10Th Annual Report 2013 National Joint Registry for England, Wales and Northern Ireland

HIPS KNEES ANKLES ELBOWS SHOULDERS PROMs

10th Annual Report 2013 National Joint Registry for England, Wales and Northern Ireland

ISSN 1745-1450 (Online) Surgical data to 31 December 2012 Prepared by

The NJR Editorial Board

NJRSC Members Mr Martyn Porter (Chairman) Mick Borroff Professor Paul Gregg Professor Alex MacGregor Mr Keith Tucker

NJR RCC Network Representatives Mr Peter Howard (Chairman) Mr Colin Esler Mr Alun John Mr Matthew Porteous

Orthopaedic Specialists Professor Andy Carr Mr Andy Goldberg

NJR Research Fellows Mr Jeya Palan

Healthcare Quality Improvement Partnership NJR Management Team and NJR Communications Rebecca Beaumont James Thornton Melissa Wright Elaine Young

Northgate Information Solutions (UK) Ltd NJR Centre, IT and data management Olivia Forsyth Anita Mistry Dr Claire Newell Dr Martin Pickford Martin Royall Mike Swanson

University of Bristol NJR Statistical support, analysis and research team Professor Yoav Ben Shlomo Professor Ashley Blom Dr Emma Clark Professor Paul Dieppe Dr Linda Hunt Dr Michèle Smith Professor Jonathan Tobias Kelly Vernon

Pad Creative Ltd (design and production)

This document is available in PDF format for download from the NJR website at www.njrcentre.org.uk

This document is available in PDF format for download from the NJR website at www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Contents

Chairman’s introduction 10

Foreword from the Chairman of the Editorial Board 12

Executive summary 13 Part 1: Annual progress...... 14 Part 2: Clinical activity 2012...... 15 Part 3: Outcomes after joint replacement 2003 to 2012...... 19

Part 1 Annual progress 23 1.1 Introduction 23 1.1.1 The National Joint Registry ...... 24 1.1.2 Management and funding ...... 24 1.1.3 Content of the 10th Annual Report ...... 25

1.2 Data completeness and quality 26 1.2.1 Key indicators...... 27 1.2.2 Operation totals ...... 29

1.3 Work of the NJR Steering Committee and its Sub-committees 33 1.3.1 Introduction...... 34 1.3.2 NJR Steering Committee...... 34 1.3.3 NJR Editorial Board - Mr Martyn Porter, Chairman...... 35 1.3.4 Implant Performance Sub-committee - Mr Keith Tucker, Chairman...... 36 1.3.5 Surgeon Outliers Sub-committee - Professor Paul Gregg, Chairman...... 36 1.3.6 NJR Research Sub-committee - Professor Alex MacGregor, Chairman...... 38 1.3.7 NJR Data Quality Group - Professor Paul Gregg, Chairman...... 39 1.3.8 NJR Regional Clinical Coordinators’ (RCC) Network - Mr Peter Howard, Chairman...... 40

1.4 Highlights 41 1.4.1 Geographic extension of the NJR ...... 42 1.4.2 Extension of the NJR: Elbows and shoulders...... 42 1.4.3 Beyond Compliance: ‘Protecting Patients, Supporting Innovation’...... 42 1.4.4 Publication of consultant-level data...... 44 1.4.5 Patient-focused initiatives ...... 44 1.4.6 NJR Feedback Services: Updates...... 44 1.4.7 Proposed changes to the Minimum Dataset ...... 45 1.4.8 Price Benchmarking...... 45 1.4.9 International developments ...... 46 1.4.10 Patient Reported Outcomes Measures (PROMs)...... 46 1.4.11 NJR Fellowships ...... 47 1.4.12 Orthopaedic Data Evaluation Panel (ODEP) overview...... 47

1.5 Finance 48 1.5.1 Income and expenditure 2012/13 ...... 49 Appendix 1 NJR Steering Committee 2012/13...... 50

www.njrcentre.org.uk 3 Appendix 2 List of papers, publications and research requests using NJR data...... 52 Appendix 3 Additional information on the NJR website...... 55

Part 2 Clinical activity 2012 56 2.1 Introduction 56 2.1.1 Hospitals and treatment centres participating in the NJR...... 57

2.2 Hip replacement procedures 2012 60 2.2.1 Primary total hip replacement procedures (THR) 2012...... 62 2.2.2 Hip revision procedures 2012 ...... 80

2.3 Knee replacement procedures 2012 84 2.3.1 Primary knee replacement procedures 2012...... 86 2.3.2 Knee revision procedures 2012...... 101

2.4 Ankle replacement procedures 2012 103 2.4.1 Primary ankle replacement procedures 2012...... 104 2.4.2 Ankle revision procedures 2012...... 109

2.5 Elbow replacement procedures 2012 111 2.5.1 Primary elbow replacement procedures 2012 (nine months) ...... 112 2.5.2 Elbow revision procedures 2012...... 115 2.5.3 Elbow components used in primary and revision procedures...... 116

2.6 Shoulder replacement procedures 2012 117 2.6.1 Primary shoulder replacement procedures 2012 (nine months) ...... 118 2.6.2 Shoulder components used in primary procedures ...... 121 2.6.3 Shoulder revision procedures 2012 (nine months)...... 121

Part 3 Outcomes after joint replacement 2003 to 2012 123 3.1 Summary of data sources and linkage 123

3.2 Outcomes after primary hip replacement 126 3.2.1 Overview of primary hip surgery...... 128 3.2.2 Revisions after primary hip surgery...... 131 3.2.3 Revisions for different causes after primary hip surgery ...... 136 3.2.4 Revisions after primary hip surgery for the main stem-cup brand combinations...... 140 3.2.5 Revisions after primary hip surgery: Effect of head sizes for polyethylene liners ...... 144 3.2.6 Mortality after primary hip surgery ...... 148 3.2.7 In-depth study: Metal-on-metal hip resurfacing ...... 150 3.2.8 Conclusions...... 158

3.3 Outcomes after primary knee replacement 159 3.3.1 Overview of primary knee surgery ...... 160 3.3.2 Revisions after primary knee surgery...... 165 3.3.3 Revisions for different causes after primary knee surgery...... 169

4 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.3.4 Revisions after primary knee surgery by main brands for TKR and UKR...... 173 3.3.5 Mortality after primary knee surgery...... 179 3.3.6 Conclusions...... 181

3.4 Outcomes after primary ankle replacement 182 3.4.1 Overview of primary ankle surgery...... 183 3.4.2 Revisions after primary ankle surgery...... 184 3.4.3 Mortality after primary ankle surgery...... 184

3.5 PROMs outcomes 186 3.5.1 Background to Patient Reported Outcome Measures (PROMs)...... 187 3.5.2 Data linkage from PROMs to HES to NJR...... 187 3.5.3 PROMs outcomes for primary hip replacements ...... 188 3.5.4 PROMs outcomes for primary knee replacements...... 196 3.5.5 Conclusions...... 201

Part 4 Trust-, Local Health Board- and unit-level activity and outcomes 2012 202 4.1 Introduction 202

4.2 Unit outlier analysis methodology 204

Trust-, Local Health Board- and unit-level data ...... 206

Part 1 tables Table 1.1 Total joint replacement procedures entered into the NJR, 2008/09 to 2012/13, recorded by country in which the procedure took place...... 29 Table 1.2 Total joint replacement procedures entered into the NJR, 2008/09 to 2012/13, recorded by procedure type.30 Table 1.3 Proportion of reported procedures by type of provider, 2008/09 to 2012/13...... 31

Part 1 figures Figure 1.1 Compliance, Consent, and Linkability Rates from 2003 to 2013...... 28 Figure 1.2 Beyond Compliance data flow...... 43

Part 2 tables Table 2.1 Total number of hospitals and treatment centres in England and Wales able to participate in the NJR and the proportion actually participating in 2012 ...... 58 Table 2.2 Number of participating hospitals, according to number of procedures performed during 2012...... 58 Table 2.3 Procedure details, according to type of provider for hip procedures in 2012...... 61 Table 2.4 Patient characteristics for primary hip replacement procedures in 2012, according to procedure type ...... 64 Table 2.5 Age and gender for primary hip replacement patients in 2012...... 65 Table 2.6 Indications for hip primary procedure based on age groups ...... 66 Table 2.7 Surgical technique for primary hip replacement patients in 2012...... 71

