 15 0% 17 0% 153 0% Other inflammatory arthropathy 91 0% 112 1% 53 1% 17 0% 273 0% Perthes 16 0% 87 0% 18 0% 23 0% 144 0% Previous arthrodesis 4 0% 15 0% 4 0% 2 0% 25 0% Infection 13 0% 22 0% 10 0% 2 0% 47 0% Psoriatic arthropathy 8 0% 13 0% 8 0% 2 0% 31 0% Seropositive rheumatoid arthritis 164 1% 109 1% 67 1% 14 0% 354 1% Slipped upper femoral epiphesis 20 0% 67 0% 20 0% 41 1% 148 0% Other 291 1% 211 1% 158 2% 87 1% 747 1% Side Bilateral 53 0% 160 1% 19 0% 41 1% 273 0% Left, unilateral 11,836 44% 9,277 45% 4,172 46% 2,749 47% 28,034 45% Right, unilateral 14,796 55% 11,253 54% 4,846 54% 3,051 52% 33,946 55%
National Joint Registry 63 THR using cement Figure 4.1 THR not using cement Type of primary hip Hybrid or reverse hybrid replacement procedure undertaken between Hip resurfacing 2005 and 2007
60%
50%
40%
30%
20%
10%
0% 2005 Q1 2005 Q2 2005 Q3 2005 Q4 2006 Q1 2006 Q2 2006 Q3 2006 Q4 2007 Q2 2007 Q2 2007 Q3 2007 Q4
4.1.1 Primary hip replacement at the time of their primary hip replacement (68.4 years procedures, 2007 vs. 65.8 years respectively, Table 4.2). Patients undergoing a resurfacing procedure were the Less than half of the 62,253 primary hip replacement youngest at an average age of 55.2 years (See Figure procedures undertaken in 2007 were cemented total 4.2). According to the American Society of hip replacements (THR) and 5,841 (9%) were hip Anaesthesiologists (ASA) grade system, about a fifth of resurfacing procedures (Table 4.1). Compared with patients undergoing a primary hip replacement in 2007 previous years, there has been a reduction in the use were graded as being “fit and healthy” prior to surgery. of cemented THR procedures and hybrid or reverse Last year about a quarter of patients were graded as hybrid procedures and a corresponding increase in the “fit and healthy”. The single largest indication recorded use of cementless procedures (Figure 4.1). There has for surgery was osteoarthritis, recorded in 94% of been an overall reduction in cemented procedures procedures. Compared to previous years, the patient from 53% during 2004 to 43% during 2007 and an characteristics described above have largely remained increase in cementless procedures from 21% during unchanged. However, there has been an increase in 2004 to 33% during 2007. hip operations in patients with mild or incapacitating disease. For example, procedures in patients with mild 4.1.1.1 Patient characteristics disease increased from 57% in 2004 to 79% in 2007.
Age and gender were recorded only for those patients who gave consent or where consent was not recorded for their personal identifiers to be entered into the NJR (82%). The average age was 66.7 years and about 60% of the patients were female (Table 4.1). On average, female patients were older than male patients
64 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.2 Age and gender for primary hip replacement in 2007
Patient Procedure Primary Total Primary Total Primary Total Primary Total Prosthetic Prosthetic Prosthetic Resurfacing Replacement Replacement not Replacement not Arthroplasty of using cement using cement classified Joint elsewhere (e.g. HYBRID) Average Age by Gender Female 14,337 9,398 4,828 1,619 30,182 Average 73.5 65.9 69.2 54.4 68.37 SD 9.1 10.9 10.9 8.9 12.14 Interquartile range 68.4 - 79.8 59.8 - 73.2 62.8 - 76.8 49.4 - 60.4 62.7 - 77.3 Male 7,473 7,008 2,802 3,026 20,309 Average 71.9 64.4 68.6 55.8 65.84 SD 9.2 11.9 10.4 9.0 12.06 Interquartile range 66.9 - 77.8 58.4 - 72.3 62.5 - 75.2 50.4 - 62.0 59.7 - 74.5 Age Group by Gender Female < 45 years 100 1% 373 4% 132 3% 213 13% 818 3% 45 - 54 years 357 2% 872 9% 287 6% 574 35% 2,090 7% 55 - 64 years 1,859 13% 2,952 31% 1,139 24% 707 44% 6,657 22% 65 - 74 years 5,282 37% 3,341 36% 1,769 37% 109 7% 10,501 35% 75 - 84 years 5,496 38% 1,593 17% 1,239 26% 15 1% 8,343 28% > 85 years 1,243 9% 267 3% 262 5% 1 0% 1,773 6% Male < 45 years 83 1% 411 6% 81 3% 343 11% 918 5% 45 - 54 years 244 3% 763 11% 203 7% 969 32% 2,179 11% 55 - 64 years 1,171 16% 2,239 32% 669 24% 1,311 43% 5,390 27% 65 - 74 years 3,103 42% 2,469 35% 1,136 41% 374 12% 7,082 35% 75 - 84 years 2,445 33% 1,021 15% 627 22% 28 1% 4,121 20% > 85 years 427 6% 105 1% 86 3% 1 0% 619 3%
National Joint Registry 65 Female 60% Figure 4.2 50% Age and gender for primary hip replacement 40% patients in 2007
30%
20% Female <55 10% Female 55-64 0% Female 65-74 Cemented Cementless Hybrid Hip THR THR THR Resurfacing Female 75+
Male 50% 45% 40% 35% 30% 25% 20% 15% Male <55 10% 5% Male 55-64 0% Male 65-74 Cemented Cementless Hybrid Hip THR THR THR Resurfacing Male 75+
4.1.1.2 Surgical techniques was used in less than 10% of all procedures. Compared with previous years, there has been a slight The surgical techniques used in procedures increase in the use of a posterior incision approach undertaken in 2007 are summarised in Table 4.3. and a reduction in the use of cemented prostheses. Patients were mainly positioned laterally. The lateral The use of a lateral patient position increased from position was used more frequently in hybrid and 83% in 2004 to 89% in 2007, while the use of a resurfacing procedures than in cemented and posterior incision approach increased from 39% in cementless procedures. The most frequently used 2004 to 49% in 2007. The reduction in the use of incision approach was posterior for cementless, hybrid cemented stems from 77% in 2004 to 65% in 2007 and resurfacing procedures and lateral for cemented and from 56% to 44% in the use of cemented cups is procedures. Use of minimally invasive surgery (MIS) consistent with the reduction seen in the overall was greatest in cementless procedures, although it number of cemented procedures (Figure 4.1).
66 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.3 Characteristics of surgical practice for primary hip replacement procedures in 2007, according to type of procedure
Primary Total Primary Total Primary Total Primary Total Prosthetic Prosthetic Prosthetic Resurfacing Replacement Replacement not Replacement not Arthroplasty of using cement using cement classified Joint elsewhere (e.g. HYBRID) Total 26,685 20,690 9,037 5,841 62,253 Laminar Flow Yes 20,277 98% 16,447 97% 7,429 98% 4,710 98% 48,863 98% No 437 2% 449 3% 138 2% 77 2% 1,101 2% Not Selected 5,971 3,794 1,470 1,054 12,289 Patient Position Lateral 22,623 85% 18,496 89% 8,603 95% 5,671 97% 55,393 89% Supine 4,062 15% 2,194 11% 434 5% 170 3% 6,860 11% Incision Anterior/Antero-lateral 4,514 17% 3,161 15% 1,054 12% 453 8% 9,182 15% Lateral (inc Hardinge) 11,966 45% 7,259 35% 2,538 28% 949 16% 22,712 36% Posterior 10,133 38% 10,212 49% 5,432 60% 4,426 76% 30,203 49% Other 66 0% 54 0% 12 0% 13 0% 145 0% Trochanteric Osteotomy 6 0% 4 0% 1 0% 0 0% 11 0% Trochanteric Osteotomy Yes 786 3% 282 1% 66 1% 145 3% 1,279 2% No 25,125 94% 19,689 99% 8,762 99% 5,484 97% 59,060 98% Not Selected 774 3% 719 209 212 1,914 Complex Osteotomy Yes 122 0% 30 0% 7 0% 14 0% 173 0% No 25,789 100% 19,941 100% 8,821 100% 5,615 100% 60,166 100% Not Selected 774 719 209 212 1,914 Minimally Invasive Surgery Yes 1,135 5% 2,094 11% 300 4% 106 2% 3,635 7% No 22,822 95% 16,198 89% 8,043 96% 5,024 98% 52,087 93% Not Selected 2,728 2,398 694 711 6,531 Incision Length Greater than 10 cm 19,598 90% 13,481 81% 7,233 92% 4,575 96% 44,887 88% Less than or equal to 10 cm 2,263 10% 3,202 19% 656 8% 180 4% 6,301 12% Not Selected 4,824 4,007 1,148 13% 1,086 11,065 Image Guided Surgery Used Yes 117 1% 122 1% 26 0% 102 2% 367 1% No 23,202 99% 17,750 99% 8,184 100% 4,879 98% 54,015 99% Not Selected 3,366 2,818 827 860 7,871 Femoral Bone graft Used Yes 68 0% 55 0% 13 0% 21 0% 157 0% No 26,617 100% 20,635 100% 9,024 100% 5,820 100% 62,096 100% Acetabular Bone graft Used Yes 474 2% 602 3% 378 4% 121 2% 1,575 3% No 26,211 98% 20,088 97% 8,659 96% 5,720 98% 60,678 97% Femoral cement Used Yes 25,909 100% 9 0% 8,191 93% 5,040 90% 39,149 65% No 2 0% 19,955 100% 637 7% 589 10% 21,183 35% Not Selected 774 726 209 212 1,921 Acetabular cement Used Yes 25,908 100% 8 0% 640 7% 270 5% 26,826 44% No 3 0% 19,956 100% 8,188 93% 5,359 95% 33,506 56% Not Selected 774 726 209 212 1,921
National Joint Registry 67 4.1.1.3 Thromboprophylaxis prescribed from 50% in 2004 to 63% in 2007. The increase can partly be attributed to similar changes in The most frequently prescribed chemical method the prescription of LMWH, which increased from 50% was low molecular weight heparin (LMWH), and in 2005 to 64% in 2007 and TED stockings which TED stockings were the most frequently prescribed increased from 53% in 2005 to 60% in 2007. A mechanical method (Table 4.4). Each of these marginal increase was also seen in the prescription of methods was prescribed for use after about 60% of all a foot pump from 24% in 2004 to 28% in 2007. There procedures. There has been an increase in procedures has been a 2% decrease in the use of Aspirin. in which both chemical and mechanical methods are
Table 4.4 Thromboprophylaxis regime for primary hip replacement patients prescribed at time of operation, according to procedure
Primary Total Primary Total Primary Total Primary Total Prosthetic Prosthetic Prosthetic Resurfacing Replacement Replacement not Replacement not Arthroplasty of using cement using cement classified Joint elsewhere (e.g. HYBRID) Total 26,685 20,690 9,037 5,841 62,253 Aspirin 6,107 23% 4,018 19% 2,433 27% 1,576 27% 14,134 23% Chloroquinine 13 0% 2 0% 2 0% 0 0% 17 0% Low Dose Heparin 186 1% 247 1% 66 1% 85 1% 584 1% Low molecular weight heparin 18,003 67% 13,666 66% 5,043 56% 3,056 52% 39,768 64% Pentasaccharide 170 1% 175 1% 103 1% 66 1% 514 1% Warfarin 392 1% 245 1% 149 2% 111 2% 897 1% Other Chemical 405 2% 219 1% 93 1% 286 5% 1,003 2% Foot pump 6,675 25% 6,212 30% 2,750 30% 1,571 27% 17,208 28% Intermittent calf compression 7,047 26% 5,270 25% 2,397 27% 1,390 24% 16,104 26% TED stockings 15,280 57% 13,175 64% 5,519 61% 3,668 63% 37,642 60% Other Mechanical 253 1% 242 1% 45 0% 28 0% 568 1% None recorded 9 0% 9 0% 1 0% 1 0% 20 0% Both Mechanical and 16,805 63% 13,327 64% 5,559 62% 3,342 57% 39,033 63% Chemical Method
4.1.1.4 Untoward intra-operative events completed, was that no untoward intra-operative event occurred. Of the 605 untoward events reported, Untoward intra-operative events were reported in 1% 41% were calcar cracks. This compares with 32% in of procedures (Table 4.5). Completion of the data field 2006. Pelvic penetration has reduced by 4% and requesting these events was not mandatory and the trochanteric fracture has reduced by 3%. default option, which held when the field was not
68 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.5 Reported untoward intra-operative events for primary hip replacement patients in 2007, according to procedure type
Primary Total Primary Total Primary Total Primary Total Prosthetic Prosthetic Prosthetic Resurfacing Replacement Replacement not Replacement not Arthroplasty of using cement using cement classified Joint elsewhere (e.g. HYBRID) Total 26,685 20,690 9,037 5,841 62,253 Not Selected 13,052 49% 9,060 44% 3,537 39% 2,638 45% 28,287 45% None 13,432 50% 11,329 55% 5,416 60% 3,184 55% 33,361 54% Event Specified 201 1% 301 1% 84 1% 19 0% 605 1% Calcar crack 43 21% 174 58% 26 31% 4 21% 247 41% Pelvic penetration 41 20% 23 8% 10 12% 4 21% 78 13% Shaft fracture 11 5% 13 4% 6 7% 0 0% 30 5% Shaft penetration 6 3% 8 3% 4 5% 0 0% 18 3% Trochanteric fracture 40 20% 37 12% 20 24% 0 0% 97 16% Other 60 30% 46 15% 18 21% 11 58% 135 22%
4.1.2 Hip revisions, 2007 4.1.2.1 Patient characteristics
A total of 6,353 hip revision procedures were reported. Table 4.6 summarises patient characteristics for the Of the 6,353 procedures (Table 4.6), 5,440 (86%) 6,353 hip revision procedures undertaken in 2007. were single stage revision procedures, 363 (6%) were Compared with previous years, the patient Stage 1 of a two-stage process, 476 (7%) procedures characteristics shown in Table 4.6 have largely were Stage 2 of a two-stage revision and 74 (1%) remained unchanged compared with 2006. There has were Excision Arthroplasty type procedures. been a reduction in aseptic loosening as a reason for Compared with previous years, there has been no revision with 55% for 2007 compared with 60% in change in the types of revision procedures carried out. 2006. A similar reduction is seen in Lysis and Pain. However, the main increase is seen in infection with 11% of revision being performed due to infection in 2007, whereas in 2006 8% of revisions were due to infection. With regard to the indication for surgery among two-stage revision procedures (Table 4.7), there was, for example, a reduction in revision for aseptic loosening from 29% in 2005 to 18% in 2007.