www.njrcentre.org.uk 5 Table 2.8 Thromboprophylaxis regime for primary hip replacement patients, prescribed at time of operation...... 73 Table 2.9 Reported untoward intra-operative events for primary hip replacement patients in 2012, according to procedure type...... 73 Table 2.10 Patient characteristics for hip revision procedures in 2012, according to procedure type...... 81 Table 2.11 Indication for surgery for hip revision procedures 2008 to 2012...... 82 Table 2.12 Components removed during hip revision procedures in 2012...... 82 Table 2.13 Components used during single-stage hip revision procedures in 2012...... 83 Table 2.14 Procedure details, according to type of provider for knee procedures in 2012...... 85 Table 2.15 Patient characteristics for primary knee replacement procedures in 2012, according to procedure type ...... 87 Table 2.16 Age and gender for primary knee replacement patients in 2012...... 91 Table 2.17 Characteristics of surgical practice for primary knee replacement procedures in 2012, according to procedure type...... 95 Table 2.18 Thromboprophylaxis regime for primary knee replacement patients, prescribed at time of operation...... 97 Table 2.19 Reported untoward intra-operative events for primary knee replacement patients in 2012, according to procedure type...... 98 Table 2.20 Patient characteristics for knee revision procedures in 2012, according to procedure type ...... 102 Table 2.21 Patient characteristics for primary ankle replacement procedures in 2012...... 105 Table 2.22 Age and gender for primary ankle replacement patients in 2012 ...... 106 Table 2.23 Characteristics of surgical practice for primary ankle replacement procedures in 2012...... 107 Table 2.24 Thromboprophylaxis regime for primary ankle replacement patients, prescribed at time of operation...... 108 Table 2.25 Reported untoward intra-operative events for primary ankle replacement patients in 2012...... 108 Table 2.26 Details for ankle revision procedures in 2012...... 109 Table 2.27 Patient characteristics for ankle revision procedures in 2012...... 110 Table 2.28 Details for replacement primary elbow procedures in 2012 (nine months) ...... 112 Table 2.29 Patient characteristics of primary elbow replacement procedures in 2012 (nine months)...... 113 Table 2.30 Characteristics of surgical practice for primary elbow replacement procedures in 2012 (nine months)...... 114 Table 2.31 Thromboprophylaxis regime for primary elbow replacement patients, prescribed at time of operation...... 115 Table 2.32 Details for elbow revision procedures in 2012 (nine months)...... 116 Table 2.33 Details for primary shoulder procedures in 2012 (nine months)...... 118 Table 2.34 Patient characteristics for primary shoulder procedures in 2012 (nine months)...... 119 Table 2.35 Characteristics of surgical practice for primary shoulder replacement procedures in 2012 (nine months) ...... 120 Table 2.36 Thromboprophylaxis regime for primary shoulder replacement patients, prescribed at time of operation...... 121 Table 2.37 Details for shoulder revision procedures in 2012 (nine months)...... 122

Part 2 figures Figure 2.1 Percentage of participating hospitals by number of procedures per annum, 2004 to 2012...... 59 Figure 2.2 Primary hip procedures by type of provider 2012...... 62 Figure 2.3 Type of primary hip replacement procedures undertaken between 2005 and 2012...... 63 Figure 2.4 Age and gender for primary hip replacement patients in 2012...... 67 Figure 2.5 Age for primary hip replacement patients between 2003 and 2012...... 68 Figure 2.6 ASA grades for primary hip replacement patients between 2003 and 2012...... 69 Figure 2.7 BMI for primary hip replacement patients between 2004 and 2012...... 70 Figure 2.8 Bone cement types for primary hip replacement procedures undertaken between 2004 and 2012...... 72

6 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.9 Top five cemented hip stem brands, trends 2003 to 2012...... 75 Figure 2.10 Top five cemented hip cup brands, trends 2003 to 2012...... 75 Figure 2.11 Top five cementless hip stem brands, trends 2003 to 2012...... 76 Figure 2.12 Top five cementless hip cup brands, trends 2003 to 2012...... 77 Figure 2.13 Top five resurfacing head brands, trends 2003 to 2012...... 78 Figure 2.14 Femoral head size, trends 2003 to 2012...... 79 Figure 2.15 Hip articulation, trends 2003 to 2012...... 79 Figure 2.16 Primary knee procedures by type of provider 2012...... 86 Figure 2.17 Type of primary knee replacement procedures undertaken between 2006 and 2012...... 88 Figure 2.18 Implant constraint for bicondylar primary knee replacement procedures between 2006 and 2012. . . . . 89 Figure 2.19 Bearing type for unicondylar implant used in primary knee replacement procedures undertaken between 2006 and 2012...... 90 Figure 2.20 Age and gender for primary knee replacement patients in 2012...... 92 Figure 2.21 ASA grades for primary knee replacement patients between 2003 and 2012...... 93 Figure 2.22 BMI for primary knee replacement patients between 2004 and 2012...... 94 Figure 2.23 Bone cement types for primary knee replacement procedures undertaken between 2003 and 2012. . . 96 Figure 2.24 Top five total condylar knee brands, trends 2003 to 2012...... 98 Figure 2.25 Top five unicondylar knee brands, trends 2003 to 2012...... 99 Figure 2.26 Top five patello-femoral knee brands, trends 2003 to 2012...... 100 Figure 2.27 Top five fixed hinged knee brands, trends 2003 to 2012...... 101

Part 3 tables Table 3.1 Summary description of datasets used for survivorship analysis...... 124 Table 3.2 Composition of person-level datasets for survivorship analysis...... 125 Table 3.3 Numbers (%) of primary hip replacements by fixation and, within each fixation sub-group, by bearing surface...... 128 Table 3.4 Percentage of primary hip replacements performed each year by type of hip fixation and constraint...... 129 Table 3.5 Distribution of consultant surgeon and unit caseload for each fixation type ...... 130 Table 3.6 Age (in years) at primary hip replacement, by fixation and main bearing surface...... 131 Table 3.7 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI), at different times after the primary operation, for each fixation/bearing surface sub-group...... 132 Table 3.8 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first hip revision. Rates shown are for all revised cases and by fixation and bearing surface...... 137 Table 3.9 Revision rates, expressed as numbers per 1,000 patient-years, for any reason, according to time interval from primary operation...... 139 Table 3.10 Revision rates (95% CI), expressed as numbers per 1,000 patient-years, for each reason, by time interval from primary operation ...... 139 Table 3.11 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at different times after the primary operation, for cup-stem brand combinations with large group sizes (>2,500 or >1,000 in the case of resurfacings) ...... 141 Table 3.12 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at different times after the primary operation for cup-stem brand combinations with large group sizes (>10,000) with further subdivision by main bearing surface (provided the sub-group size >1,000)...... 143 Table 3.13 Kaplan-Meier estimates of the cumulative percentage mortality (95% CI), at different times after primary hip operation, for all cases and by age/gender...... 149 Table L1 Description of comparison groups: number of cases (percentage of total)...... 154 Table L2 Predicted revision rates for 55-year-old males by prosthesis and head size (95% CI)...... 155

www.njrcentre.org.uk 7 Table L3 Predicted revision rates for 55-year-old females by prosthesis and head size (95% CI)...... 155 Table L4 Comparison of most commonly used brands; predicted revision rates for 55-year-old patients (95% CI)...... 156 Table L5 Reasons for revision (95% CI), expressed as incidence per 1,000 patient-years, by articulation and fixation ...... 157 Table 3.14 Numbers and percentages of primary knee replacements by fixation method and bearing type...... 161 Table 3.15 Percentage of primary knee replacements performed each year by method of fixation and, within each fixation group, by bearing type...... 162 Table 3.16 Distribution of consultant surgeon and unit caseload for each fixation type ...... 163 Table 3.17 Age (in years) at primary operation for different types of knee replacement; by fixation and bearing type. . 164 Table 3.18 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at specified times after primary knee replacement, by fixation and bearing type...... 166 Table 3.19 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first knee revision. Rates shown are for all revised cases and by fixation type ...... 170 Table 3.20 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first knee revision. Rates shown are for each fixation/bearing surface sub-group...... 171 Table 3.21 Kaplan-Meier estimated cumulative percentage probability of first revision (95% CI) of a primary total knee replacement by main type of implant brand at the indicated number of years after primary operation). . . 174 Table 3.22 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) of a primary unicompartmental knee replacement by main type of implant brand at the indicated number of years after primary operation...... 175 Table 3.23 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) of a total knee replacement at the indicated number of years after primary operation, by main implant brands and type of fixation and constraint...... 176 Table 3.24 Age and gender distribution of patients undergoing all types of primary knee replacement operations for the period 2003 to 2012...... 179 Table 3.25 Kaplan-Meier estimated cumulative percentage probability (95% CI) of a patient dying at the indicated number of years after a primary knee joint replacement operation (i) by age group and gender and (ii) for all patients ...... 180 Table 3.26 Number of primary ankle operations by year ...... 183 Table 3.27 Number of primary ankles by ankle brand...... 183 Table 3.28 Reasons for ankle revision (not mutually exclusive)...... 184 Table 3.29 Overall outcomes after primary hip surgery ...... 189 Table 3.30 Bias in Q2 completion of EQ-5D Index, EQ-5D Health Scale (VAS), Oxford Hip Score...... 190 Table 3.31 Changes in EQ-5D Index for hip primaries with index scores at both time points...... 190 Table 3.32 Changes in EQ-5D Health Scale Score (VAS) for hip primaries with scores at both time points ...... 192 Table 3.33 Changes in Oxford Hip Score for hip primaries with scores at both time points...... 193 Table 3.34 Overall outcomes after primary knee surgery...... 196 Table 3.35 Bias in Q2 completion of EQ-5D Index, EQ-5D Health Scale (VAS), Oxford Knee Score...... 197 Table 3.36 Changes in EQ-5D Index for knee primaries with index scores at both time points...... 198 Table 3.37 Changes in EQ-5D Health Scale score (VAS) for knee primaries with scores at both time points. . . . . 198 Table 3.38 Changes in Oxford Knee Score for knee primaries with scores at both time points ...... 198

Part 3 figures Figure 3.1 Patients with hip, knee and ankle primary operations within the survivorship data sets...... 125 Figure 3.2 Comparison of cumulative hazard of first revision for cemented hips with different bearing surfaces (with 95% CI)...... 133 Figure 3.3 Comparison of cumulative hazard of first hip revision for uncemented hips with different bearing surfaces (with 95% CI)...... 134

8 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.4 Comparison of cumulative hazard of a first hip revision for hybrid (not including reverse hybrid) hips with different bearing surfaces (with 95% CI)...... 135 Figure 3.5 (i) Comparison of the cumulative hazard for revision for different head sizes for metal-on-polyethylene monobloc cup...... 145 Figure 3.5 (ii) Comparison of the cumulative hazard for revision for different head sizes for metal-on-polyethylene with metal shell/polyethylene liners...... 146 Figure 3.5 (iii) Comparison of the cumulative hazard for revision for different head sizes for ceramic-on-polyethylene with polyethylene monobloc cup...... 147 Figure 3.5 (iv) Comparison of the cumulative hazard for revision for different head sizes for ceramic-on-polyethylene with metal shell/polyethylene liners...... 148 Figure L1 Cumulative hazard of revision after resurfacing by gender (with 95% CI)...... 152 Figure L2 Estimated cumulative incidence of revision for 55-year-old male by prosthesis type...... 153 Figure L3 Estimated cumulative incidence of revision for 55-year-old female by prosthesis type ...... 153 Figure 3.6 Cumulative hazard (x100) of a first revision for different types of primary knee replacement at increasing years after the primary surgery (with 95% CI)...... 168 Figure 3.7 Comparison of the cumulative hazard (x100) of a knee prosthesis first revision for different bearing types at increasing years after the primary surgery when the primary arthroplasty method of fixation was cemented only (with 95% CI)...... 169 Figure 3.8 (i) Histogram to compare the distributions of the EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=80,394). At Q1...... 194 Figure 3.8 (ii) Histogram to compare the distributions of the EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=80,394). At Q2...... 194 Figure 3.9 (i) Histogram to compare the distributions of the Oxford Hip Score between Q1 and Q2 in cases with scores at both time points (n=92,133). At Q1...... 195 Figure 3.9 (ii) Histogram to compare the distributions of the Oxford Hip Score between Q1 and Q2 in cases with scores at both time points (n=92,133). At Q2...... 195 Figure 3.10 (i) Histogram to compare the distributions of EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=84,031). At Q1...... 199 Figure 3.10 (ii) Histogram to compare the distributions of EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=84,031). At Q2...... 199 Figure 3.11 (i) Histogram to compare the distributions of the Oxford Knee Score between Q1 and Q2 in cases with scores at both time points (n=93,353). At Q1...... 200 Figure 3.11 (ii) Histogram to compare the distributions of the Oxford Knee Score between Q1 and Q2 in cases with scores at both time points (n=93,353). At Q2...... 200

Glossary Glossary ...... 232

www.njrcentre.org.uk 9 Chairman’s introduction Laurel Powers-Freeling

It has been another busy year for the National Joint structure, funding model and major activities support Registry (NJR), which continues to develop and our aims. Key developments from this review include mature as the world’s largest orthopaedic registry and the establishment of a new Executive Committee to show leadership in analysis and communication of handle the rapidly expanding day-to-day management insights we glean from our rich pool of data. Normally of the NJR; establishment of a new Medical Advisory my ‘thanks to colleagues’ goes at the end of the Committee intended to provide a forum to gather Chairman’s Introduction, but given the extraordinary views from our medical stakeholders, in particular the efforts and change we have seen this year, I want medical societies; the creation of a new NJR Medical to acknowledge up-front my gratitude to the NJR Director executive post to partially relieve the burden Steering Committee, and in particular to note the carried by volunteers on the Steering Committee; outstanding contributions made by our Vice Chairman, and, we have established a formal Patient Network Professor Paul Gregg, Mr Martyn Porter, Mr Keith to ensure that the voice of the patient is heard in all Tucker, Professor Alex MacGregor and Mr Peter the work we do. In addition, we plan to broaden the Howard. Without their commitment and generosity of membership of the NJR Sub-committees to include time and spirit, much of the work of the NJR would a wider representation of stakeholders than those not have been possible. I would also like to thank our exclusively on the NJR Steering Committee. Work is partners: University of Bristol, led by Professor Ashley ongoing to finalise these arrangements. Blom; Northgate Information Solutions; as well as the NJR management team at the Healthcare Quality In September 2012, we launched our new Annual Improvement Partnership, led by Elaine Young, who Clinical Reports, which were circulated to Trust, receives our special thanks for her dedication and Local Health Board and independent provider hard work. Chief Executives detailing a range of performance information for their hospitals. This was a major step Historically, the data collection and analysis done by forward in providing units with feedback on the data the NJR has focused on hip and knee replacement, held for them by the NJR. and from April 2010 ankle replacement, and has done so only for procedures done in England and Wales. In Another important initiative in which the NJR has April 2012, however, we extended our data collection been an integral member is the establishment of the to include elbow and shoulder joint replacements, initiative Beyond Compliance, which invites implant as well as implementing a two-year pilot study of manufacturers on a voluntary basis to go beyond Patient Reported Outcomes Measures for shoulders, the requirements of the implant CE mark and enter with data collected from patients at six months post- a rigorous, ongoing monitoring process for newly- surgery. As well as new joint types, we have also developed implants. The NJR supports the British extended the NJR geographically with the inclusion Orthopaedic Association and the Medicines and of data from Northern Ireland, and are currently in Healthcare products Regulatory Agency, who are the discussions to incorporate data from the Isle of Man. joint leaders of this programme, by providing data and operational support. We are proud to be a part of this Given the growth and development of the NJR, world-leading development. we took the opportunity in this, our tenth year of operation, to step back and review our purpose In support of better cost-effectiveness of implant and strategic aims, as well as considering how our purchasing by Trusts, the NJR was commissioned by the Department of Health under the QIPP Orthopaedic

10 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Procurement Group (see page 45) to look at pricing and with intentions to make a further appointment data across NHS Trusts and Local Health Boards in this year. England and Wales; 35 units have been involved in • NJR Communications – we have developed a the pilot study to date. By matching actual cost data new communications strategy for 2013/14, which with NJR data, a potential suite of tools will enable has included the recent launch of NJR social media healthcare organisations to analyse and compare platforms, including Twitter, Facebook and our spend and usage of orthopaedic implant products. first eBulletin news service, with more innovations The goal is to develop a model that can be planned for next year. extended nationally. • New Annual Report format – work is ongoing NHS England recently announced through the that will see NJR Annual Reporting data presented ‘Everyone Counts’ initiative that outcome of surgery at online, taking into account a stakeholder survey of individual consultant level would be published by the preferences and views. end of June 2013 for ten clinical areas, one of which is Finally, we will be holding a special celebration of the hip and knee, and which is supported by data from the NJR’s ‘10th Birthday’ this September with a special NJR. At the same time, the NJR has also developed stakeholder reception to launch this Annual Report. a surgeon and hospital profile system that has been used as the publication vehicle. The magnitude and ...... complexity of this exercise for orthopaedics has relied on a huge amount of work and collaboration between While I gave my thanks to colleagues at the beginning the NJR and the British Orthopaedic Association, of this introduction, I would like to close by mentioning with support from the specialist societies, the British the members of the NJRSC who will be stepping- Association for Surgery of the Knee and the British Hip down during 2013/14, which include Professor Paul Society. Despite very tight deadlines, we were able Gregg, Mick Borroff and Professor Alex MacGregor to publish data on the new NJR system, which we – all of whom have been on the Steering Committee intend to develop further in the coming year to provide since its inception. I would also like to thank Andrew surgeons and trusts with the ability to customise Woodhead who has served on the NJRSC since profiles that are made available to the public. 2007, and Dean Sleigh who stepped down in March 2013 and has served on the NJRSC since 2008. Each In June this year, the NJR hosted the 2nd Annual of them can rightly take pride in having built the NJR Congress for the International Society of Arthroplasty and in the importance of the work we do today. Registries (ISAR) in Stratford-upon-Avon, providing us with an opportunity to showcase our work and influence the development of future international cooperation. We look forward to the 3rd ISAR Yours sincerely, Congress in Boston, USA in 2014.