National Joint Registry 69 Table 4.6 Patient characteristics for hip revision procedures in 2007, according to procedure type
Hip Single Stage Hip Stage 1 of 2 Hip Stage 2 of 2 Hip Excision Total Revision Stage Revision Stage Revision Arthroplasty
Total 5,440 86% 363 6% 476 7% 74 1% 6,353
Age (excludes consent 'no') 4,325 281 368 56 5,030
Average age 71.0 68.6 67.6 71.6 70.1
SD of Average 11.8 10.2 11.7 12.4 12.1
Interquartiles 64.0 - 79.1 61.1 - 75.8 60.4 - 75.9 70.0 - 82.0 63.5 - 78.7
n n n n n
Gender 4,325 281 368 56 5,030
Female 2,577 60% 136 48% 175 48% 31 55% 2,919 58%
Male 1,748 40% 145 52% 193 52% 25 45% 2,111 42%
Patient Physical Status
P1 - Fit and healthy 786 14% 51 14% 53 11% 2 3% 892 14% P2 - Mild disease, not incapacitating 3,247 60% 194 53% 283 59% 26 35% 3,750 59% P3 - Incapacitating systemic disease 1,336 25% 107 29% 137 29% 43 58% 1,623 26% P4 - Life threatening disease 67 1% 11 3% 3 1% 3 4% 84 1% P5 - Not expected to survive 24 hours 4 0% 0 0% 0 0% 0 0% 4 0%
Indications for Surgery
Aseptic Loosening 3,303 61% 67 18% 76 16% 19 26% 3,465 55%
Lysis 981 18% 42 12% 37 8% 11 15% 1,071 17%
Pain 1,064 20% 52 14% 46 10% 11 15% 1,173 18%
Dislocation/subluxation 789 15% 19 5% 14 3% 14 19% 836 13%
Periprosthetic fracture 398 7% 14 4% 15 3% 6 8% 433 7%
Infection 85 2% 281 77% 318 67% 39 53% 723 11%
Malalignment 344 6% 4 1% 6 1% 4 5% 358 6%
Fractured acetabulum 120 2% 3 1% 5 1% 0 0% 128 2%
Fractured stem 4 0% 2 1% 8 2% 0 0% 104 2%
Fractured femoral head 30 1% 0 0% 2 0% 0 0% 32 1%
Incorrect sizing/head socket mismatch 52 1% 1 0% 1 0% 0 0% 54 1%
Wear of acetabular component 739 14% 6 2% 4 1% 5 7% 754 12%
Dissociation 68 1% 0 0% 1 0% 0 0% 69 1%
Other 331 6% 9 2% 38 8% 2 3% 380 6%
Side
Bilateral 3 0% 0 0% 0 0% 0 0% 3 0%
Left, unilateral 2,791 51% 192 53% 232 49% 35 47% 3,250 51%
Right, unilateral 2,646 49% 171 47% 244 51% 39 53% 3,100 49%
70 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.7 Indication for surgery for hip revision procedures in 2005 to 2007
Indications for Single Stage Revision 2005 2006 2007 Total number of procedures 5,141 5,328 5,440 Aseptic Loosening 3,352 65% 3,377 63% 3,303 61% Lysis 1,127 22% 1,130 21% 981 18% Pain 968 19% 1,051 20% 1,064 20% Infection 86 2% 104 2% 85 2% Indications for Stage 1 of 2 Stage Revision 315 366 363 Aseptic Loosening 91 29% 77 21% 67 18% Lysis 58 18% 56 15% 42 12% Pain 44 14% 59 16% 52 14% Infection 232 74% 296 81% 281 77%
4.1.2.2 Components removed and components used
Both the acetabular and femoral components were removed in over half of all revision procedures (Table 4.8). However, comparing the different types of revision procedures indicates that both components were more likely to be removed during a two-stage revision process or with a hip Excision Arthroplasty type procedure than during a single stage revision. About a third of the femoral or acetabular components used during single stage revision procedures were cemented (Table 4.9).
National Joint Registry 71 Table 4.8 Components removed during hip revision procedures in 2007
Hip Single Stage Hip Stage 1 of 2 Hip Excision Total Revision Stage Revision Arthroplasty
Total 5,328 366 63 5,757
Both components (acetabular and 2,752 52% 309 84% 44 70% 3,105 54% femoral)
Acetabular +/- femoral head 1,329 25% 15 4% 3 5% 1,347 23% Femoral component alone 917 17% 30 8% 10 16% 957 17% Femoral head alone 175 3% 0 0% 0 0% 175 3% None 155 3% 12 3% 6 10% 173 3%
Table 4.9 Components used during single stage hip revision procedures in 2007
Number of Hips Total 5,440 Femoral Prosthesis 5,345 Cemented 1,775 33% Cementless 2,720 51% Not Revised 850 16% Acetabular Prosthesis 5,345 Cemented 1,547 29% Cementless 3,345 63% Not Revised 453 8%
4.1.3 Hip re-operations other than revision, 2007
A total of 344 re-operations procedures following an initial hip replacement were recorded in 2007. Table 4.10 details the procedures performed and Table 4.11 details the characteristics of patients undergoing re-operations.
72 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
Table 4.10 Types of hip re-operation other than revisions in 2007
Total 344 Type of re-operation Socket Augmentation 48 14% Wound Exploration 40 12% Open Reduction Of Dislocation 26 8% Orif Femur 14 4% Excision Heterotopic Bone 8 2% Orif Trochanter 10 3% Focal Bone Graft Only Femur 7 2% Focal Bone Graft Only Acetabulum 5 1% Other Reoperation 186 54%
Table 4.11 Characteristics of patients undergoing hip re-operations other than revision
Total Re-operations 306 Number with patient details 247 Average age 70.6 SD of Average 11.8 Interquartile Range 63.3 – 79.7 Gender 247 Female 156 63% Male 91 37% Patient physical status P1 – Fit and healthy 26 8% P2 – Mild disease, not incapacitating 189 62% P3 – Incapacitating systemic disease 84 27% P4 – Life threatening disease 7 2% P5 – Not expected to survive 24 hours 0 0% Side Bilateral 0 0% Left, unilateral 155 51% Right, unilateral 151 49%
National Joint Registry 73 4.2 Outcomes following primary Some patients had two primary hip replacement procedures, one on each side, which occurred on hip replacement, 2003 – 2007 different dates and were both linked to a HES record. In such cases, to avoid including a patient twice in the This section presents analyses of revision rates linked database, only the earliest primary procedure according to prosthesis type, with special attention was retained. For each patient with a linked primary given to revision of hip resurfacing procedures, bearing procedure, all hospital episodes recorded in HES surface, and head size. Other topics are mortality rates succeeding the primary procedure were identified. and the occurrence of pulmonary embolism according From these succeeding episodes, the first hip revision to the method of thromboprophylaxis. The analyses that occurred on the same side as the primary was are based on data on primary hip replacement then identified in HES. Where the side of the primary procedures undertaken between 1 st April 2003 and or the revision was not recorded, the first revision that 30 th September 2007 that were linked to an episode in occurred after the primary procedure was assumed to the HES database. be a revision of that primary. In HES, revision procedures were identified using the procedure codes As explained in Section 2.2, linkage was not possible W373, W383, W393 and W582 (with Z843 – hip joint) for independently funded procedures undertaken at indicating revision of a cemented, cementless, independent sector. A further issue is that analyses of unclassified and hip resurfacing primary procedure, revision rates can only be carried out if information is respectively. In addition, revision codes indicating available about whether the patients are still alive at conversion from one type of replacement to another, the end of the follow-up period (e.g. 30 th September as well as attention to the replacement joint in which 2007). Data from the mortality registry of the Office for there was renewal of the prosthesis, were also National Statistics is included in HES. Unfortunately, identified as revisions. Codes indicating other types of no mortality data is available in PEDW. Therefore, the attention to the hip replacement joint, such as results presented in this Section are based on NJR- correction of displacement of prosthesis, were HES linked patients only. considered to be re-operations and were not included.
4.2.1 Revision Using the revisions identified in HES as described above, revision rates of primary hip replacement were This section summarises survival to revision of then estimated according to various patient implants used in primary hip replacement procedures characteristics and provider types. The revision rates that were linked to a HES procedure. 102,179 NJR- were estimated using the Kaplan-Meier survival HES linked primary hip replacement procedures that analysis method that allows prostheses that are were undertaken between 1 st April 2003 and 30 th followed up for unequal lengths of time after implant to September 2007 and for which both NJR and HES be included. 30 th September 2007 or time of death indicated that it was a primary procedure were were considered as end of follow-up. identified. 4.2.1.1 Prosthesis type In HES, primary hip replacements for the patients with linked procedures were identified using the following The overall revision rate following primary hip OPCS-4 procedure codes: replacement was 0.6% (95% CI 0.5% – 0.6%) at 1 year and 1.2% (95% CI 1.1% – 1.3%) at 3 years. W371 – Primary total prosthetic replacement of hip Figure 4.3 shows estimates of implant survival to joint using cement revision according to prosthesis type up to 3 years W381 – Primary total prosthetic replacement of hip after primary replacement. joint not using cement W391 – Primary total prosthetic replacement of hip joint not elsewhere classified W581 – Primary resurfacing arthroplasty of joint.
74 National Joint Registry National Joint Registry 5th Annual Report for England and Wales
1.00 Figure 4.3 Implant survival of HES- 0.98 linked primary hip replacement prosthesis 0.96 according to prosthesis type, 2003 – 2007 0.94
Cemented THR 0.92
K Hybrid THR a p l a Cementless THR n 0.90 - M
e Hip Resurfacing i e 0123 r s
u Analysis time (years from operation) r v i v a l e
Res vision rates varied according to type of prosthesis t 4.2.1.2 Age and gender i (p Age <65 years Age 65+ years Figure 4.4 1.00 1.00 Implant survival to revision of HES-linked 0.98 0.98 primary hip replacement prosthesis according to 0.96 0.96 age and prosthesis type, 2003 – 2007 0.94 0.94 Cemented THR 0.92 0.92 K K Hybrid THR a a p p l l a a Cementless THR n n 0.90 0.90 - - M M Hip Resurfacing e e 0 1 2 3 0 1 2 3 i i e e r r s s Analysis time (years from operation) Analysis time (years from operation) u u r r v v i i v v a a l l e e s s t t i i m m a a t t e e National Joint Registry 75 The 3-year revision rates in patients younger than 65 prosthesis than in the group of younger patients; but years were lowest in those who had a cemented the revision rate after resurfacing was higher: 0.7% prosthesis (1.0%, 95%CI 0.8% – 1.3%) and highest (95%CI 0.6% – 0.8%) in those with a cemented after resurfacing (2.7%, 95%CI 2.2% – 3.4%). prosthesis, 1.5% (95%CI 1.3% – 1.8%) for those with Corresponding revision rates in those with cementless a cementless prosthesis, 1.2% (95%CI 0.9% – 1.6%) prosthesis were 2.2% (95%CI 1.8% – 2.6%) and in for those with a hybrid, and 3.8% (95%CI 1.8% – those with a hybrid 1.5% (1.1% – 2.0%) (See Figure 4.4). 7.7%) for resurfacing. However, the statistical evidence that the pattern of revision rates depends on In patients of 65 years and older, the 3-year revision age is weak (p = 0.2). rates were lower for cemented, cementless and hybrid Female Male 1.00 1.00 Figure 4.5 Implant survival to 0.98 0.98 revision of HES-linked primary hip replacement 0.96 0.96 prosthesis according to gender, 2003 – 2007 0.94 0.94 Cemented THR 0.92 0.92 K K Hybrid THR a a p p l l a a Cementless THR n n 0.90 0.90 - - M M e e Hip Resurfacing i i 0 1 2 3 0 1 2 3 e e r r s s Analysis time (years from operation) Analysis time (years from operation) u u r r v v i i v v a a l l e e s s t t i i m m Overall, revision rates are lower in women (1.1%, Given the strong association between the patients a a t t 95%CI 1.0% – 1.2%) than in men (1.4%, 95%CI 1.3% age and sex and the type of prosthesis used, a e e – 1.6%; p < 0.001; see Figure 4.5). However, the multivariable analysis was carried out (Table 4.12). pattern of revision rates according to prosthesis type This analysis also includes physical status and seemed to be related to the patient’s sex (p –value for procedure provider type on 3-year revision rates. interaction 0.02). In women, the 3-year revision rate In Table 4.12, we firstly present revision rates that was 0.6% (95%CI 0.5% – 0.8%) for cemented are not adjusted for other factors. For example, the prostheses, 1.7% (95%CI 1.5% – 2.0%) for revision rates according to prosthesis types do not cementless prostheses, 1.3% (95%CI 0.9% – 1.7%) take into account that patients undergoing hip for hybrids and 3.6% (95%CI 2.7% – 4.8%) for resurfacing are on average younger than those resurfacing. In men, corresponding rates were 0.9% undergoing other types of procedures. (95%CI 0.8% – 1.1%) for cemented prostheses, 2.0% (95%CI 1.7% – 2.4%) for cementless prostheses, Secondly, a multivariable model was used to adjust 1.4% (95%CI 1.0% – 1.9%) for hybrids, and 2.3% simultaneously for confounding factors, such as age, (95%CI 1.9% – 3.3%) for resurfacing. In other words, sex and physical status, when comparing rates for women have higher revision rates than for resurfacing, different procedure types. The results of this type of but lower revision rates for all other prostheses. analysis are expressed as hazard ratios (i.e. the rate of revision within 3 years after replacement at a particular 76 National Joint Registry National Joint Registry 5th Annual Report for England and Wales level of a risk factor divided by the rate at the reference 4.2.1.4 Revision rates for the most level of that factor). For example, the hazard ratio frequently used implant brands shown for hip resurfacing suggest that the risk of revision for a patient who undergoes a resurfacing Table 4.14 and 4.15 show revision rates for cemented procedure is 3.6 times the risk for a patient of the and cementless stems and cups that were that were same age, sex and physical status who undergoes a most frequently used according to the NJR (and at cemented procedure. Likewise, the hazard ratio for least 1,000 of each entered). The highest 3-year hybrid hip replacements indicates that the risk of revision revision rates of the cemented stems were observed for hybrid procedures in the first 3 years after replacement for the CPT and Stanmore modular and the lowest is 1.5 times higher than that of cemented procedures. rate for the Exeter V40. Significance testing provides some evidence that the revision rates of the cemented The hazard ratios presented in Table 4.12 suggest that stems depend on brand (p = 0.04). The highest prosthesis type is the most influential factor on implant revision rates of the cementless stems were observed survival following primary hip replacement. Hip for the SL-Plus and the lowest for the Omnifit and resurfacing procedures have statistically significantly Synergy. Also, for the cementless stems there is some higher revision rates than cemented THR procedures, statistical evidence that the revision rates depend on after adjusting for age, sex, ASA grade and provider brand (p = 0.04). type. For reasons of simplicity, the results presented ignore that the pattern of revision rates according to The 3-year revision rates of the cemented cups varied prosthesis type was related to the patients’ age and sex. from 0.9% for the Exeter Duration to 0.4% for the 4.2.1.3 Comparison of revision rates with Cenator, but the statistical evidence that brand has an effect on the revision rates is very weak (p=0.49). other national registries There was greater variation in the 3-year revision rates of the cementless cups. Highest revision rates was A comparison of revision rates observed in the NJR observed for the CSF and the lowest for the Reflection with those reported in other national registries showed Cementless and statistical testing indicates that it is that the rates reported in the NJR were generally lower unlikely that these difference are due to chance alone than those reported elsewhere. Table 4.13 (p < 0.001). summarises results of a number of national registries that have published quantitative figures on revision rates in patients who had a hip replacement since 4.2.2. Revision rates after 2000. The revision rates observed in Norway and hip resurfacing Australia are higher than observed in the NJR, whereas those observed in New Zealand are very similar. Table 4.16 gives the revision rates for patients who underwent a hip resurfacing according to age, sex, The interpretation of these differences is not physical status and provider type. This table straightforward and a number of explanations have demonstrates that the 3-year revision rates are slightly to be considered. First, the completeness of the higher in patients who are young and in female patients, identification of revisions might be higher in Norway but these difference were not statistically significant and Australia than in England and New Zealand. (p = 0.6 and p = 0.1, respectively). Figures 4.6 shows Second, the definitions of what constitutes a revision that older patients seem to have increased revision may differ between the registries. For example, the rates early after the surgical procedure. Men and women NJR explicitly excludes re-operations other than have similar revision rates during the first year after revisions. Third, data validation procedures in some surgery, after which women seem to have higher rates. registries that check the completeness of the reporting of revisions may lead to selective inclusion of patients The hazard ratios calculated with a multivariable model who had a revision which in turn would overestimate adjusting for all factors included in Table 4.16 did not the revision rate. These possible explanations as well provide any statistical evidence for a difference in as others need to be explored in more detail before revision rates between older and younger patients and any conclusions about the quality of orthopaedic men and women. However, it has to be taken into services can be calculated. account that we calculated these hazard ratios assuming that the relative differences in the revision rates are constant over time after surgery which as shown in Figure 4.6 is not the case. National Joint Registry 77 The reasons for revision of resurfacing replacements In the patients with mild and incapacitating disease differ from those of other procedures (table 4.17). there was strong statistical evidence that the revision Periprosthetic fractures (predominantly of the femoral rates depend on bearing surface (both p<0.001). neck) account for 25% of revisions of resurfacings, Implants with at least one metal surface had lower compared with 5 – 11% for other procedure types. rates than those with ceramic in patients with mild Pain is also a commoner feature for revision, 23% disease. In patients with incapacitating disease or compared with 7 – 10%. Conversely revision because worse the ceramic on ceramic and metal on metal of dislocation is lower, 5% compared with 20 – 27%. implants had the lowest rates. Although the effect of bearing surface seemed to be strong in these patients, Two hip resurfacing brands were entered at least after adjustment for age, sex and prosthesis type the 1,000 times (Table 4.18). The revision rate of the BHR statistical evidence of this effect modification was was lower than that of the Corment 2000 (p = 0.004). weaker ( p=0.03). Implant survival according to physical status is shown in Fig 4.8. 4.2.3 Revision rates according to bearing surface and head size Table 4.22 presents the 3 year revision rates according to head size for patients who received a Of the 102,179 patients who underwent a primary hip primary hip replacement with a metal on polyethylene replacement, data on bearing surface were entered in articulation. The 3 year revision rate is higher in the NJR for 99,397. Table 4.19 presents the 3-year patients who had a head size of 28 mm than in revision rates for the four most frequently entered patients with other sizes (p = 0.01), but the effect on articulation combinations, excluding all resurfacing head size on revision rates disappeared after procedures and implants with large heads (>32mm). adjustment for age, sex, physical status and The revision rates are highest for ceramic on prosthesis type (p = 0.3). polyethylene and ceramic on ceramic (p < 0.001). However, multivariable analysis adjusting for age, sex, 4.2.4 Mortality physical status, and prosthesis type did not confirm this observation (p=0.9). This section summarises mortality rates for all 102,179 first primary hip replacement carried out between 1 st April Metal on polyethylene was used in 81% of the patients 2003 and 30 th September 2007 that were linked to a HES aged 65 or older, whereas it is used in only 47% of those record. younger than 65 (Table 4.20). In the older patient group, there is some evidence that revision rates differ according The mortality rate was 0.7% % (95%CI 0.6% – 0.7%) at to bearing surface (p = 0.01) with metal on polyethylene 3 months and 6.2% (95%CI 5.9% – 6.4%) at 3 years having the lowest revision rate (Figure 4.7). However, after surgery. Table 4.23 shows the results of a the effect of bearing surface seemed to disappear multivariable analysis of mortality within the first 3 years when a multivariable model that adjusted for age, sex, after surgery. According to this analysis with physical status and prosthesis type (p = 0.3). In the adjustment of all factors included in the table, mortality younger patient group, patients who had a metal on is higher in older people and those with a poorer metal prosthesis had a lower revision rate than physical status. Mortality is lower in women than in men. patients who had received a prosthesis with another Even with multivariable adjustment, mortality of patients bearing surface but the statistical evidence for an who had hip resurfacing is lower than that of patients effect of bearing surface is weak (Figure 4.7; p = 0.2). who received another type of prosthesis . Similarly, patients treated in an NHS or independent treatment Metal on polyethylene was used in 61% of the most centre have a lower mortality. The explanation of the active patients, whereas it was used in 75% of those mortality being lower in these group is patient selection with mild disease and 71% of those considered based on factors that are not or only partially included incapacitated. Metal on metal seems to have the in the multivariable analysis. lowest revision rate in fit and healthy patients (Table 4.21) although the statistical evidence for an effect of bearing surface is weak (P = 0.5). 