As we move from this year to next, the NJR has a number of new activities and plans that we believe will further support our stakeholders, including:

• Patient Implant Cards – we are undertaking a pilot study with the aim of introducing a patient implant card designed to improve online access to Laurel Powers-Freeling information for patients and also provide feedback to Chairman, National Joint Registry Steering Committee the NJR. • NJR Research Fellows – due to the success of the NJR Research Fellows posts, the programme will continue, with one new Fellow now appointed

www.njrcentre.org.uk 11 Foreword from the Chairman of the Editorial Board

The 2013/14 year marks the 10th anniversary of the For the first time Patient Reported Outcome Measures are National Joint Registry. By 31 March 2013 the NJR reported. This represents a major step forward looking held more than 1.4 million records and nearly 200,000 at other factors apart from revision which may affect submissions were received on hip, knee, ankle, shoulder, the outcome. It is still relatively early days both in terms and elbow replacements. The in-year compliance is over of understanding how this data should be statistically 90% and the overall compliance, since 2003, is 86.8%. analysed and how the findings should be incorporated. It was interesting to note that 55% of joint replacement With such a large database the NJR is clearly having a activity carried out in independent hospitals was funded larger influence on the evidence base. However, it will by the NHS. probably take another ten years before the full effects and benefits of bearing performance are fully appreciated. The previous trend in terms of the uncemented fixation continues and the use of metal-on-polyethylene and The availability of new materials and a better ceramic-on-ceramic bearings predominate. The use of understanding of the biomechanics of joints have resulted the metal-on-metal articulation has declined to very low in many new innovations being available to orthopaedic levels. The knee replacement trends continue very much surgeons. An important role of the NJR has been to as they have done over the lifespan of the registry with maintain a close surveillance on the early adoption of new the posterior cruciate retaining bicondylar cemented knee technologies and our implant performance Scrutiny Group being the procedure of choice. Despite higher observed regularly review clinical publications and results reported revision probability with the unicondylar knees these have by international registries. It is important that this review of still maintained about 8% of the market. other data takes place because the numbers of any new product implanted in one country are often too small for An important change this year in Part Three of the report any definitive decisions on performance to be made within has been the use of the Kaplan-Meier survivorship the first couple of years following the initial implantations. estimates for all information in the tables instead of the Nelson-Aalen cumulative hazard used previously, The planning for the 11th Annual Report is already although the latter technique has been used for the underway and I would like to point out that the format graphs. The details of this are explained in Part Three of next year’s report is likely to be significantly different. itself. It is also important to note that the data has not We are intending to concentrate on providing online and been fully risk adjusted. Instead it has been presented interactive report features. The traditional printed report according to a combination of fixation attributes and in will be a much briefer summary and in this context the the larger groups also with different bearing attributes 10th Annual Report is likely to be last of the larger as we did last year. Instead of considering survival of a report formats. single component, both the stem and cup brand have been considered together and in the larger groups further Once again I would like to thank everybody who has broken down in relation to the bearing used within the contributed to the registry, not least to the patients for brand companies. This allows a more realistic view on allowing their data to be recorded and to all surgeons for the effect of the ‘whole’ hip replacement. However, as entering their data in to the NJR. the data has not been fully risk adjusted care needs to be taken interpreting the results and, as always, I would recommend looking at the upper and lower confidence intervals around the mean to ascertain whether any differences are likely to be statistically significant.

Martyn Porter Chairman, Editorial Board

12 www.njrcentre.org.uk Executive summary Part 1: Annual progress

The 10th Annual Report of the National Joint Registry NJR Editorial Board, and the NJR Data Quality Group. for England, Wales and Northern Ireland is the formal The reports from the respective Chairman of the public report for the period 1 April 2012 to 31 March Implant Performance Sub-committee and the Surgeon 2013 (Part One). Also included are statistics on joint Outlier Sub-committee outline how outlier analysis is replacement activity for the period 1 January to 31 undertaken and include the high level outcomes of the December 2012 (Part Two) and survivorship and monitoring process for 2012/13. more detailed statistical analysis on hip and knee joint replacement surgery using data from 1 April The work of the NJR has continued to expand 2003 to 31 December 2012 (Part Three). Part Four throughout 2012/13. In April 2012, details of shoulder shows indicators for hip and knee joint replacement and elbow joint replacement procedures were procedures by Trust and unit based on the 2012 added to the data collection, Northern Ireland joined calendar year. the NJR in February 2013, and the NJR began the next phase of its PROMs study: the follow up, at The NJR began collecting data on hip and knee three years, of approximately 43,000 hip and knee replacement operations on 1 April 2003. Data replacement patients. The report also outlines two collection on ankle replacements began on 1 April other major developments being supported by the 2010 and on shoulder and elbow replacements on 1 NJR: Price Benchmarking and Beyond Compliance. April 2012. Data collection in Northern Ireland began Price Benchmarking is sponsored by the Department in February 2013. The total number of procedures of Health and is currently in the second phase of its recorded in the NJR exceeded 1.4 million records by pilot. The service, when fully implemented, will provide 31 March 2013, with 2012/13 having the highest ever a comparison of implant prices across NHS England annual number of submissions at 196,403. and NHS Wales. Beyond Compliance is a voluntary, post-market surveillance process that goes beyond the The NJR uses rates of compliance (case normal, regulatory requirements for the introduction, ascertainment), patient consent, and linkability (the to market, of new implants. Using a central data ability to link a patient’s primary procedure to a revision repository consisting of data from numerous sources, procedure) as its key data quality indicators. Details of new implants can be more closely followed up with the how these are calculated are included in the report. At intention of them being introduced to market in a more 91%, 2012/13 saw the highest annual rate of consent controlled fashion than has been possible previously. recorded, whilst linkability remained constant at 96%. The NJR has continued to increase its presence The NJR Steering Committee and its sub-committees internationally and is a full member of both the are vital to the running and development of the NJR. International Society of Arthroplasty Registers (ISAR) A key focus of the steering committee throughout the and the International Consortium of Orthopaedic last year has been to review the NJR’s priorities and its Registers (ICOR). The NJR recently hosted ISAR’s 2nd governance and operating model, in recognition of the Annual Congress (June 2013). great increase in scope and responsibilities of the NJR since its formation in November 2002.

The report highlights the ongoing work of the Implant Performance Sub-committee, the Surgeon Outlier Sub- committee, the NJR Research Sub-committee, the

14 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Part 2: Clinical activity 2012

Part Two of the NJR 10th Annual Report includes data The proportion of patients within age ranges has on clinical activity – volumes and surgical technique, in remained virtually the same in the 10 years that the relation to hip, knee and ankle procedures carried out NJR has been recording data with about one third of in England and Wales between 1 January 2012 and patients being in the age range 70 to 79 and just over 31 December 2012. It also includes information on 1% being less than 40 years at the time of surgery. elbow and shoulder prostheses between 1 April 2012 The average body mass index (BMI) increased very and 31 December 2012. To be included in the report slightly to 28.7 compared to 28.6 in 2011. all procedures must have been entered into the NJR by 28 February 2013. The ASA distribution is identical to last year with only 15% being ASA grade 1 and the majority of During 2012 there were 413 orthopaedic units that patients being ASA grade 2. However, there were were active and this included 240 NHS hospitals and important differences in patients ASA grades between 163 independent units. The remaining units were those treated within NHS hospitals compared to independent sector treatment centres. A total of 96% independent hospitals. For example, 21% of patients of units submitted at least one procedure carried out treated in NHS hospitals had an ASA grade 3 within the calendar year 2012 to the NJR. The average compared to just 7% in independent hospitals and number of hip replacements submitted per unit was patients with ASA grade 1 (fittest patients) were 218 and for knee replacement 235. just 11% in the NHS and almost double (21%) in independent hospitals. These data would suggest Just over 1% of activity in NHS hospitals is that a much fitter cohort of patients are treated in independently funded but in independent hospitals the independent hospitals which is important when majority of work (55%) is now funded by the NHS. comparing outcomes between these two sectors.

Hip replacement procedures As previously noted, both the age and gender of A total of 86,488 hip procedures were recorded on the the patient have strong associations with the type NJR in 2012 which represented a 7.5% increase from of replacement carried out. For example, in female last year. Of these 76,448 were primary and 10,040 patients less than 50 years of age about 68% were were revision procedures. The ‘revision’ burden now cementless replacements compared to just 21% in stands at 12% of total hip activity compared to 11% in the age range between 80 and 89. 2011. In terms of fixation of hip replacement in 2012 The most common diagnosis was osteoarthritis but the trend was very similar to 2011 where cementless again the age of the patient had a strong influence fixation predominates at 43% with a slight reduction in on the diagnosis and in young patients, as expected, cemented hip replacement to 33% (compared to 36% there was a much higher incidence of patients with last year) and a slight increase in the use of hybrid a diagnosis of avascular necrosis and hip dysplasia fixation at 20% (compared to 18% last year). The compared to the more elderly cohort of patients. use of hip resurfacing has now fallen to 1% and large head metal-on-metal replacements to 2%. Patient The surgical techniques were very similar to those characteristics have remained fairly constant with an recorded in 2011, the lateral position being used in average age of 68.7 years and a predominance of 93% of cases and a slight increase in the use of the female patients (60%). posterior approach to 61% compared to 59% last year.