78 National Joint Registry National Joint Registry 5th Annual Report for England and Wales 4.2.5 Thromboprophylaxis The thromboprophylaxis was identified from the NJR and the occurrence of pulmonary embolism in the first three months after surgery was identified from HES using the relevant ICD 10 codes in any of the first four diagnosis fields. Table 4.24 shows that there are only small differences in the characteristics of patients who received different forms of thromboprophylaxis. Similarly Table 4.25 demonstrates small differences in patients who received different forms of chemical thromboprophylaxis. There were only small differences in the occurrence of pulmonary embolism (p = 0.1; Table 4.26). Multivariable analysis did not change these differences appreciably. Similarly, there was no evidence to suggest that mortality rates within 90 days of operation were different across the thromboprophylaxis groups (unadjusted p=0.1; adjusted p=0.5; Table 4.26). The 3-month pulmonary embolism rate was 0.6% (95%CI 0.5% – 0.7%) in the 55,210 patients who received LMWH and 0.5% (95%CI 0.4% – 0.6%) in the 16,662 patients who received aspirin (p =0.01). A multivariable analysis of this difference, adjusting for age, sex, physical status and the use of mechanical prophylaxis, suggest that the pulmonary embolism rate in patients who received LMWH is about 1.2 (95%CI 0.9 – 1.6) times higher than that in patients who received aspirin, but the effect was no longer statistically significant (p = 0.1). The 3-month mortality rate was 0.6% (95%CI 0.5% – 0.7%) in the patients who received (LMWH) . The corresponding rate in patients who received aspirin was higher at 0.7% (95%CI 0.6% – 0.9%; p=0.02). After adjustment for age, sex, physical status and the use of mechanical prophylaxis, the hazard ratio comparing to aspirin is 0.8(95%CI 0.6—0.9) indicating a significantly lower mortality in those who receive (p=0.01). The observed effect of the different forms of thromboprophylaxis on the pulmonary embolism and mortality rates has to be interpreted with caution. It cannot be excluded that, despite using multivariable analysis to adjust for age, sex and physical status, other differences in the treatment groups may underlie these effect, a phenomenon often called “confounding by indication”. National Joint Registry 79 Table 4.12 Revision rate at 3 years for primary hip replacement procedures, 1 st April 2003 to 30 th September 2007 that were linked to a HES episode. 6 Factor Category Number of Revision rate 7 Hazard ratio 8 patients (95% CI) (95% CI) Age < 65 years 35,288 1.8% (1.6% - 2.0%) 1.0 65 - 74 years 36,881 1.0% (0.9% - 1.1%) 0.8 (0.6 - 0.9) 75 + years 30,009 0.9% (0.7 % - 1.0%) 0.8 (0.6 - 0.9) Gender Male 41,220 1.4% (1.3 % - 1.6%) 1.0 Female 60,926 1.1% (1.0% - 1.2%) 0.9 (0.8 - 1.0) Patient physical status P1 - Fit and healthy 23,432 1.3% (1.1% - 1.5%) 1.0 P2 - Mild disease not incapacitating 63,382 1.2% (1.1 % - 1.3%) 1.3 (1.1 - 1.5) P3+ - Incapacitating systemic disease or worse 15,365 1.2% (1.0% - 1.5%) 1.2 (1.5 - 1.9) Procedure type Total replacement using cement 54,769 0.7% (0.6% - 0.8%) 1.0 Total replacement not using cement 28,590 1.8% (1.6% - 2.1%) 2.4 (2.1 - 2.9) Hybrid total replacement 11,501 1.3% (1.1% - 1.6%) 1.7 (1.3 - 2.1) Hip resurfacing 7,319 2.8% (2.3% - 3.5%) 3.6 (2.9 - 4.6) Provider type NHS hospital 88,419 1.3% (1.2% - 1.4%) 1.0 Independent hospital 4,198 0.8% (0.5% - 1.3%) 0.6 (0.4 - 0.9) NHS Treatment Centre 6,868 1.1% (0.7% - 1.6%) 0.7 (0.5 - 0.9) Independent Treatment Centre 2,694 No 3-year follow-up data 0.6 (0.3 - 1.2) 6. Procedures linked to a HES episode were those undertaken in England and exclude those funded privately in the independent sector (See section 2.2 of Part 2) 7. Calculated using the Kaplan-Meier survival analysis method 8. Hazard ratios are based on a multivariable model. Relative hazard of revision within 3 years of primary hip replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses. 80 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.13 Hip revision rates (95% confidence intervals) in other national registries 9 Registry Time period (n) 1 year 3 years Definition of revision Hip Replacement Norwegian 2000-2005 Cemented - 1.8% Exchange or removal of femur, femoral head, acetabulum, all components or complete arthrodesis. Cementless 2.6% Australian 1999-2005 Cemented (12,494) 1.11% 2.14% Revision procedures are categorised as major or minor. A major revision involves removal and/or (0.93% - 1.32%) (1.87% -2.46%) replacement of either femoral stem or acetabular cup or shell. A minor revision is a revision where a major component has not been removed or replaced. Examples of this include exchange of femoral head and or acetabular insert. Cementless (42,937) 1.71% 2.79% (1.58% - 1.84%) (2.61% - 2.98%) Hybrid (29,441) 1.45% 2.38% (1.32% - 1.60%) (2.18% - 2.59%) Resurfacing (7,205) 1.98% 3.12% (1.67% - 2.35%) (2.67% - 3.64%) New Zealand 1999-2005 All primary hips (35,998) 0.3% 1.2% A revision is a new operation in a previously (-) (-) replaced hip joint during which one of the components is exchanged, removed, manipulated or added. It includes excision arthroplasty and amputation but not a soft tissue procedure. 9. Norwegian Arthroplasty Register. Report 2006. Available: http://www.haukeland.no/nrl/Report2006.pdf; Australian Orthopaedic Association. National Joint Replacement Registry, Annual Report 2006. Available: https://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2006.pdf; New Zealand Orthopaedic Association. National Joint Registry, Seven Year Report, January 1999 to December 2005. Available: http://www.cdhb.govt.nz/NJR/reports/7-year-report-master.pdf National Joint Registry 81 Table 4.14 Revision rate at 3 years according to stem brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 10 Type Brand Number of patients Revision rat e 11 (95% CI) Cemented Stems Exeter V40 30,872 0.7% (0.5% - 0.8%) Charnley 9,055 0.8% (0.6% - 1.0%) C-stem 6,280 0.8% (0.6% - 1.2%) CPT 4,427 1.1% (0.7% - 1.5%) Stanmore modular 1,449 1.0% (0.5% - 2.1%) Elite Plus 1,072 0.8% (0.4% - 1.7%) Cementless Stems Corail 9,955 1.4% (1.1% - 1.9%) Furlong 7,564 2.0% (1.6% - 2.5%) Omnifit 1,690 1.1% (0.6% - 1.8%) SL-Plus 1,483 2.4% (1.4% - 3.9%) ABG II 1,265 1.8% (1.1% - 2.8%) Synergy 1,087 1.0% (0.5% - 2.1%) 10. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 11. Calculated using the Kaplan-Meier survival analysis method 82 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.15 Revision rate at 3 years according to cup brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 12 Type Brand Number of patients Revision rate 13 (95% CI) Cemented Cups Contemporary 9,984 0.8% (0.5% - 1.1%) Elite Plus Ogee 7,680 0.6% (0.4% - 0.8%) Charnley 5,716 0.8% (0.6 % - 1.2%) Exeter Duration 4,591 0.9% (0.6% - 1.3%) Charnley Ogee 4,577 0.8% (0.5% - 1.1%) Elite Plus 3,292 0.7% (0.4% - 1.2%) ZCA 2,345 0.7% (0.4% - 1.4%) Opera 2,190 0.4% (0.1% - 1.2%) Low Profile Muller 1,320 0.7% (0.3% - 1.7%) Cenator 1,195 0.4% (0.1% - 1.0%) Cementless Cups Pinnacle 7,173 1.0% (0.6 % - 1.5%) CSF 7,027 1.9% (1.6% - 2.4%) Trident 5,000 1.1% (0.7% - 1.7%) Trilogy 4,647 1.6% (1.2% - 2.0%) Duraloc 3,351 1.7% (1.3% - 2.3%) ABG II 1,627 1.2% (0.7% - 1.9%) EPF-Plus 1,346 2.0% (1.1% - 3.6%) Reflection Cementless 1,272 0.6% (0.3% - 1.2%) 12. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 13. Calculated using the Kaplan-Meier survival analysis method National Joint Registry 83 Table 4.16 Revision rate at 3 years according to patient characteristics and provider for patients who received a primary hip resurfacing between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 14 . Factor Category Number of Revision Rate s15 Hazard ratio 16 patients (95% CI) (95% CI) Age < 55 years 3,555 3.0% (2.3 - 4.1) 1 >=55 3,764 2.5% (1.9 - 3.3) 1.1 (0.8 - 1.6) Sex Male 4,481 2.3% (1.8 - 3.1) 1 Female 2,838 3.6% (2.7 - 4.8) 1.1 (0.8 - 1.7) Patient physical status P1 - Fit and healthy 3,637 2.5% (1.9 - 3.4) 1 P2 - Mild disease not incapacitating 3,368 2.9% (2.1 - 3.9) 1.3 (0.9 - 1.8) P3+ - Incapacitating systemic disease or worse 314 5.4% (2.5 - 11.7) 2.2 (1.2 - 4.2) Provider type NHS hospital 6,425 2.9% (2.3 - 3.5) 1 Independent hospital 300 1.2% (0.4 - 3.8) 0.5 (0.2 - 1.6) NHS Treatment Centre 547 3.8% (1.6 - 8.5) 1.1 (0.6 - 2.1) Independent Treatment Centre 47 0 n/a 14. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 15. Calculated using the Kaplan-Meier survival analysis method 16. Relative hazard of revision within 3 years of primary hip replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 1.00 1.00 Figure 4.6 Implant survival to 0.98 0.98 revision of HES-linked primary hip resurfacings according to age and 0.96 0.96 gender 0.94 0.94 K K a a p p l l a 0.92 a 0.92 n n - <55 - Male M M e e i i e e 55+ Female r r s s 0.90 0.90 u u r r v v i i v v 0 1 2 3 0 1 2 3 a a l l Analysis time (years from operation) Analysis time (years from operation) e e s s t t i i m m a a t t e e 84 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.17 Reasons for revision according to primary prosthesis type for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007. These results are based on linkage of primary replacements and revisions within the NJR only. Cause of Failure Total hip Total hip Hybrid or Hip Total replacement replacement reverse hybrid resurfacing using cement not using total hip cement replacement Total 489 571 191 259 1,510 Perisprosthetic Fracture 27 6% 62 11% 9 5% 47 25% 145 10% Pain 46 9% 55 10% 13 7% 43 23% 157 10% Aseptic Loosening 85 17% 117 20% 43 23% 61 32% 306 20% Infection 119 24% 70 12% 35 18% 13 7% 237 16% Malalignment 38 8% 65 11% 19 10% 21 11% 143 9% Dislocation/Subluxation 115 24% 112 20% 51 27% 10 5% 288 19% Lysis 18 4% 15 3% 6 3% 10 5% 49 3% Fractured Stem 6 1% 7 1% 2 1% 6 3% 21 1% Fractured Acetabulum 1 0% 3 1% 0 0% 1 1% 5 0% Fractured Femoral Head 4 1% 7 1% 0 0% 6 3% 17 1% Incorrect sizing/head socket mismatch 4 1% 5 1% 3 2% 3 2% 15 1% Dissociation of Liner 2 0% 7 1% 0 0% 0 0% 9 1% Other 24 5% 46 8% 10 5% 38 20% 118 8% Table 4.18 Revision rate at 3 years according to resurfacing cup brands for primary hip replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode. 17 Type Brand Number of patients Revision rate 18 (95% CI) Resurfacing BHR 5,163 2.2% (1.7% – 2.8%) Corment 2000 1,183 3.2% (2.1% – 4.6%) 17. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 18. Calculated using the Kaplan-Meier survival analysis method National Joint Registry 85 Table 4.19 Revision rate at 3 years according to bearing surface for primary hip replacement procedures (excluding all resurfacing procedures) undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 19 Bearing surface Number of patients Revision rate 20 (95% CI) Hazard ratio adjusted for age, sex and physical status 21 Metal on polyethylene 59,210 0.9% (0.8% - 1.0%) 1.0 Ceramic on polyethylene 11,586 1.4% (1.1% - 1.7%) 1.0 (0.8 - 1.2) Ceramic on ceramic 6,039 1.5% (1.2% - 2.0%) 0.9 (0.7 - 1.2) Metal on metal 4,013 0.8% (0.5% - 1.3%) 0.9 (0.6 - 1.4) 19. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 20. Calculated using the Kaplan-Meier survival analysis method 21. Relative hazard of revision within 3 years of primary hip replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses Table 4.20 Revision rate at 3 years according to bearing surface and age for primary hip replacement procedures 22 undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 23 Bearing surface Number of patients Revision rate 24 (95% CI) Age <65 Metal on polyethylene 10,135 1.3% (1.0 - 1.6) Ceramic on polyethylene 4,884 1.6% (1.2 - 2.2) Ceramic on ceramic 5,543 1.5% (1.1 - 2.1) Metal on metal 1,063 0.5% (0.2 - 1.3) Age 65+ Metal on polyethylene 49,074 0.8% (0.7 - 0.9) Ceramic on polyethylene 6,702 1.1% (0.9 - 1.5) Ceramic on ceramic 1,496 1.4% (0.8 - 2.5) Metal on metal 2,950 0.9% (0.6 - 1.6) 22. excluding all resurfacing procedures and implants with large heads (>32mm) 23. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 24. Calculated using the Kaplan-Meier survival analysis method 86 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.21 Revision rate at 3 years according to bearing surface and patient physical status for primary hip replacement procedures 25 undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 26 Bearing surface Number of patients Revision rate 27 (95% CI) P1 - Fit and Healthy Metal on polyethylene 9,542 1.1% (1.0% - 1.4%) Ceramic on polyethylene 2,730 1.0% (0.6% - 1.6%) Ceramic on ceramic 2,198 1.1% (0.7% - 2.0%) Metal on metal 951 0.4% (0.2% - 1.2%) P2 - Mild Disease not incapacitating Metal on polyethylene 39,169 0.8% (0.7 - 0.9) Ceramic on polyethylene 7,148 1.4% (1.1 - 1.8) Ceramic on ceramic 3,376 1.9% (1.3 - 2.6) Metal on metal 2,379 0.9% (0.5 - 1.5) P3+ Incapacitating disease or worse Metal on polyethylene 10,499 1.0% (0.7 - 1.2) Ceramic on polyethylene 1,708 1.8% (1.1 - 2.8) Ceramic on ceramic 465 0.9% (0.3 - 2.3) Metal on metal 683 1.3% (0.5 - 4.6) 25. excluding all resurfacing procedures and implants with large heads (>32mm) 26. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 27. Calculated using the Kaplan-Meier survival analysis method National Joint Registry 87 <65 65+ Figure 4.7 1.00 1.00 Implant survival to e e revision of HES-linked t t a a 0.98 0.98 primary conventional m m i i t t s s total hip replacements e e l l a a according to age and v v i i 0.96 0.96 v v r r bearing surface u u s s r r e e i i e e 0.94 0.94 M M - - n n a a l l Metal on polyethylene p p a a 0.92 0.92 K K Ceramic on polyethylene Ceramic on cermaic 0.90 0.90 Metal on metal 0 1 2 3 0 1 2 3 Analysis time (years from operation) Analysis time (years from operation) P1 - Fit & Healthy P2 - Mild Disease P2 + Incapacitated Figure 4.8 1.00 1.00 1.00 Implant survival to revision of HES-linked e e e t t t 0.98 0.98 0.98 primary conventional a a a m m m total hip replacements i i i t t t s s s e e e according to patient l l l a a a v v v physical status and i i i 0.96 0.96 0.96 v v v r r r u u u bearing surface s s s r r r e e e i i i e e e M M M 0.94 0.94 0.94 - - - n n n a a a l l l Metal on polyethylene p p p a a a K K K Ceramic on polyethylene 0.92 0.92 0.92 Ceramic on ceramic Metal on metal 0.90 0.90 0.90 0 1 2 3 0 1 2 3 0 1 2 3 Analysis time (years) Analysis time (years) Analysis time (years) 88 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.22 Revision rate at 3 years according to head size for patients who received a primary hip replacement with a metal on polyethylene articulation undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 28 Bearing surface Number of patients Revision rate 29 Hazard ratio adjusted for (95% CI) age, sex, physical status, and prosthesis type 22 mm 12,362 0.7% (0.6% - 0.9%) 1 26 mm 3,442 0.5% (0.3% - 0.9%) 0.6 (0.3 - 1.2) 28 mm 40,957 1.0% (0.9% - 1.2%) 1.1 (0.8 - 1.4) 32 mm 2,449 0.5% (0.3% - 1.0%) 0.7 (0.4 - 1.4) Other 786 1.9% (0.8% - 4.4%) 1.4 (0.6 - 3.1) 28. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 29. Calculated using the Kaplan-Meier survival analysis method National Joint Registry 89 Table 4.23 Mortality within 3 years according to patient characteristics, prosthesis type and provider for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode. 30 Factor Category Number of Mortality rates 31 Hazard ratio 32 patients (95% CI) (95% CI) Age 33 < 65 years 35,288 2.3% (2.1 - 2.5) 1 65 - 74 years 36,881 4.9% (4.6 - 5.2) 1.8 (1.6 - 2.0) 75 + years 30,009 12.3% (11.8 - 12.9) 4.3 (3.8 - 4.7) Sex 34 Male 41,220 6.9% (6.6 - 7.3) 1 Female 60,926 5.6% (5.4 - 5.9) 0.7 (0.6 - 0.7) Patient physical status P1 - Fit and healthy 23,432 2.9 % (2.6 - 3.2) 1 P2 - Mild disease not incapacitating 63,382 5.6% (5.3 - 5.9) 1.4 (1.3 - 1.6) P3+ - Incapacitating systemic disease or worse 15,365 14.3% (13.5 - 15.1) 3.4 (3.0 - 3.8) Prosthesis type Total replacement using cement 54,769 7.4 % (7.1 - 7.1) 1 Total replacement not using cement 28,590 5.0% (4.6 - 5.4) 0.9 (0.9 - 1.0) Hybrid total replacement 11,501 5.0% (4.5 - 5.6) 0.9 (0.8 - 1.0) Hip resurfacing 7,319 1.2% (0.8 - 1.7) 0.4 (0.3 - 0.6) Provider type NHS hospital 88,419 6.4% (6.2 - 6.6) 1 Independent hospital 4,198 3.2% (2.5 - 4.0) 0.6 (0.5 - 0.7) NHS Treatment Centre 6,868 5.9% (4.9 - 7.1) 0.9 (0.8 - 1.1) Independent Treatment Centre 2,694 N/A 0.5 (0.3 - 0.7) 30. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 31. Calculated using the Kaplan-Meier survival analysis method 32. Relative hazard of revision within 3 years of primary hip replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 33. Age missing for 1 record 34. Sex missing or unknown for 33 records 90 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 4.24 Patient characteristics according to thromboprophylaxis Factor Category Thromboprophylaxis None Chemical Mechanical Both Total only only n=16,942 n=61,236 102,179 n=20,563 n (%) n (%) n (%) n (%) Age 35 < 65 years 542 (38%) 8,136 (36%) 6,005 (35%) 20,605 (34%) 35,288 (35%) 65 - 74 years 540 (37%) 8,166 (36%) 5,960 (35%) 22,215 (36%) 36,881 (36%) 75 + years 356 (25%) 6,261 (28%) 4,977 (30%) 18,415 (30%) 30,009 (29%) Sex 36 Male 572 (40%) 9,146 (41%) 6,712 (40%) 24,790 (41%) 41,220 (40%) Female 866 (60%) 13,413 (59%) 10,225 (60%) 36,422 (60%) 60,926 (60%) Patient physical status P1 - Fit and healthy 327 (23%) 5,204 (23%) 4,316 (25%) 13,585 (22%) 23,432 (23%) P2 - Mild disease not incapacitating 843 (58%) 13,667 (61%) 10,137 (60%) 38,735 (63%) 63,382 (62%) P3+ - Incapacitating or worse 268 (19%) 3,691 (16%) 2,489 (15%) 8,916 (15%) 15,365 (15%) Table 4.25 Patient characteristics according to chemical thromboprophylaxis 37 Factor Category Chemical Thromboprophylaxis Aspirin Low molecular Total 16,662 weight heparin 71,872 n (%) 55,210 n (%) n (%) Age 38 < 65 years 5,940 (36%) 18,408 (33%) 24,348 (34%) 65 - 74 years 5,906 (35%) 20,268 (37%) 26,174 (37%) 75 + years 4,816 (29%) 16,533 (30%) 21,349 (30%) Sex 39 Male 6,873 (41%) 22,080 (40%) 28,953 (40%) Female 9,789 (59%) 33,103 (60%) 42,892 (60%) Patient physical status P1 - Fit and healthy 4,135 (25%) 12,062 (22%) 16,197 (23%) P2 - Mild disease not incapacitating 10,343 (62%) 34,556 (63%) 44,899 (62%) P3+ - Incapacitating or worse 2,184 (13%) 8,592 (15%) 10,776 (15%) 35. Age missing for 1 record 36. Sex missing or unknown in 33 records 37. Aspirin or Low-molecular weight heparin 38. Age missing for 1 record 39. Sex missing or unknown for 33 records National Joint Registry 91 5th Annual Report National Joint Registry for England and Wales Table 4.26 Pulmonary embolism 40 and mortality at 3 months according to method of thromboprophylaxis for patients who received a primary hip replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode Method of Number of patients Pulmonary embolism rate 41 Hazard ratio adjusted for age, thromboprophylaxis (95% CI) sex, and physical status 42 None 1,438 0.5% (0.2% - 1.0%) 1 Chemical only 20,563 0.7% (0.6% - 0.8%) 1.4 (0.6 - 2.9) Mechanical only 16,942 0.7% (0.5% - 0.8%) 1.3 (0.6 - 2.8) Both 61,236 0.5% (0.5% - 0.6%) 1.1 (0.5 - 2.2) Method of Number of patients Mortality rate (95% CI) Hazard ratio adjusted for age, thromboprophylaxis sex, and physical status None 1,438 0.7% (0.4% - 1.3%) 1 Chemical only 20,563 0.6% (0.5% - 0.8%) 0.9 (0.5 - 1.7) Mechanical only 16,942 0.7% (0.6% - 0.9%) 1.0 (0.5 - 1.9) Both 61,236 0.6% (0.6% - 0.7%) 0.9 (0.5 - 1.6) 40. Death not related to pulmonary embolism was considered as a censoring event 41. Calculated using the Kaplan-Meier survival analysis method 42. Relative hazard of revision within 3 years of primary hip replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 92 National Joint Registry part 2 5. knee replacement procedures National Joint Registry 93 5.1 Knee replacement 5.3). Where recorded minimally invasive surgery (MIS) was used in half of unicondylar knee replacement procedures, 2007 procedures, compared with use in 3% of all other types of knee replacement procedures. Compared This section summarises the patient characteristics with previous years, the surgical techniques used in of all 68,654 primary, 3,377 revision and 449 primary knee replacements have largely remained re-operation knee replacement procedures undertaken unchanged. There was, however, an increase in the between 1 st January 2007 and 31 st December 2007 use of MIS in unicondylar knee replacements from and entered into the NJR by 29 th February 2008. 37% in 2004 to 50% in 2007. The use of cement for Fields for which data entry is optional are indicated in patella implantation increased from 42% in 2006 to the tables of results. All optional fields, apart from 59% in 2007. “untoward intra-operative events”, were reported in more than 80% of procedures. 5.1.1.3 Thromboprophylaxis 5.1.1 Primary knee replacement The most frequently prescribed chemical method was LMWH , and TED stockings were the most frequently procedures, 2007 prescribed mechanical method (Table 5.4). Compared with previous years, there has been an increase in the Over 83% of the 68,654 primary knee replacement prescription of a combined chemical and mechanical procedures undertaken in 2007 were cemented total regime from 49% in 2004 to 63% in 2007. The knee replacement (TKR) procedures and 5,420 (8%) increase can be attributed to similar changes in the were unicondylar knee replacements (Table 5.1). prescription of LMWH , which increased from 49% in Compared with previous years the types of primary 2004 to 62% in 2007, and TED stockings which knee replacement procedures undertaken have largely increased from 55% in 2004 to 61% in 2007. A remained the same (Figure 5.1). marginal increase was also seen in the prescription of a foot pump from 24% in 2004 to 28% in 2006. A 5.1.1.1 Patient characteristics similar trend in the prescribed thromboprophylaxis regime was found in primary hip replacement The average age was 69.8 years and 57% of the procedures. patients were female. Patients undergoing a patello- femoral replacement were the youngest, at an average 5.1.1.4 Untoward intra-operative events age of 61 years, with more than 80% of these being female. On average, female patients were one year Table 5.5 shows that untoward intra-operative events older than male patients at the time of their primary were rare, reported in 0.4% of procedures. knee replacement (70.1 years and 69.3 years, Completion of the data field requesting these events respectively, Table 5.2). According to the American was not mandatory and the default option, which held Society of Anaesthesiologists (ASA) grade system, when the field was not completed, was that no 15% of the patients undergoing a primary knee untoward intra-operative event occurred. Of the 261 replacement procedure were graded as “fit and untoward events reported, over 40% were fractures, healthy” (Table 5.1). The single largest indication which is an increase of 7% increase compared with recorded for surgery was osteoarthritis, recorded in 2006. There has been a decrease of 8% in ‘Other’ 97% of all primary procedures. Compared with untoward intra-operative events compared with 2006. previous years, the patient characteristics described above have largely remained unchanged. However, there has been an increase in patients with mild or 5.1.2 Knee revisions, 2007 incapacitating disease prior to surgery. For example, patients with mild disease increased from 62% in 2004 A total of 3,377 knee revision procedures were to 71% in 2007. reported. Of the 3,377 procedures, 2,489 (74%) were single stage revision operations, 374 (11%) were 5.1.1.2 Surgical techniques Stage 1 of a two-stage process, 498 (14%) procedures that were Stage 2 of a two-stage revision (Table 5.6). A further 16 (<1%) procedures were The most common surgical approach was the medial recorded, comprising 14 conversions of previous knee parapatellar, used in over 90% of procedures (Table 94 National Joint Registry National Joint Registry 5th Annual Report for England and Wales replacements to arthrodesis and 2 knee amputations. W421 Primary total prosthetic replacement of knee Compared with previous years, there has been no joint not elsewhere classified (nec) change in the types of revision procedures carried out. W581 Primary resurfacing arthroplasty of joint (used only if procedure code Z846 (knee joint) is also 5.1.2.1 Patient characteristics specified) W521 Primary prosthetic replacement of articulation of The mean age of knee revision patients was 69.7 bone using cement (nec). years (Table 5.6). There were more female (52%) than male patients (48%). Aseptic loosening was the most Some patients had two primary knee replacement common indication for single stage revision and procedures, one on each side, which occurred on infection was the most common indication for two- different dates and were both linked to a HES record. stage revision. Compared with previous years, the In such cases, to avoid including a patient twice in the patient characteristics described above have largely linked database, only the earliest primary procedure remained the same. However, there was an increase was retained. For each linked patient, all hospital in patients with mild or incapacitating disease. For episodes recorded in HES that succeeded the primary example, procedures in patients with mild disease procedure were identified. From these succeeding increased from 59% in 2004 to 66% in 2007. episodes, the first knee revision that occurred on the same side as the primary procedure was then 5.1.2.2 Components removed identified in HES. Where the side of the primary or the revision was not given, the first revision that occurred In three-quarters of the revision procedures, the after the primary procedure was assumed to be a femoral and tibial components were both removed revision of that primary. In HES, revision procedures (Table 5.7). were identified using the procedure codes W403, W413, W423, W582 (only with Z846) and W523 indicating revision of a cemented, cementless, 5.2 Outcomes following primary unclassified, knee resurfacing, and replacement of knee replacement, 2003 – 2007 articulation of bone primary procedure, respectively. In addition, revision codes indicating conversion from This section presents analyses of revision rates one type of replacement to another, as well as according to prosthesis type, including unicondylar attention to the replacement joint in which there was knee replacement. Mortality and the occurrence of renewal of the prosthesis, were also identified as pulmonary thromboembolism according to method of revisions. Codes indicating other types of attention to thromboprophylaxis are described as well. Just as the knee replacement joint, such as correction of explained before, the analyses are based on data of displacement of prosthesis, were considered to be primary joint replacements undertaken between 1 st re-operations and were not included as revisions. April 2003 and 30 th September 2007 that were linked to an episode in the HES database. Using the revisions identified in HES as described above , revision rates were then estimated for various patient 5.2.1 Revision characteristics, as well as procedure and provider types. The number of linked patello-femoral knee This section summarises revision rates of implants replacements was too low to allow reliable estimates used in primary knee replacement that were linked to of revision rates to be obtained and these procedures a HES procedure. 