www.njrcentre.org.uk 15 Minimally-invasive surgery (MIS) was recorded as There were 10,040 hip revision procedures recorded being used in 5% and image-guided surgery was used in 2012 of which 88% were single-stage revisions. very infrequently in just 0.3% of cases. Aseptic loosening was the most commonly recorded indication for revision surgery in 40%, infection in 12% In terms of thromboprophylaxis a multi-modal that increased to 13% for adverse soft tissue reactions regime was commonly used with low molecular (1,330 cases). Patients undergoing revision surgery are weight Heparin being used in 73% of cases and TED less fit than patients undergoing primary replacement stockings in 66%. with one third of patients being graded as ASA grade 3 for staged revision surgery. NHS hospitals incurred In 2012, 146 brands of femoral stem, 101 brands of most of the burden of revision operations – 83% acetabular cup and eight brands of resurfacing cup compared to independent hospitals 15%. This differs were used. Using the Orthopaedic Data Evaluation to the activity of primary surgery where NHS hospitals Panel (ODEP) ratings only about half of all femoral carried out 69% of primary surgery compared to 27% and acetabular components had an ODEP rating in the independent sector. When the revision is carried in primary arthroplasty. However, the components out in a single operation both the cup and stem are used with the full 10A benchmark rating was 88% for removed in 45% of cases, the acetabular component cemented stems, 69% for cementless stems, 34% in only in 30% and the femoral stem only in 14%. The cemented cups, 3% for cementless cups and 63% for revision components used during single-stage hip resurfacing cups. revision is approximately equal between cemented In terms of brand choice the Exeter V40 stem was and cementless (28% versus 29%) whereas on the most commonly used in 66% of cemented hips and acetabular side cementless components are used the contemporary cup used in about 34% of cases. A three times more frequently (58% versus 18%) than rapid rise of the Exeter rim fit cup introduced in 2010 cemented components. was observed. This now has a market share of about 12%, obtained in a very short time period. Knee replacement procedures In 2012, 90,842 knee replacement procedures were In regard to cementless fixation the Corail stem has entered into the NJR representing an increase of 7.3% maintained its position as the most commonly used compared to 2011. In these procedures 84,833 were stem in about 46% of cases and the Pinnacle socket primaries and 6,009 were revision procedures. The in 33%. The use of resurfacing in 2012 fell to 1,075 revision burden of knee replacement is just over 6.5% compared to 1,883 in 2011. The Birmingham hip which is just over half of the revision burden of hip resurfacing is the most popular implant used. replacements. The type of knee replacements used has remained remarkably constant over the last three The preference to use large diameter heads to to four years with 86% being cemented total knee improve stability continues with about 30% of femoral replacements, 8% being unicondylar replacements, heads being 36 millimetres, another 30% being 32 3% total knee replacements not using cement and millimetres and nearly 40% being 28 millimetres. Head only about 1% being patello-femoral replacements. sizes over 36 millimetres are now only rarely used as is the smaller head size of 22.25 millimetres. In terms of surgical technique a medial parapatellar approach was used in 93% of procedures. MIS In terms of bearing combinations the use of metal- was used in just 2% of total knee replacements but on-polyethylene (this includes standard and cross link much more frequently in unicondylar replacements polyethylene) remains the most common selection in (46%). The patella was resurfaced in 38% of primary just under 60% of cases followed by a ceramic-on- cemented knee replacements but in just 7% of ceramic articulation. There is a slight increase in the cementless total knee replacements. As in hip use of ceramic-on-polyethylene and the use of metal- replacements thromboprophylaxis tended to be multi- on-metal articulation continues to fall. modal with low molecular weight Heparin being used in 72% of cases and TED stockings in 70%. The PFC

16 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Sigma was the most commonly used total condylar less reliance on calf or foot compression compared to knee replacement being used in just under 37% of hip and knee replacements. cases. For the unicondylar replacement there was a slight decrease in the use of the Oxford which had just The main implant used was the DePuy Mobility over 60% of the market share. in 52% followed by the Zenith ankle replacement manufactured by Corin in 20%. A total of 6,009 knee revision procedures were reported in 2012 representing an increase of 17% Of the 50 ankle revision procedures most, 46, were compared to 2011. The vast majority of these carried out on the NHS. The main indicators for revision (78%) were single-stage revision operations. As were undiagnosed pain (36%), suspicion of infection in hip revisions, patients undergoing revision knee (26%), aseptic loosening of the tibial component (18%) replacements tended to be less fit than patients and aseptic loosening of the talar component (16%). undergoing primary replacement and 85% of the revisions were carried out within NHS hospitals Elbow replacement procedures compared to 69% of primaries indicating that it is the This is the first year that we have reported on elbow NHS that picks up a disproportionate amount of the replacement and shoulder replacement surgery with knee revision burden. Indications for revision surgery data collection starting on 1 April 2012 and therefore were recorded as aseptic loosening in 32% of cases the reporting period is nine months rather than the full and infection in 22%. calendar year.

Ankle replacement procedures There were 288 elbow replacement procedures which include 214 primary and 74 revision procedures. This is only the second full year of recording ankle Many of the revision procedures clearly relate to replacements with data collection starting on 1 April previous periods of implantation. 89% of the primaries 2010. The compliance has improved from 64% last were total replacements and 9% were radial head year to 77%. replacements. The indications for surgery were more Of the ankle replacements recorded for the calendar variable compared to other procedures in this report year 2012 there were 590 ankle replacements including with osteoarthritis only comprising 32%, inflammatory 540 primaries and 50 revisions. The majority, 86%, arthritis 33% and trauma or trauma sequelae were funded by the NHS and the patient characteristics comprising a further 37%. The average age of patients in terms of ASA and BMI were not dissimilar to that of was 65 (female) and 67 (male). 89% were treated in patients undergoing hip and knee replacement. Unlike the NHS. hip replacements male gender was recorded in 58% The posterior approach was used in 89% of cases and of cases and osteoarthritis was the most common humeral bone graft in 16%. 59% had low molecular diagnosis in 84% and inflammatory arthritis in 13%. weight Heparin and 54% TED stockings. Additional information is recorded on ankles including tibia-hindfoot alignment of 46% being in neutral position Of the 74 revision procedures, 50% were carried and the range of ankle dorsiflexion, 42% having out for aseptic loosening, 18% for infection, 12% for dorsiflexion between five and 20 degrees and 54% instability and 19% for peri-prosthetic fracture. The having plantarflexion between five and 15 degrees. Coonrad Morrey prosthesis manufactured by Zimmer had 45% of the market share. In terms of surgical technique the anterior approach was the most common in 97% and in 11% Achilles Shoulder replacement procedures tendon lengthening was carried out. Bone grafts were used in 15%. Shoulder replacements are again reported for nine months of the calendar year 2012 for reasons For prophylaxis low molecular weight Heparin was explained under the elbow section. used in 80%. 56% had TED stockings but there was

www.njrcentre.org.uk 17 A total of 2,225 shoulder replacement procedures were recorded including 1,968 primaries and 257 revisions. Of the primary procedures there was quite a diverse mix of prostheses with 27% being primary total replacements, 15% hemi-arthroplasty, 6% resurfacing, 22% hemi-resurfacing and 30%, or 597, reverse prostheses. Osteoarthritis was the primary diagnosis in 61% of cases but 24% of patients were reported to have cuff tear arthropathy. The mean age of female patients (73.2 years) was greater than that of male patients (68.8 years) and 72% of patients were female.

The most common surgical approach was the Delta pectoral approach in 75%. Humeral bone graft was used in 12%. A tenotomy of the long head of biceps was carried out in 45% of patients and the rotator cuff condition was described as normal in just 43%. 57% of patients received low molecular weight Heparin.

Of the 257 revision procedures, the indications for surgery were variable; just 14% for aseptic loosening, 30% were conversion of hemi to total and 25% were revision for cuff insufficiency.

18 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Part 3: Outcomes after joint replacement 2003 to 2012