112,556 NJR-HES linked primary are therefore excluded from the results presented knee replacements were identified could be included. below. Revision rates were estimated using the Kaplan-Meier survival analysis method that allows In HES, primary knee replacements for the patients prostheses that are followed up for unequal lengths of with linked procedures were identified using the time after implant to be included (see also 4.2.1). following OPCS-4 procedure codes: W401 Primary total prosthetic replacement of knee joint using cement W411 Primary total prosthetic replacement of knee joint not using cement National Joint Registry 95 5.2.1.1 Prosthesis type cemented prostheses, 1.7% (95%CI 1.3% – 2.4%) for cementless prostheses, 1.1% (95%CI 0.5% – 2.3%) The overall revision rate following primary knee for hybrids, and 2.4% (95% 1.6% – 3.5%) for replacement was 0.3% (95%CI 0.3% – 0.4%) at 1 year unicondylar prosthesis. In men, corresponding and 1.2% (95%CI 1.1% – 1.3%) at 3 years after percentages were 1.3% (95%CI 1.2% – 1.5%) for surgery. Figure 5.3 shows estimates of implant survival cemented prostheses, 1.2% (95%CI 0.9% – 1.8%) for to revision according to prosthesis type up to 3 years cementless prostheses, 2.4% (95%CI 1.3% – 4.6%) after primary replacement. for hybrids, and 1.7% (95%CI1.0% – 2.5) for unicondylar prostheses. In other words, men seem to Revision rates varied according to type of prosthesis have higher revision rates for hybrids and lower (p < 0.001). The 3-year revision rates were lowest in revision rates for unicondylar. patients who received a cemented prosthesis (1.2%, 95%CI 1.1% – 1.3%) and highest in those who A multivariable analysis was carried out to investigate received a unicondylar knee replacement (2.0%, the association between prosthesis type and revision 95%CI 1.5% – 2.7%). The 3-year revision rate was in the first 3 years after surgery adjusted for age, sex 1.5% (95%CI 1.2% – 1.9%) in those who received a physical status and provider type (Table 5.8; see cementless prosthesis and 1.7% (95%CI 1.0% – 4.2.1.2 for further explanation). 2.8%) in those who received a hybrid. Figure 5.3 also demonstrates that there are hardly any differences The results of this multivariable analysis indicate that among the three prosthesis type in the first year after even after adjustment for other factors prosthesis type surgery. The revision rate for unicondylar knee seems to be an important factor determining revision replacements is lower than reported last year. rates (p = 0.01). For example, the risk of revision in the first 3 years after surgery was about 1.4 times higher in 5.2.1.2 Age and gender patients who received a unicondylar knee replacement than in patients who received a cemented prosthesis. Younger patients had higher revision rates than older 5.2.1.3 Comparison of revision rates with patients for all types of prosthesis (Figure 5.4). In other national registries patients younger than 65 years, the 3-year revision was 2.0% (95%CI 1.7% – 2.3%) in patients with a Just as was found for hip replacement, revision rates cemented prosthesis, 2.6% (95%CI 1.8% – 3.7%) in after knee replacement reported in the NJR were those with a cementless prosthesis, 2.1% (95%CI generally lower than those reported in other registries 0.9% – 4.9%) in those with a hybrid, and 2.9% (95%CI (Table 5.9). These differences should be interpreted 2.0% – 4.1%) in those who received a unicondylar. carefully, and all alternative explanations given in Section 4.2.1.3 should be taken into account. In patients of 65 years and older, the 3-year revision rates were 0.9% (95%CI 0.8% – 1.0%) in patients with a cemented prosthesis, 1.1% (95%CI0.8% – 1.5%) in 5.2.1.4 Revision rates for the most those with a cementless prosthesis, 1.5% (95%CI frequently used brands 0.8% – 2.8%) in those with a hybrid, and 1.2% (95%CI 0.7% – 2.0%) in those with a unicondylar prosthesis. The revision rates for the most frequently used brands of total condylar knee replacement are presented in Overall, revision rates were lower in women (1.1%, Table 5.10 The highest revision rate was observed for 95%CI 1.0% – 1.3%) than in men (1.3%, 95%CI 1.2% the Kinemax and the lowest rate for Nexgen, Genesis – 1.5%; p = 0.003; see Figure 5.5). The pattern of 2 and Endoplus. Statistical significance test provides revision rates according to prosthesis type seem to be some evidence that the revision rates depend on different between men and women, but the statistical brand (p = 0.03). The 3-year revision rates for the five evidence for an interaction between prosthesis type most popular brands are similar to those reported last and sex is weak (p = 0.3). In women, the 3-year year, although the rate for Scorpio implants has fallen revision rate was 1.1% (95%CI 0.9% – 1.2%) for from 1.6% (95%CI 1.0 – 2.5) to 1.2 (95%CI 0.9 – 1.6). 96 National Joint Registry National Joint Registry 5th Annual Report for England and Wales The Oxford Partial Knee is the only brand of 5.2.4 Thromboprophylaxis unicondylar knee replacement that was entered in the NJR and linked to HES for more than 1,000 patients Thromboprophylaxis was identified from the NJR and (n = 3,510). All other brands were found less than 500 the occurrence of pulmonary embolism in the first times. The next most popular brands of unicondylar three months after surgery was identified from HES knee replacements were the MG (n = 405) and (see 4.2.3 for further explanation). Table 5.12 shows Preservation (n = 313). The 3-year revision rate for the that there are only small differences in the Oxford Partial Knee was 1.9% (95%CI 1.4% – 2.6%). characteristics of patients who received different forms of thromboprophylaxis. Similarly, Table 5.13 5.2.2 Unicondylar knee replacements demonstrates small differences in patients who received different forms of chemical In 4,356 (90%) of the 4,825 linked primary unicondylar thromboprophylaxis. knee replacements a medial approach was used. The 3-year revision rate for these was 2.1(95%CI 1.6% – The pulmonary embolism and mortality rates at 3 2.9%). Only 119 procedures were recorded as having months after surgery were both highest for those who taken a lateral approach and the remainder were received chemical prophylaxis only but this increase recorded as sub-vastus (n = 38) or other. Minimally was not statistically significant even with adjustment invasive surgery was used in 2,249 (47%) of the linked for age, sex and physical status (Table 5.14). primary unicondylar knee replacements, for which the 3-year revision rate was 1.5% (95%CI 0.9% – 2.7%). The 3-month pulmonary embolism rate was 0.4% The corresponding rate for procedures where (95%CI 0.4% – 0.5%) in the 58,871 who had LMWH minimally invasive surgery was not used was 2.4% and 0.5% (95%CI 0.4% – 0.6%) in the 21,343 who (95%CI 1.7 – 3.4%). Statistical testing provided some had aspirin (p = 0.4). A multivariable analysis, adjusting evidence that the use if minimally invasive surgery has for age, sex and physical status and the use of an important effect (p – 0.07) but this comparison is mechanical prophylaxis, suggests that the pulmonary unadjusted for confounders. embolism rate with is about 0.9 (95%CI 0.7 – 1.1) times that with aspirin (p = 0.3). 5.2.3 Mortality The 3-month mortality rate was 0.6% (95%CI 0.5% – This section summarises mortality rates for all 111,723 0.6%) in the patients who received LMWH . The first primary knee replacement (excluding patello- corresponding rate in patients who received aspirin femoral replacements) carried out between 1 st April was lower at 0.5% (95%CI 0.4% – 0.6%; p = 0.08). 2003 and 30 th September 2007 that were linked to a A multivariable model, with adjustment for age, sex, HES record. physical status and the use of mechanical prophylaxis, indicated that the mortality rate for those patients who The mortality rate was 0.5% (95%CI 0.5% – 0.6%) at 3 received LMWH is about 1.2 (95%CI 1.0 – 1.5) times months and 6.1% (95%CI 6.0% – 6.4%) at 3 years higher than in those who were given aspirin. As after surgery. Table 5.11 shows the results of a described in section 4.2.5, other differences in the multivariable analysis of mortality within the first 3 treatment groups may underlie these effects, despite years after surgery. After adjustment for all factors using multivariable analysis to adjust for age, sex and included in the table, mortality is higher in older people physical status. and in those with a poorer physical status. Mortality is lower in women than in men. Mortality also seems to be lower in those who received a unicondylar replacement than in those who received another type. Similarly, patients treated in the independent sector had a lower mortality than those treated in the NHS. As explained in 4.2.2, the most likely explanation for the latter finding is patient selection. National Joint Registry 97 Table 5.1 Patient characteristics for primary knee replacement procedures in 2007, according to type of procedure Primary Total Primary Total Primary Total Unicondylar Patello- Total Prosthetic Prosthetic Prosthetic Knee Femoral Replacement Replacement Replacement Replacement Replacement using cement not using not classified cement elsewhere (e.g. HYBRID) Total 56,937 83% 4,452 6% 872 1% 5,420 8% 973 1% 68,654 Number with Patient Data 46,809 3,580 730 4,434 800 56,353 Mean Age 70.5 68.6 70.2 64.7 61.8 69.8 Sd 9.2 9.4 9.6 9.6 12.1 9.49 64.2 - 62.2 - 63.9 - 58.2 - 53.1 - 63.3 - Interquartile 77.3 75.4 77.2 71.3 70.9 76.8 Number with Gender information 46,809 3,580 730 4,434 800 56,353 Male 19,768 42% 1,686 47% 318 44% 2,297 52% 156 20% 24,225 43% Female 27,041 58% 1,894 53% 412 56% 2,137 48% 644 81% 32,128 57% Patient Physical Status P1 - Fit and healthy 7,688 14% 746 17% 142 16% 1,358 25% 254 26% 10,188 15% P2 - Mild disease, not incapacitating 40,628 71% 3,110 70% 621 71% 3,608 67% 628 65% 48,595 71% P3 - Incapacitating systemic disease 8,348 15% 575 13% 106 12% 437 8% 87 9% 9,553 14% P4 - Life threatening disease 243 0% 16 0% 3 0% 14 0% 3 0% 279 0% P5 - Not expected to survive 24 hours 30 0% 5 0% 0% 3 0% 1 0% 39 0% Indications for Surgery Osteoarthritis 55,091 97% 4,355 98% 840 96% 5,364 99% 928 95% 66,578 97% Avascular necrosis 221 0% 14 0% 0 0% 32 1% 0 0% 267 0% other inflammatory arthropathy 300 1% 8 0% 10 1% 2 0% 2 0% 322 0% Infection 43 0% 2 0% 1 0% 0 0% 1 0% 47 0% rheumatoid arthritis 1,280 2% 49 1% 14 2% 4 0% 3 0% 1,350 2% Trauma 312 1% 17 0% 6 1% 12 0% 7 1% 354 1% Other 413 1% 36 1% 15 2% 43 1% 42 4% 549 1% Side Bilateral 368 1% 86 2% 15 2% 129 2% 40 4% 638 1% Left, unilateral 26,847 47% 2,108 47% 416 48% 2,747 51% 426 44% 32,544 47% Right, unilateral 29,722 52% 2,258 51% 441 51% 2,544 47% 507 52% 35,472 52% 98 National Joint Registry National Joint Registry 5th Annual Report for England and Wales 90% Figure 5.1 80% Type of primary knee 70% replacement procedure 60% undertaken between 50% 2005 and 2007 40% 30% Patello-Femoral Replacement 20% TKR Hybrid 10% 0% TKR not using cement TKR using cement Unicondylar 2 0 0 5 Q 2 1 0 0 5 Q 2 2 0 0 5 Q 3 2 0 0 5 Q 2 4 0 0 6 Q 2 1 0 0 6 Q 2 2 0 0 6 Q 2 3 0 0 6 Q 2 4 0 0 7 Q 2 1 0 0 7 Q 2 2 0 0 7 Q 2 3 0 0 7 Q 4 National Joint Registry 99 Table 5.2 Age and gender for primary knee replacements in 2007 Primary Total Primary Total Primary Total Unicondylar Patello- Total Prosthetic Prosthetic Prosthetic Knee Femoral Replacement Replacement Replacement Replacement Replacement using cement not using not classified cement elsewhere (e.g. HYBRID) Average Age by Gender Female 27,041 1,894 412 2,137 644 32,128 Mean 70.9 68.8 70.3 64.5 61.1 70.1 SD 9.3 9.7 10.1 10.0 12.2 9.68 64.5 - 62.1 - 62.9 - 57.8 - 52.6 - 63.5 - Interquartile range 77.8 76.0 77.7 71.7 70.5 77.4 Male 19,768 1,686 318 2,297 156 24,225 Mean 70.0 68.5 69.9 64.7 64.0 69.3 SD 9.0 9.1 9.1 9.3 11.8 9.2 63.9 - 62.3 - 65.0 - 58.7 - 56.3 - 63.1 - Interquartile range 76.5 75.1 76.1 71.7 73.2 76.1 Age Group By Gender Female < 45 years 156 1% 22 1% 5 1% 39 2% 51 8% 273 1% 45 - 54 years 1,180 4% 127 7% 28 7% 327 15% 150 23% 1,812 6% 55 - 64 years 5,840 22% 525 28% 89 22% 787 37% 191 30% 7,432 23% 65 - 74 years 10,034 37% 696 37% 139 34% 620 29% 168 26% 11,657 36% 75 - 84 years 8,557 32% 453 24% 136 33% 344 16% 73 11% 9,563 30% > 85 years 1,274 5% 71 4% 15 4% 20 1% 11 2% 1,391 4% Male < 45 years 127 1% 14 1% 4 1% 48 2% 13 8% 206 1% 45 - 54 years 913 5% 87 5% 15 5% 254 11% 21 13% 1,290 5% 55 - 64 years 4,735 24% 504 30% 58 18% 907 39% 41 26% 6,245 26% 65 - 74 years 7,939 40% 657 39% 148 47% 758 33% 51 33% 9,553 39% 75 - 84 years 5,399 27% 380 23% 84 26% 309 13% 30 19% 6,202 26% > 85 years 655 3% 44 3% 9 3% 21 1% 0 0% 729 3% 100 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Female 40% Figure 5.2 35% 30% Age and Gender 25% for primary knee 20% replacement 15% patients in 2007 10% 5% 0% Patello- Hybrid Cementless Cemented Unicondylar Femoral Knee Replacement Replacement Male 50% 45% 40% 35% 30% <55 25% 20% 55-64 15% 65-74 10% 5% 75+ 0% Patello- Hybrid Cementless Cemented Unicondylar Femoral Knee Replacement Replacement National Joint Registry 101 Table 5.3 Characteristics of surgical practice for primary knee replacement procedures in 2007, according to type of procedure Primary Total Primary Total Primary Total Unicondylar Patello- Total Prosthetic Prosthetic Prosthetic Knee Femoral Replacement Replacement Replacement Replacement Replacement using cement not using not classified cement elsewhere (e.g. HYBRID) No of Knees 56,937 4,452 872 5,420 973 68,654 Laminar Flow Yes 43,982 98% 3,676 99% 629 89% 4,254 97% 768 97% 53,309 98% No 899 2% 54 1% 79 11% 110 3% 23 3% 1,165 2% Not Selected 12,056 722 164 1,056 182 14,180 Surgical Approach Lateral parapatellar 665 1% 83 2% 13 1% 161 3% 50 5% 972 1% Medial parapatellar 52,818 93% 4,094 92% 759 87% 4,919 91% 850 87% 63,440 92% Mid-Vastus 43 0% 0% 0% 3 0% 4 0% 50 0% Other 2,510 4% 236 5% 54 6% 260 5% 60 6% 3,120 5% Sub-Vastus 901 2% 39 1% 46 5% 77 1% 9 1% 1,072 2% Minimally Invasive Surgery Used Yes 1,659 3% 166 4% 11 1% 2,732 57% 67 9% 4,635 8% No 48,904 97% 3,603 96% 755 99% 2,028 43% 691 91% 55,981 92% Not Selected 6,374 683 106 660 215 8,038 Image Guided Surgery Used Yes 1,067 2% 166 4% 8 1% 101 2% 13 2% 1,355 2% No 49,404 98% 3,595 96% 761 99% 4,639 98% 739 98% 59,138 98% Not Selected 6,466 691 103 680 221 8,161 Femoral Bone Graft Used Yes 143 0% 43 1% 5 1% 5 0% 2 0% 198 0% No 56,794 100% 4,409 99% 867 99% 5,415 100% 971 100% 68,456 100% Tibial Bone Graft Used Yes 121 0% 12 0% 5 1% 3 0% 0% 141 0% No 56,816 100% 4,440 100% 867 99% 5,417 100% 973 100% 68,513 100% Femoral Cemented Used Yes 55,245 100% 1 0% 145 17% 5,101 97% 813 86% 61,305 92% No 0 0% 4,329 100% 694 83% 135 3% 133 14% 5,291 8% Not Selected 1,692 122 33 184 27 2,058 Tibial Cement Used Yes 55,245 100% 1 0% 678 81% 5,098 97% 1 0% 61,023 92% No 0 0% 4,329 100% 161 19% 138 3% 945 100% 5,573 8% Not Selected 1,692 122 33 184 27 2,058 Patella Cement Used Yes 22,915 63% 231 14% 359 68% 236 12% 867 96% 24,608 59% No 13,430 37% 1,396 86% 167 32% 1,781 88% 39 4% 16,813 41% Not Selected 1,746 305 33 208 27 2,319 Not Replaced 18,846 2,520 313 3,195 40 24,914 102 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 5.4 Thromboprophylaxis characteristics for primary knee replacement patients prescribed Primary Total Primary Total Primary Total Unicondylar Patello- Total Prosthetic Prosthetic Prosthetic Knee Femoral Replacement Replacement Replacement Replacement Replacement using cement not using not classified cement elsewhere (e.g. HYBRID) Total 56,937 4,452 872 5,420 973 68,654 Aspirin 13,560 24% 1,092 25% 164 19% 1,542 28% 258 27% 16,616 24% Chloroquinine 18 0% 1 0% 1 0% 4 0% 0 0% 24 0% Low Dose Heparin 607 1% 17 0% 0 0% 42 1% 7 1% 673 1% Low molecular weight heparin 35,996 63% 2,683 60% 493 57% 2,848 53% 467 48% 42,487 62% Pentasaccharide 350 1% 11 0% 1 0% 32 1% 10 1% 404 1% Warfarin 589 1% 68 2% 13 1% 29 1% 7 1% 706 1% Other Chemical 766 1% 27 1% 3 0% 26 0% 15 2% 837 1% Foot pump 15,101 27% 1,857 42% 330 38% 1,360 25% 325 33% 18,973 28% Intermittent calf compression 14,081 25% 1,525 34% 225 26% 1,552 29% 210 22% 17,593 26% TED stockings 34,650 61% 3,065 69% 586 67% 3,327 61% 543 56% 42,171 61% Other Mechanical 582 1% 40 1% 14 2% 34 1% 15 2% 685 1% None recorded 719 1% 9 0% 6 1% 65 1% 16 2% 815 1% Both Chemical and 36,006 63% 3,153 71% 541 62% 3,337 62% 507 52% 43,554 63% Mechanical Method Table 5.5 Reported untoward intra-operative events for primary knee replacement patients in 2007 according to procedure type Primary Total Primary Total Primary Total Unicondylar Patello- Total Prosthetic Prosthetic Prosthetic Knee Femoral Replacement Replacement Replacement Replacement Replacement using cement not using not classified cement elsewhere (e.g. HYBRID) Number of Knees 56,937 4,452 872 5,420 973 68,654 Not Selected 28,344 2,450 360 2,245 413 33,812 None Specified 28,366 1,989 509 3,157 560 34,581 Total Specified 227 13 3 18 0 261 0.40% Fracture 97 43% 5 38% 0 0% 13 72% 0 0% 115 44% Patella Tendon avulsion 21 9% 2 15% 0 0% 0 0% 0 0% 23 9% Ligament Injury 28 12% 2 15% 0 0% 2 11% 0 0% 32 12% Other 81 36% 4 31% 3 100% 3 17% 0 0% 91 35% National Joint Registry 103 Table 5.6 Patient characteristics for knee revision procedures in 2007, according to procedure type Knee Single Knee Stage 1 Knee Stage 2 Knee Knee Procedure Type Stage of 2 Stage of 2 Stage Conversion to Total Amputation Revision Revision Revision Arthrodesis Number of Knees 2,489 74% 374 11% 498 15% 14 0% 2 0% 3,377 Number with patient data 2,057 306 410 13 1 2,787 Average age 69.8 69.2 668 69.5 95.5 69.67 SD of Average 10.2 10.3 9.3 9.3 10.12 63.0 - 62.3 - 63.5 - 62.4 - 95.5 - 63.0 - Interquartile Range 77.4 76.3 76.3 70.3 95.5 77.3 Gender Female 1,147 56% 126 41% 179 44% 7 54% 0 0% 1,459 52% Male 910 44% 180 59% 231 56% 6 46% 1 100% 1,328 48% Patient physical status P1 317 13% 48 13% 52 10% 0 0% 0 0% 417 12% P2 1,680 67% 216 58% 319 64% 8 57% 1 50% 2,224 66% P3 468 19% 99 26% 124 25% 5 36% 1 50% 697 21% P4 23 1% 11 3% 3 1% 1 7% 0 0% 38 1% P5 1 0% 0 0% 0 0% 0 0% 0 0% 1 0% Indications for surgery Aseptic Loosening 1,165 47% 39 10% 62 12% 1 7% 0 0% 1,267 38% Pain 538 22% 25 7% 22 4% 3 21% 0 0% 588 17% Lysis 305 12% 40 11% 25 5% 0 0% 0 0% 370 11% Wear of polythene component 389 16% 16 4% 13 3% 0 0% 0 0% 418 12% Instability 442 18% 21 6% 10 2% 1 7% 0 0% 474 14% Infection 73 3% 314 84% 399 80% 10 71% 0 0% 796 24% malalignment 240 10% 3 1% 9 2% 0 0% 0 0% 252 7% Dislocation/subluxation 118 5% 8 2% 7 1% 1 7% 0 0% 134 4% Periprosthetic Fractured 98 4% 6 2% 4 1% 0 0% 0 0% 108 3% Stiffness 120 5% 6 2% 6 1% 3 21% 1 50% 136 4% Implant fracture 46 2% 1 0% 1 0% 0 0% 0 0% 48 1% component dissociation 55 2% 2 1% 4 1% 0 0% 0 0% 61 2% other 214 9% 8 2% 17 3% 0 0% 1 50% 240 7% Side Bilateral 3 0% 1 0% 0 0% 0 0% 0 0% 4 0% Left, unilateral 1,170 47% 156 42% 234 47% 5 36% 1 50% 1,566 46% Right, unilateral 1,316 53% 217 58% 264 53% 9 64% 1 50% 1,807 54% 104 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 5.7 Combinations of Implants removed during Single and Stage 1 of 2 Stage knee revision procedures in 2007 Knee Single Stage Knee Revision Total Revision (Stage 1 of 2) Number of Knees 2,489 374 2,863 Femoral component and not tibial component 218 9% 7 2% 225 8% Tibial component and not femoral component 157 6% 4 1% 161 6% Femoral component and tibial component 1,779 71% 335 90% 2,114 74% Other combination (not femoral or tibial component) 335 13% 28 7% 363 13% 1.00 Figure 5.3 Implant survival of HES- 0.98 linked primary knee replacement prosthesis according to prosthesis 0.96 type, 2003 – 2007 0.94 Cemented TKR 0.92 Cementless TKR K a p l Hybrid THR a n 0.90 - M Unicondylar e 0 1 2 3 i e r Analysis time (years from operation) s u r v i v a l e s <65 65+ t i m 1.00 1.00 a Figure 5.4 t e Survival to revision of e e t t a a 0.98 0.98 HES-linked primary knee m m i i t t replacement prosthesis s s e e l according to age and l a a 0.96 0.96 v v i i prosthesis type, v v r r u u 2003 – 2007 s s r r e e i i 0.94 0.94 e e M M - - n n a a l l p p Cemented TKR a a 0.92 0.92 K K Cementless TKR 0.90 0.90 Hybrid TKR 0 1 2 3 0 1 2 3 Unicondylar Analysis time (years from operation) Analysis time (years from operation) National Joint Registry 105 male female 1.00 1.00 Figure 5.5 Implant survival to e e t 0.98 t 0.98 revision of HES-linked a a m m i i t t primary knee s s e e replacement prosthesis l l a 0.96 a 0.96 v v i i according to gender and v v r r u u prosthesis type, s s r r e e 2003 – 2007 i 0.94 i 0.94 e e M M - - n n a a l l p p Cemented TKR a 0.92 a 0.92 K K Cementless TKR 0.90 0.90 Hybrid TKR 0 1 2 3 0 1 2 3 Unicondylar Analysis time (years from operation) Analysis time (years from operation) 106 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 5.8 Revision rate at 3 years for primary knee (excluding patello femoral) replacement procedures, 1 st April 2003 to 30 th September 2007 and linked to a HES episode. 43 Factor Category Number of Revision rate 44 Hazard ratio 45 patients (95% CI) (95% CI) Age 46 < 65 years 29,946 2.1% (1.9% - 2.4%) 1.0 65 - 74 years 43,618 1.1% (1.0% - 1.2%) 0.6 (0.5 - 0.7) 75 + years 38,517 0.7% (0.6 % - 0.9%) 0.4 (0.4 - 0.5) Sex 47 Male 47,296 1.3% (1.2 % - 1.5%) 1.0 Female 64,400 1.2% (1.0% - 1.3%) 0.9 (0.7 - 1.0) Patient physical status P1 - Fit and healthy 27,232 1.5% (1.4% - 1.7%) 1.0 P2 - Mild disease not incapacitating 66,256 1.1% (1.0 % - 1.2%) 0.8 (0.7 - 0.9) P3+ - Incapacitating systemic disease or worse 16,235 1.0% (0.8% - 1.3%) 0.8 (0.6 - 1.0) Procedure type Total replacement using cement 97,207 1.2% (1.1% - 1.3%) 1.0 Total replacement not using cement 8,113 1.5% (1.2% - 1.9%) 1.3 (1.0 - 1.7) Hybrid total replacement 1,578 1.7% (1.0% - 2.8%) 1.6 (1.0 - 2.6) Unicondylar 4,825 2.0% (1.5% - 2.7%) 1.4 (1.0 - 1.8) Provider type NHS hospital 95,067 1.3% (1.2% - 1.4%) 1.0 Independent hospital 5,337 1.0% (0.7% - 1.4%) 1.0 (0.7 - 1.3) NHS Treatment Centre 8,234 1.0% (0.7% - 1.4%) 0.7 (0.5 - 1.0) Independent Treatment Centre 3,085 No 3-year follow-up data 0.4 (0.1 - 1.1) 43. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 44. Calculated using the Kaplan-Meier survival analysis method 45. Relative hazard of revision within 3 years of primary knee replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 46. Age missing for 2 records 47. Sex missing or unknown for 27 records National Joint Registry 107 Table 5.9 Knee revision rates (95% confidence intervals) in other national registries 48 Registry Time period (n) 1 year 3 years Definition of revision Knee Replacement Swedish 1995-2004 Total knee replacement of (48,838) 0.81% 2.36% Revisions are defined as removal, osteoarthritis (0.73% - 0.90%) (2.21% - 2.55%) addition and exchange of components. Re-operations such Total knee replacement for 0.83% 2.53% (4,130) as lavage or synovectomy are not rheumatoid arthritis (0.59% - 1.16%). (2.06% - 3.17%) registered as revisions if some part (typically a polyethylene inlay) has Unicondylar knee replacement (9,755) 1.46% 4.95% not been exchanged/removed of osteoarthritis (1.23% - 1.73%) (4.5% - 5.49%) during the operation. Australian 1999-2005 Total knee replacement (107,802) 1.05% 2.84% Revision procedures are categorised (0.99% - 1.12%) (2.71% - 2.96%) as major or minor. A major revision involves the removal and/or Unicondylar knee replacement (18,398) 2.37% 6.24% replacement of a major component. (2.15% - 2.62%) (5.83% - 6.69%) This is defined (with the exception of the patella) as a component that interfaces with bone i.e. either the femoral and/or tibial component. A minor revision is a revision where a major component has not been removed or replaced. Examples of this include patella replacement, tibial insert exchange, or both. New Zealand 1999-2005 All primary knees (23,565) 0.3% (-) 1.7% (-) Revision is defined as a new operation in a previously replaced knee joint during which one or more of the components are exchanged, removed, manipulated or added. It includes arthrodesis or amputation, but not a soft tissue procedure. 48. Norwegian Arthroplasty Register. Report 2006. Available: http://www.haukeland.no/nrl/Report2006.pdf; 2 Australian Orthopaedic Association. National Joint Replacement Registry, Annual Report 2006. Available: https://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2006.pdf; New Zealand Orthopaedic Association. National Joint Registry, Seven Year Report, January 1999 to December 2005. Available: http://www.cdhb.govt.nz/NJR/reports/7-year-report-master.pdf; Swedish Knee Arthroplasty Register. Annual Report 2006. Available: http://www.knee.nko.se/english/online/uploadedFiles/107_SKAR2006_Englv2.pdf 108 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 5.10 Revision rate at 3 years according to brands for primary total condylar knee replacement procedures undertaken between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 49 Brand Number of patients Revision rate 50 (95% CI) PFC Sigma 37,322 1.1% (1.0% - 1.3%) AGC 14,127 1.1% (0.9% - 1.4%) Nexgen 14,025 0.8% (0.6% - 1.1%) Scorpio 8,247 1.2% (0.9% - 1.6%) Kinemax 6,293 1.7% (1.4% - 2.2%) LCS 4,633 1.1% (0.8% - 1.7%) Genesis 2 3,711 0.9% (0.6% - 1.6%) Endoplus 2,425 0.9% (0.4% - 1.8%) Profix 2,274 1.2% (0.7% - 2.2%) Insall-Burstein 2 1,821 1.3% (0.8% - 2.1%) 49. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 50. Calculated using the Kaplan-Meier survival analysis method National Joint Registry 109 Table 5.11 Mortality within 3 years according to patient characteristics, prosthesis type and provider for patients who received a primary knee replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 51 . Factor Category Number of Mortality 52 Hazard ratio 53 patients rates(95% CI) (95% CI) Age 54 < 65 years 29,946 2.0% (1.7 - 2.3) 1 65 - 74 years 43,618 4.5% (4.2 - 4.8) 2.1 (1.8 - 2.4) 75 + years 38,517 11.1% (10.6 - 11.6) 5.4 (4.8 - 6.1) Sex Male 47,296 7.4% (7.1 - 7.8) 1 Female 64,400 5.2% (5.0 - 5.4) 0.6 (0.6 - 0.7) Patient physical status P1 - Fit and healthy 27,232 4.6 % (4.3 - 4.9) 1 P2 - Mild disease not incapacitating 66,256 6.0% (5.7 - 6.4) 1.2 (1.1 - 1.3) P3+ - Incapacitating systemic disease or worse 16,235 11.4% (10.6 - 12.3) 2.1 (1.9 - 2.3) Prosthesis type Total replacement using cement 97,207 6.3 % (6.1 - 6.6) 1 Total replacement not using cement 8,113 5.7% (5.0 - 6.5) 0.9 (0.8 - 1.1) Hybrid total replacement 1,578 5.5% (4.1 - 7.3) 1.0 (0.7 - 1.3) Unicondylar 4,825 3.0% (2.3 - 3.9) 0.6 (0.4 - 0.7) Provider type NHS hospital 95,067 6.3% (6.1 - 6.6) 1 Independent hospital 5,337 3.6% (3.0 - 4.4) 0.6 (0.5 - 0.8) NHS Treatment Centre 8,234 6.5% (5.5 - 7.7) 1.0 (0.9 - 1.1) Independent Treatment Centre 3,085 N/A 0.6 (0.4 - 0.8) 51. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 52. Calculated using the Kaplan-Meier survival analysis method 53. Relative hazard of revision within 3 years of primary knee replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 54. Age missing for 2 records 55. Sex missing or unknown for 27 records 110 National Joint Registry National Joint Registry 5th Annual Report for England and Wales Table 5.12 Patient characteristics according to thromboprophylaxis Factor Category Thromboprophylaxis Chemical Mechanical None only only Both Total N=1,581 N=22,649 N=19,564 N=67,927 111,723 Age 56 < 65 years 434 (27%) 6,275 (28%) 5,298 (27%) 17,939 (26%) 29,946 (27%) 65 - 74 years 629 (40%) 8,949 (39%) 7,519 (38%) 26,521 (39%) 43,618 (39%) 75 + years 518 (33%) 7,425 (33%) 6,747 (35%) 23,467 (35%) 38,157(34%) Sex 57 Male 695 (44%) 9,741 (43%) 8,184 (42%) 28,676 (42%) 47,296 (42%) Female 886 (56%) 12,908 (57%) 11,377 (58%) 39,229 (58%) 64,400 (58%) Patient physical status P1 - Fit and healthy 823 (52%) 6,519 (28%) 5,441 (28%) 14,449 (21%) 27,232 (24%) P2 - Mild disease not incapacitating 612 (39%) 12,618 (56%) 11,509 (59%) 43,517(64%) 68,256 (61%) P3+ - Incapacitating or worse 146 (9%) 3,513 (16%) 2,614 (13%) 9,962 (15%) 16,235 (15%) 56 Age missing for 2 records 57 Sex missing or unknown for 27 records Table 5.13 Patient characteristics according to chemical thromboprophylaxis Factor Category Chemical Thromboprophylaxis 58 Low molecular Aspirin weight heparin Total N=21,343 N=21,343 N=80,214 Age 59 < 65 years 5,855 (27%) 15,617 (27%) 21,472 (27%) 65 - 74 years 8,323 (39%) 23,105 (39%) 31,428 (39%) 75 + years 7,165 (34%) 20,147 (34%) 27,312 (34%) Sex 60 Male 9,092 (43%) 24,956 (42%) 34,048 (42%) Female 12,252 (57%) 33,891 (58%) 46,142 (58%) Patient physical status P1 - Fit and healthy 5,504 (26%) 13,230 (22%) 18,734 (23%) P2 - Mild disease not incapacitating 13,089 (61%) 36,553 (62%) 49,642 (62%) P3+ - Incapacitating or worse 2,750 (13%) 9,088 (16%) 11,838 (15%) 58. Aspirin or Low-molecular weight heparin 59. Age missing for 2 records 60. Sex missing for 24 records National Joint Registry 111 5th Annual Report National Joint Registry for England and Wales Table 5.14 Pulmonary embolism 61 and mortality rates 62 at 3 months according to method of thromboprophylaxis for patients who received a primary knee replacement between 1 st April 2003 and 30 th September 2007 that were linked to a HES episode 63 Method of Number of patients Pulmonary embolism rate Hazard ratio adjusted for age, thromboprophylaxis (95% CI) sex, and physical status 64 None 1,581 0.4% (0.2% - 0.9%) 1 Chemical only 22,650 0.5% (0.5% - 0.6%) 1.4 (0.6 - 3.2) Mechanical only 19,564 0.4% (0.3% - 0.5%) 1.0 (0.4 - 2.2) Both 67,928 0.4% (0.4% - 0.5%) 1.1 (0.5 - 2.5) Number of patients Mortality rate (95% CI) Hazard ratio adjusted for age, sex, and physical status None 1,581 0.4% (0.2% - 0.9%) 1 Chemical only 22,650 0.6% (0.5% - 0.7%) 1.4 (0.6 - 3.1) Mechanical only 19,564 0.4% (0.4% - 0.6%) 1.1 (0.5 - 2.5) Both 67,928 0.5% (0.5% - 0.6%) 1.3 (0.6 - 2.9) 61. Death not related to pulmonary embolism was considered as a censoring event 62. Calculated using the Kaplan-Meier survival analysis method 63. Procedures linked to a HES episode were those undertaken in England and excludes those funded privately in the independent sector (See section 2.2 of Part 2) 64. Relative hazard of revision within 3 years of primary knee replacement compared to a patient with a reference level of the factor (hazard ratio=1), adjusted for all other factors included in the analyses 112 National Joint Registry glossary National Joint Registry 113 A Acetabular component The portion of a total hip replacement prosthesis that is inserted into the acetabulum – the socket part of a ball and socket joint. Acetabular cup (hip) See Acetabular component. Acetabular prosthesis See Acetabular component. Arthroplasty A procedure where natural joint, or part of a natural joint, is replaced by an artificial prosthesis. ASA American Society of Anaesthesiology (ASA) scoring system for grading the overall physical condition of the patient as ‘fit and healthy’, ‘mild disease, not incapacitating’, ‘incapacitating system disease’ or ‘life threatening disease’. B Bilateral operation Operation performed on both sides, e.g. left and right knee procedures carried out during a single operation. Brand (of prosthesis) The brand of a prosthesis (or implant) is the manufacturer’s product name, e.g. the Charnley brand for hips, the Rotaglide Plus brand for knees. C Case ascertainment Proportion of all relevant joint replacement procedures performed in England and Wales that are entered in the NJR. Case mix Term used to describe the variation in the surgical practice in terms of factors such as indications for surgery, patient age, sex etc. Cement gun A pressurised container used to insert bone cement into bony cavities. Cement pressuriser A device used to aid surgeon in optimising the strength of adhesion between bone cement and bone. Compliance The percentage of all total joint procedures which were performed in an individual unit that have been entered into the NJR within any given period. Cup See Acetabular component. D Data collection periods for The NJR Annual Report Part 1 is about data collected between 1 st April 2007 annual report analysis and 31 st March 2008 – the 2007/08 financial year. The NJR Annual Report Part 2 analyses data on hip and knee procedures undertaken between 1st January and 31 st December 2007 inclusive – the 2007 calendar year. Data quality Completeness, accuracy and linkability of NJR data. F Femoral component (hip) Part of total hip joint that is inserted into the femur (thigh bone) of the patient. It normally consists of a stem and head (ball). Femoral component (knee) Portion of a knee prosthesis that is used to replace the articulating surface of the femur (thigh bone). Femoral head Ball shaped portion of the femur that forms part of the ball and socket hip joint. Femoral prosthesis Portion of a total joint replacement used to replace damaged parts of the femur (thigh bone). Femoral stem See Femoral component (hip). 114 National Joint Registry H Head See Femoral head. Healthcare provider NHS or independent sector organisation that provides healthcare; in the case of the NJR, orthopaedic hip and knee replacement surgery. HES Hospital Episode Statistics Hybrid procedure Joint replacement procedure where cement is used on one articulated surface and the other is cementless. I Image guided surgery Surgery performed by the surgeon, using real time images (normally x-rays) to help the alignment and positioning of prosthetic components. Indication (for surgery) Reason for surgery. The NJR system allows for more than one indication to be recorded. L Laminar flow (in theatres) System which ensures a continuous flow of ‘clean’ air around the patient during surgical procedures. Levy Additional payment placed on the sales of specific hip and knee implants to cover the costs associated with on-going operations and development of the NJR. Linkable percentage Linkable percentage is an estimate of the percentage of all relevant procedures that have been entered into the NJR and may be linked via NHS number to other procedures performed on the same patient. Linkable procedures Procedures entered into the NJR database that are linkable to a patient’s previous or subsequent procedures by the patient’s NHS number. M MDS Minimum data set, the set of data fields collected by the NJR. Some of the data fields are mandatory (i.e. they must be filled in). Fields that relate to patients’ personal details must only be completed where informed patient consent has been obtained. MDS 1 (MDSv1) Minimum data set version one, used to collect data from 1 st April 2003. MDS 1 closed to new data entry on 1 st April 2005. MDS 2 (MDSv2) Minimum data set version two, introduced on 1 st April 2004. MDS 2 replaced MDS 1 as official data set on 1 st June 2004. MDS 3 (MDSv3) Minimum data set version two, introduced on 1 st November 2007 as the new official data set. Minimally invasive surgery Surgery performed using special instruments that allow very small incisions to be made (MIS) in the tissues of the patient. Mixing and matching Also know as ‘cross-breeding’. Hip replacement procedure where a surgeon chooses to implant a femoral component (incorporating a metallic or ceramic modular head) from one manufacturer with an acetabular component (incorporating a polyethylene bearing surface) from another. Modular Component composed of more than one piece, e.g. a modular acetabular cup shell component. National Joint Registry 115 N NHS National Health Service. NICE National Institute for Health and Clinical Excellence. NICE benchmark See ODEP ratings. NJR National Joint Registry for England and Wales. Since 1 st April 2003, the NJR has collected and analysed data on hip and knee replacements. It covers both the NHS and Independent Healthcare Sectors to ensure complete recording of national activity in England and Wales. NJR Centre National co-ordinating centre for the NJR. NJR StatsOnline Web facility for viewing and downloading NJR statistics on www.njrcentre.org.uk NSTS NHS Strategic Tracing Service. Used to source missing NHS numbers and to determine when patients recorded on the NJR have died. O ODEP Orthopaedic Data Evaluation Panel of the NHS Purchasing and Supply Agency (PASA). ODEP ratings ODEP ratings are the criteria for prostheses for primary total hip replacement product categorisation against NICE benchmarks. The categorisation based on the NICE benchmarks for products – pre-entry benchmark (products commercially available that are involved in postmarket clinical follow-up studies); entry benchmark (3 years, 5 years and 7 years; level A – acceptable evidence, level B – weak evidence); full benchmark (10 years; level A – strong evidence, level B – reasonable evidence, level C – weak evidence). Under each year is a level that is unacceptable evidence, where product should only be used as part of a clinical trial. OPCS-4 Office of Population, Censuses and Surveys: Classification of Surgical Operations and Procedures, 4 th Revision. P PASA NHS Purchasing and Supply Agency. Patella resurfacing Replacement of the surface of the patella (knee cap) with a prosthesis. Patello-femoral knee Procedure involving replacement of femoral condyles and resurfacing of the patella. replacement Patello-femoral prosthesis Two piece knee prosthesis that provides a prosthetic articulation surface between the (knee) patella and femoral condyles. Patient consent Patient personal details may only be submitted to the NJR where explicit informed patient consent has been given. If a patient does not give consent, only the anonymous operation and implant data may be submitted. Patient physical status See ASA. Patient procedure Type of procedure carried out on a patient, e.g. primary total prosthetic replacement using cement. PEDW Patient Episode Database Wales, the Welsh equivalent to Hospital Episode Statistics (HES) in England. Primary hip/knee First total joint replacement operation performed on any individual patient. replacement Posterior Lipped Special design of cup. Augmentation Device (PLAD) Prosthesis Orthopaedic implant using joint replacement procedures, e.g. a total hip or a unicondylar knee. Pulsatile lavage Pulsed jet of sterile water used to clean the bony surfaces prior to the implantation of a total joint replacement. 116 National Joint Registry R Re-operation other than Procedures following a primary replacement that do not require component removal or revision replacement. Resurfacing (hip) Resurfacing of the femoral head with a surface replacement femoral prosthesis and insertion of an acetabular cup with or without cement. Reverse hybrid procedure Hip procedure where the acetabular prosthesis is cemented and the femoral prosthesis (hip) is not cemented. Revision hip/knee Operation performed to remove and replace one or more components of a total joint replacement prosthesis for whatever reason. S Socket augmentation Dislocation of the joint. Stem See Femoral component (hip). Surgical approach Surgical technique used by a surgeon to expose the bone prior to joint replacement whilst minimising the damage to surrounding tissues. T TED stockings Pressurised stockings worn by patients following surgery. These help to prevent blood clots forming in the blood vessels of the legs. THR Total hip replacement (total hip arthroplasty). Replacement of the femoral head with a stemmed femoral prosthesis and the insertion of an acetabular cup, with or without cement. Thromboprophylaxis Drug or other post-operative regime prescribed to patients with the aim of preventing blood clot formation in the postoperative period. TKR Total knee replacement (total knee arthroplasty). Replacement of both tibial and both femoral condyles, with or without resurfacing of the patella and with or without cement. Total condylar knee Type of knee prosthesis that replaces the complete contact area between the femur and the tibia of a patient. Treatment centre (TC) Treatment centres are dedicated units that offer elective and short stay surgery and diagnostic procedures in specialities such as ophthalmology, orthopaedic and a range of other conditions. These include hip and knee replacements. Treatment centres may be NHS or privately funded. Trochanter Bony protuberance of the femur found just below the femoral head. Trochanteric osteotomy Temporary incision of the trochanter, used to aid exposure of hip joint during some types of total hip replacement. Type (of prosthesis) Type of prosthesis is the generic description of a prosthesis, e.g. modular cemented stem (hip), patella-femoral joint (knee). U Unicondylar arthroplasty Replacement of one tibial condylar and one femoral condyl, with or without resurfacing of the patella. Unicondylar knee See Unicondylar arthroplasty. replacement Unilateral operation Operation performed on one side only, e.g. left hip. National Joint Registry 117 k NJR Centre Contact Details u . o c . s m National Joint Registry m o c Northgate Information Solutions (UK) Ltd m m . Peoplebuilding 2 w w Peoplebuilding Estate w d t L Maylands Avenue s n o Hemel Hempstead i t a c i Hertfordshire n u HP2 4NW m m o C M & M Telephone 0845 345 9991 y b d Fax 0845 345 9992 e c u Email [email protected] d o r P Website www.njrcentre.org.uk