Part Three of the 10th Annual Report describes the metal resurfacing peaked at 10.8% in 2006 and was clinical outcomes represented by survivorship data just 1.3% in 2012. and mortality in relation to hip and knee replacements carried out in England and Wales between 1 April The unadjusted Kaplan-Meier survivorship estimates 2003 and 31 December 2012. This represents nearly are presented for each of these fixation and bearing ten years of survivorship information. combinations within the report. The lowest revision rate at nine years was that of the ceramic-on-polyethylene It also includes for the first time pre- and post- cemented hip replacement with a cumulative probability operative Patient Reported Outcome Measures of first revision of just 1.84% (95% CI of 1.51% (PROMs) collected since April 2009. to 2.24%). Considering all bearing combinations overall, cemented hip replacement had a cumulative The National Joint Registry contains over 1.4 million percentage revision probability of 2.71% (2.57% to operations but because of exclusions in relation to 2.87%) at nine years after operation compared to patient identifiers, linkage issues and other causes, the 6.71% (6.40% to 7.05%) with uncemented fixation, survivorship analysis was based on 539,372 primary once again the ceramic-on-polyethylene bearing had hips with 11,780 linked first revisions, 589,028 primary the lowest revision rate within the uncemented fixation knees and 11,666 linked first revisions and 1,417 group. The hybrid and reversed hybrid groups were ankles with 9 linked first revisions. somewhere between the cemented and uncemented fixation groups with a nine-year revision probability of This year we have used Kaplan-Meier estimates and 3.42% (3.10% to 3.76%) for hybrids and 3.37% (2.52% the cumulative chance of revision of first implant or to 4.33%) for reverse hybrids. death of patient but we have continued to use Nelson- Aalen cumulative hazard estimates for the graphs. When a metal-on-metal articulation was used with Details of these statistical methodologies are explained cementless fixation the nine-year revision probability in the appendices to this report. of a first implant was 17.66% (15.93% to 19.56%). This was similar to the metal-on-metal resurfacing Hip replacement procedures which had a revision probability of 12.31% (11.62% to The data describes the number of patients and 13.04%) at nine years. percentages of implant fixation combinations with those of bearing surface combinations and illustrates The number of revisions by indication for revision are the change of usage of both fixation and bearing presented per 1,000 patient-years for fixation and between 2003 and 2012 inclusive. This shows bearing, for example for dislocation/subluxation there interesting changes over the years, for example in were 0.89 (0.83-0.96) revisions per 1,000 patient- 2003, 55.3% of all hip replacements were metal-on- years with cemented fixation but this was just 0.43 polyethylene cemented replacements but in 2012 this (0.34 to 0.53) when metal-on-metal resurfacing was has nearly halved to just 28.6%. In comparison the used. Soft tissue reaction was recorded in 0.02 metal-on-polyethylene uncemented hip replacement (0.01 to 0.05) of cemented metal-on-polyethylene was used in just 6.2% of cases in 2003 and this has replacements compared to 3.55 (3.05 to 4.13) of almost trebled to 17.8% in 2012. The ceramic-on- metal-on-metal resurfacing and 5.50 (4.90 to 6.20) of ceramic uncemented hip replacement has also shown metal-on-metal uncemented replacement. The revision a rapid increase from just 3.5% in 2003 to a nearly six rates following surgery were not linear, for example fold increase to 19.2% in 2012. The use of metal-on- the revision rate for 1,000 patient-years within one

www.njrcentre.org.uk 19 year of surgery was 7.71 (7.47 to 7.96) but this fell to The cumulative probability of death at different times 4.60 (4.45 to 4.75) between one and three years from after the primary surgery is presented broken down primary operation. by gender and age group. The risk of death within 90 days of surgery overall was 0.51% (0.50% to 0.53%). The Kaplan-Meier cumulative probabilities of a first In female patients this ranged from 0.21% (0.16% prosthesis revision are listed for a large number of cup to 0.27%) in patients less than 55 years to 1.31% stem brand combinations. For example, the cemented (1.22% to 1.41%) in patients 80 years and above and Exeter V40 – Elite Plus Ogee socket combination had at nine years after primary surgery the cumulative a revision probability after nine years of 1.57% (1.30% probability of death in male patients less than 55 years to 1.90%) and the uncemented Furlong HAC stem – was 4.39% (3.88% to 4.96%) compared to 65.67% CSF had a cumulative probability of first revision nine (63.62% to 67.72%) for patients aged 80 and above. years after surgery of 2.97% (2.59% to 3.40%). One of the in-depth studies, that of metal-on-metal For the cup stem brand combinations with over hip resurfacing was published in The Lancet in 10,000 cases the brand combinations were further October 2012. In this paper the effects of age, gender broken down to examine the effect of bearing surface, and femoral head size are considered and the best for example, the Exeter V40 Contemporary metal- outcomes were observed in male patients aged 55 on-polyethylene bearing had a nine-year first revision years with a head size of 54 millimetres where the probability of 2.33% (2.00% to 2.70%) but when revision rate at seven years was 2.47% (1.90% the ceramic-on-polyethylene bearing was used with to 3.20%). the same brands the cumulative chance of revision after nine years was 1.87% (1.20% to 2.90%). The Knee replacement procedures Corail Pinnacle group was interesting as the effect of In comparison to hip replacements, the changes in the bearing could be clearly seen. Eight years post fixation, constraint and type of knee replacement has surgery, the chance of first revision was 5.89% (5.11% not varied as much over the observation period 2003 to 6.79%). However, within this group with a Corail to 2012. The use of cemented total replacements Pinnacle combination, if a ceramic-on-polyethylene increased from 81.5% in 2003 to 86.9% in 2012 bearing was used the probability of first revision was and within cemented groups the unconstrained just 1.84% (1.34% to 2.51%) whereas when a metal- (posterior-cruciate retaining) fixed knee increased on-metal bearing was used the cumulative probability from 53.3% in 2003 to 59.8% in 2012. Despite some was 11.10% (9.06% to 13.58%). concerns with higher revision rates seen amongst With resurfacing clear differences could be seen unicondylar replacements, the use of these remains between the brands. Eight years after the primary fairly static between 8% and 9% and similarly for surgery, the probability of first revision with the patello-femoral replacements (which are relatively low Birmingham hip (BHR) resurfacing was 6.61% (6.12% volume), the use of these ranged from just 1.0% in to 7.13%) compared to 29.69% (27.03% to 32.55%) 2003 to 1.3 % in 2012. when the ASR was used. It was reassuring that the most commonly used Cumulative revision probabilities were also considered replacement, the all cemented total replacement as a function of head size and head material when unconstrained and fixed, had a low cumulative used with a polyethylene liner. This allowed multiple probability of first revision of just 2.90% (95% CI of comparisons, for example, when a metal head 2.77% to 3.04%) at nine years. The cemented mobile was used with a polyethylene liner contained in a constrained condylar knees had a revision probability metal shell, the revision probabilities were higher for of 4.48% (2.87% to 6.96%) at nine years. With the larger 44 millimetre head compared to the 28 uncemented and hybrid replacements the posterior- millimetre head. stabilised knees had a revision probability of 6.70% (5.15% to 8.70%) at nine years. As previously noted

20 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

unicondylar knees had a high revision probability of probability of death was 3.55% (2.88% to 4.37%) at 11.57% (10.92% to 12.26%) at nine years and the nine years compared to 51.00% (49.55% to 52.46%) patello-femoral replacement had a revision probability of in patients 80 years or older. 16.11% (14.09% to 18.39%) at the same time period. Ankle replacement procedures The stated reasons for revision by fixation are described Because of the more recent inclusion of ankles and as revision rates per 1,000 patient-years. The revision relatively small numbers, limited data are presented rate was 1.10 (1.06 to 1.15) revisions for infection and on ankle replacements. There were 1,417 primary for component size mismatch 0.73 (0.70 to 0.77). There operations in a population with a median age of 67 was a further breakdown of reasons for revision by years, 57.2% of patients were male. Nearly all ankle fixation and degree of constraint. replacements (99.6%) were cemented. The Mobility The revision probabilities are described up to nine brand was used in 56.5% of cases. There were only years following surgery for the main brands; compared nine revisions recorded. Mortality was relatively low to hip replacement there was less variation in revision following ankle surgery with a 90-day cumulative probability. For example, the PFC Sigma bicondylar mortality of 0.14% (0.04% to 0.58%) and a two-year knee had a cumulative revision probability of 2.45% cumulative probability of 1.40% (0.69% to 2.83%). (2.30% to 2.61%) at nine years and the Genesis 2 2.43% (2.11% to 2.79%). Some knees performed less Patient Reported Outcome Measures (PROMs) well including the Kinemax with a revision probability For the first time the NJR reports on PROMs. PROMs of 4.30% (3.81% to 4.86%) at nine years and the data, which is part of an NHS-funded initiative, is Rotaglide Plus which had a revision probability of collected separately from the NJR. This collection 5.45% (4.19% to 7.07%). When patello-femoral joints has only been in action since April 2009 and relates were considered the seven-year revision probability to a pre-operative questionnaire and a follow-up of the Avon knee was 10.94% (9.49% to 12.60%) but questionnaire about six months following surgery. for the Journey PFJ Oxinium it was 17.52% (12.98% Data includes EQ-5D Index, EQ-5D Health Scale and to 23.41%). There was also variation with unicondylar Oxford Hip/Knee scores. knees, for example the Zimmer unicompartmental knee had a revision probability of 5.27% (4.15% to The original PROMs data file had 445,134 entries 6.69%) at nine years compared to the Preservation but for a variety of reasons a substantial proportion which had a revision probability of 17.31% (14.43% of these entries could not be matched to NJR data to 20.68%). because of data quality issues, duplication and incomplete data. When considering the type of constraint within the brand class there was generally low variation but Of the 124,136 PROMs entries linked to primary hip there were some differences, for example within the operations in the NJR, 99.7% of questionnaires were Nexgen brand the cumulative revision probability completed pre-operatively but only 75.6% post- with a cemented unconstrained fixed tibial insert was operatively. For hip replacement surgery 85.6% of 2.44% (2.03% to 2.93%) at nine years compared to patients described themselves as being much better 3.62% (3.20% to 4.41%) when the Nexgen posterior- compared to before the operation, 10% reported stabilised fixed insert was used. having had a wound problem, 5.8% bleeding since the operation, 7.3% recorded they had been re-admitted The cumulative probability of patient death at different to hospital since the operation and 2.1% declared they times after knee replacement had a similar pattern had had further surgery following the original operation. to that of hip replacement. For all cases the 90-day mortality probability was 0.36% (0.34% to 0.39%). In relation to change in EQ-5D, the median change At nine years this increased to 22.27% (21.95% to following surgery was 0.380 (IQR 0.175 to 0.694) 22.59%). In female patients less than 55 years the and for the patients treated with a cemented metal-

www.njrcentre.org.uk 21 on-polyethylene hip replacement the median change was 0.413 (0.159 to 0.694) and this compared to metal-on-metal resurfacing where the median change was 0.309 (0.165 to 0.484). However, the median pre- operative score for resurfacing was higher (0.587) for metal-on-metal compared to metal-on-polyethylene cemented (0.293).

The overall median change in EQ-5D Health Scale score (VAS) was 9 (IQR -2 to 20). This tended to be lower for all cemented where the median was 6 (-5 to 20) compared to uncemented where the median was 10 (-1 to 22).

The median change in Oxford hip score was 21 (IQR 14 to 28) for all cases and the variation between fixation types of bearings was very low.

For knee replacement surgery 70.8% of patients described themselves as being much better compared to how they were before the operation, 13.4% described having had a wound problem, 7.9% bleeding, 9.6% having been re-admitted and 3.3% having further surgery.

The median change in EQ-5D Index for all knee primaries (with scores at both time points) was 0.275 (IQR 0.069 to 0.568). The median change was lower for patello-femoral replacement (0.105). The median change in EQ-5D Healthscale score (VAS) was 3 (-7 to 15) for all cases but 0 for patello-femoral replacement (-10 to 15). The median change in Oxford knee score for all cases was 16 (9 to 22) but again was lower for patello-femoral joints at 10 (4 to 18).

22 www.njrcentre.org.uk Part 1 Annual progress

1.1 Introduction 1.1.1 The National Joint Registry 6. Support suppliers in the routine post- market surveillance of implants and provide The National Joint Registry for England, Wales and information to clinicians, patients, hospital Northern Ireland has collected hip and knee joint management and the regulatory authorities replacement data since 2003. The NJR has recorded In July 2012, the NJR Steering Committee, along with data on ankle replacement surgery since April 2010 invited stakeholders, held a one-day workshop to and from April 2012, data on shoulder and elbow define the key activities necessary to support those joint replacement surgery. Northern Ireland has been aims and their relative priorities. submitting data since February 2013. The NJR is currently the largest register of its kind in the world, 1.1.2 Management and funding with over 1.4 million recorded procedures. The NJR is managed by the Healthcare Quality The NJR’s purpose and aims are set out in its Improvement Partnership (HQIP) under a contract with strategic plan and are summarised below: NHS England as part of the delivery of the National Clinical Audit and Patient Outcomes Programme Mission statement: (NCAPOP). HQIP support the work of the NJR ‘The purpose of the National Joint Registry Steering Committee and all its sub-committees. The for England, Wales and Northern Ireland is to NJR Steering Committee, which met four times in collect high quality and relevant data about joint 2012/13, is responsible for overseeing the strategic replacement surgery in order to provide an early direction and running of the NJR. Last year its status warning of issues relating to patient safety. changed from that of an advisory non-departmental public body (ANDPB) to Departmental Expert In a continuous drive to improve the quality Committee. Its members are currently appointed by of outcomes and ensure the quality and cost the Department of Health (DH) Appointments Team effectiveness of joint replacement surgery, the following a formal recruitment process and the current NJR will monitor and report on outcomes, and list of members and their declarations are listed support and enable related research.’ in Appendix 1. The Steering Committee has a lay Strategic goals: Chairman, Ms Laurel Powers-Freeling. 1. Monitor in real time the outcomes achieved There are currently six sub-committees which support by brand of prosthesis, hospital and surgeon, the work of the NJR Steering Committee: and highlight where these fall below an expected performance in order to allow prompt • The Editorial Board Chairman, Mr Martyn Porter investigation and to support follow-up action

2. Inform patients, clinicians, providers and • The Implant Performance Sub-committee Chairman, Mr Keith Tucker commissioners of healthcare, regulators and implant suppliers of the outcomes • The Surgeon Outliers Sub-committee achieved in joint replacement surgery Chairman, Professor Paul Gregg 3. Evidence variations in outcome achieved across • The Research Sub-committee surgical practice in order to inform best practice Chairman, Professor Alex Macgregor 4. Enhance patient awareness of joint • The Regional Clinical Coordinators’ (RCC) replacement outcomes to better inform Network patient choice and patients’ quality of Chairman, Mr Peter Howard experience through engagement with • The Data Quality Group patients and patient organisations Chairman, Professor Paul Gregg 5. Support evidence-based purchasing of joint replacement implants for healthcare providers to support quality and cost effectiveness

24 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

In early 2013, to reflect its change in status, 1.1.3 Content of the 10th the Steering Committee started a review of the NJR’s governance and committee structure and Annual Report composition. The outcomes of the review are included The format of this report is similar to the 9th in Section 1.3.2 below. Annual Report:

The NJR services are delivered under three • Part One is a general outline of the work of the separate contracts: NJR for the financial year 1 April 2012 to 31 March 2013. In addition to summary statistics relating • The NJR Centre, managed and staffed by to the NJR’s key data quality indicators, this Northgate Information Solutions (UK) Ltd. section includes a summary of the major activities Northgate is responsible for the management undertaken by the NJR and individual reports on the and development of the NJR’s IT infrastructure, work of the sub-committees. software applications, data management and reporting services. This work is complemented by • Part Two is a description of joint replacement activity the NJR Service Desk, a team who provide day- reported to the NJR as having been carried out in the to-day information and support to stakeholders, calendar year 1 January to 31 December 2012. and the NJR Regional Coordinators (RC), an • Part Three provides an analysis of survivorship eight-strong team providing on-site support to of hip, knee and ankle replacement procedures orthopaedic units in hospitals. carried out between 1 April 2003 and 31 December • The NJR statistical support, analysis and research 2012. Data from the Hospital Episodes Statistics team, based at the University of Bristol. The team (HES) service, Patient Episode Database for is responsible for the delivery of statistical analyses Wales (PEDW), and the English Patient Reported of NJR data and data from other sources, and for Outcomes Measures (PROMs) programme are also developing the statistical methodologies for the included in the analysis. identification of potential outlier performance. • Part Four provides a series of reports about Their role also includes ad hoc data analyses, clinical activity and outcomes at both Trust and in addition to those included in the NJR Annual unit level. Report, that are central to the work of the Implant Performance Group and the Surgeon Outliers Sub-committee. • NJR Communications, managed by HQIP. This is a programme of stakeholder and multi-media communication to support the delivery of the strategic plan including the publication of the Public and Patient Guide to the NJR Annual Report. The HQIP communications team also support the Editorial Board and Regional Clinical Coordinators’ Network. Currently the NJR is funded through a levy raised on the sale of hip, knee, ankle, elbow, and shoulder implants. HQIP manages the levy payment collection and holds the NJR budget on behalf of the Steering Committee. The funding model is being reviewed and is likely to change in financial year 2013/14.

www.njrcentre.org.uk 25 Part 1 1.2 Data completeness and quality National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

1.2.1 Key indicators likely to be minor variations between the two because of coding differences. This comparison does not Whilst NHS hospitals in England and Wales have include privately funded procedures that take place in always been ‘expected’ to submit data to the NJR, it the independent sector in England and Wales as this has always been mandatory for independent sector data is not submitted to either HES or PEDW. The units in England and Wales since the registry started. overall compliance rate from 1 April 2003 to 31 However, the Standard NHS Contract for Acute March 2013 was 86.8%. Services was amended in April 2011 (Section 12.1.2) and now states that all providers shall participate in audits relevant to the service they provide within NCAPOP, of which the NJR is part. The submission Why are there sometimes more procedures of complete data to the NJR is, therefore, now reported than levies collected, leading to mandatory for all NHS Trusts and Foundation Trusts compliance rates greater than 100%? within England. The Welsh Government has agreed • Bulk buying of implants in one year or at that the NJR is mandatory for all NHS Wales hospitals different times in the year and the Northern Ireland Health and Social Care Board has written NJR data entry into NHS Trust contracts, • Not all levies collected by the implant suppliers this includes all NHS-funded procedures. • Not all revision procedures use a leviable component Performance against the three indicators of data quality (compliance, consent, and linkability) has continued to improve year on year, although the provision of continual support to orthopaedic units is Consent required to maintain and improve performance levels. The consent rate compares the number of records These figures are available throughout the year from submitted where the patient has agreed to their NJR StatsOnline on the NJR website. personal data being stored on the NJR database with the number of procedures recorded on the NJR.2 Compliance It is a requirement in England, Wales, and Northern The compliance rate is the proportion of procedure Ireland that patients ‘opt in’ to have their personal records submitted to the NJR compared with the data held by the NJR. Patient details are essential for levy returns for the number of implants sold.1 It is linking patient procedures in order to monitor joint impossible to establish a one to one link between replacement procedure outcomes and it follows that, a single levy and the use of the implant and this without high rates of patient consent, the NJR will not comparison is subject to a number of factors, such achieve its goals. Patients are known to rarely decline as variation in the procurement cycle throughout the consent, and a number of units regularly achieve 100% year. It is often the case that more procedures are consent rates. Failures to record positive consent are reported than levies are collected, leading to a positive usually due to the lack of robust processes in hospitals (>100%) compliance rate, followed by periods where which ensure that the completed consent form is there are more levies raised than procedures reported. available to the person entering the procedure details. For individual NHS Trusts, compliance can also be The consent rate for 2012/13 was 91.0%. measured against data held in the Hospital Episodes Statistics (HES) service and the Patient Episode Database for Wales (PEDW) service, though there are

1 For compliance analysis only, the number of procedures excludes the following procedures: re-operations other than revision; stage 1 of a 2-stage revision; excision arthroplasty; amputation; and conversion to arthrodesis. These are excluded because they do not include the implantation of a component attracting the levy. 2 Personal information includes NHS number, surname, date of birth and postcode.

www.njrcentre.org.uk 27 Linkability Low rates of linkability adversely affect the ability of the NJR to monitor clinical and implant performance. The linkability rate compares the number of records submitted with the patient’s NHS number with the Where the NHS number is missing, tracing is attempted number of procedures recorded on the NJR. The using the NHS Demographics Batch Service. This relies NHS number is required to link all primary and revision on the patient’s name, date of birth and postcode being procedures relating to a single patient.3 correctly entered. The linkability rate for 2012/13 was 95.6%. The ability to link all operations relating to a single patient is vital in determining clinical outcomes. Operations are linked using the patient’s NHS number.

Figure 1.1 Compliance, Consent, and Linkability Rates from 2003 to 2013. Source: Procedures entered into the NJR 1 April 2003 to 31 March 2013 and levy submissions to NJR by implant suppliers.

120

100

60 %

20 © National Joint Registry 2013

0 Financial year 2003/042004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 Compliance 43.4 68.3 81.6 80.6 95.3 91.6 114.3 106.8 90.3 91.0 Consent 63.3 64.3 74.0 81.2 84.4 87.6 87.9 89.0 90.5 91.0 Linkability 57.2 59.6 69.6 78.0 91.8 94.7 95.3 95.8 96.0 95.6

Compliance Consent Linkability

3 NJR data is submitted for NHS number tracing and the ‘linkability’ figure includes NHS numbers that were traced subsequent to the operation details being submitted to the NJR.

28 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

1.2.2 Operation totals Wales each year from 2008/09 to 2012/13. As for the previous five years, the number of knee replacement By 31 March 2013, 1,456,756 hip, knee, ankle, elbow, procedures (98,288) exceeded the number of hip and shoulder replacement procedures had been replacement procedures (94,044) in 2012/13 (51% reported to the NJR. There were 196,403 procedures and 49% as a proportion). Table 1.1 also records submitted in 2012/13. ankle procedures from April 2010 and elbow and shoulder procedures from April 2012. Northern Table 1.1 shows the total number of hip and knee Ireland’s joint replacement data is included from procedures recorded on the NJR in England and February 2013.

Table 1.1 Total joint replacement procedures entered into the NJR, 2008/09 to 2012/13, recorded by country in which the procedure took place.

2008/09 2009/10 2010/11 2011/12 2012/13 England hip 71,845 74,676 82,077 83,787 88,830 England knee 75,971 79,173 86,581 87,378 92,698 England shoulder 2,881 England elbow 345 England ankle 423 518 679 Wales hip 4,900 3,919 4,004 4,533 5,144 Wales knee 5,651 4,611 4,506 5,099 5,557 Wales shoulder 110 Wales elbow 13 Wales ankle 25 40 36 Northern Ireland hip 70 © National Joint Registry 2013 Northern Ireland knee 33 Northern Ireland shoulder 5 Northern Ireland elbow 0 Northern Ireland ankle 2 Total procedures 158,367 162,379 177,616 181,355 196,403

Source: Procedures entered into the NJR 1 April 2008 to 31 March 2013.

www.njrcentre.org.uk 29 Operation types reported, whilst the proportion of revisions (less shoulders and elbows) has increased by 0.6% to 9.3% Table 1.2 below shows the number of procedures from 8.7% in 2011/12. reported by type from 1 April 2008 to 31 March 2013. Primary operations make up 91.5% of all procedures

Table 1.2 Total joint replacement procedures entered into the NJR, 2008/09 to 2012/13, recorded by procedure type.

2008/09 2009/10 2010/11 2011/12 2012/13 Hip primary 69,342 70,268 76,927 78,405 82,837 Hip revision 7,327 8,245 9,117 9,914 11,207 Hip re-operation4 76 82 37 1 0 Knee primary 77,020 78,561 85,272 86,637 91,682 Knee revision 4,514 5,147 5,766 5,840 6,606 Knee re-operation4 88 76 49 0 0 Ankle primary 418 538 650 Ankle revision 30 20 67 Shoulder primary 2,666 © National Joint Registry 2013 Shoulder revision 330 Elbow primary 267 Elbow revision 91 158,367 162,379 177,616 181,355 196,403

Source: Procedures entered into the NJR 1 April 2008 to 31 March 2013.

4 Re-operation information was not collected on the first version of the Minimum Dataset (MDSv1) from 1 April 2003 to 31 March 2004. It was included on MDSv2 from 1 April 2004 and removed from MDSv3 which came into use on 1 December 2007. However, some units are continuing to use MDSv2 which is why some re-operations continue to be reported. The figures are included for completeness only.

30 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Procedures by provider type There are no NHS treatment centres or ISTCs in either Wales or Northern Ireland and as the numbers for There are four types of organisations in England Northern Ireland only cover a six week period, they are carrying out hip, knee, ankle, elbow, and shoulder joint grouped together as a provider. replacement surgery: Table 1.3 shows the proportion of procedures by type • NHS hospitals of provider. • NHS treatment centres • Independent sector hospitals • Independent sector treatment centres (ISTCs)

Table 1.3 Proportion of reported procedures by type of provider, 2008/09 to 2012/13.

2008/09 2009/10 2010/11 2011/12 2012/13 Hips 69.4% 71.6% 71.7% 70.8% 71.2% NHS hospitals 53,231 56,272 61,707 62,570 66,952 26.1% 24.4% 24.3% 25.2% 25.5% Independent hospitals 20,040 19,164 20,881 22,256 23,947 4.5% 4.0% 4.1% 4.0% 3.3% ISTCs 3,474 3,159 3,493 3,494 3,075 - - - - 0.1% Northern Ireland - - - - 70 All NJR 76,745 78,595 86,081 88,320 94,044 Knees 70.2% 72.5% 72.2% 70.6% 70.7% NHS hospitals 57,270 60,707 65,781 65,330 69,501 24.2% 22.5% 23.1% 24.8% 25.6% Independent hospitals 19,773 18,859 21,044 22,898 25,135 5.6% 5.0% 4.7% 4.6% 3.7% ISTCs 4,579 4,218 4,262 4,249 3,619 - - - - <1%

Northern Ireland © National Joint Registry 2013 - - - - 33 All NJR 81,622 83,784 91,087 92,477 98,288 Ankles - - 79.0% 76.0% 78.2% NHS hospitals - - 354 424 561 - - 17.6% 19.4% 18.3% Independent hospitals - - 79 108 131 - - 3.3% 4.7% 3.2% ISTCs - - 15 26 23 - - - - 0.3% Northern Ireland - - - - 2 All NJR - - 448 558 717 Continued >

www.njrcentre.org.uk 31 Table 1.3 (continued)

2008/09 2009/10 2010/11 2011/12 2012/13 Shoulders - - - - 82.6% NHS hospitals - - - - 2,476 - - - - 14.7% Independent hospitals - - - - 439 - - - - 2.5% ISTCs - - - - 76 - - - - 0.2% Northern Ireland - - - - 5 All NJR - - - - 2,996 Elbows - - - - 94.1% NHS hospitals © National Joint Registry 2013 - - - - 337 - - - - 5.6% Independent hospitals - - - - 20 - - - - 0.3% ISTCs - - - - 1 - - - - 0.0% Northern Ireland - - - - 0 All NJR - - - - 358

Source: Procedures entered into the NJR 1 April 2008 to 31 March 2013.

Further information is available on the NJR website under ‘NJR 10th Annual Report’.

32 www.njrcentre.org.uk Part 1

1.3 Work of the NJR Steering Committee and its Sub-committees 1.3.1 Introduction It is clear that, as the NJR has matured over its first ten years, the NJR’s governance structure and its This section highlights the work undertaken by the NJR resources, designed for a simpler set of activities and Steering Committee and its sub-committees throughout outputs, could not be sustained with an increasing 2012/13. The reports not only reflect the continually work load, a growing list of strategic priorities, increasing requirements placed on the NJR by its changes in the healthcare environment, greater numerous stakeholders but the commitment and efforts stakeholder expectations, and rapidly increasing of the NJR Steering Committee members who receive international interest and collaboration. no remuneration, other than travel and accommodation expenses, for their considerable effort. In order to ensure that appropriate focus on all areas of business was maintained, decisions could be made in a The membership of the NJR Steering Committee timely and efficient manner, representation included all and membership of and attendance at meetings for major stakeholders, resource capacity was appropriate all sub-committees for 2012/13 can be found online to the effort required, and that roles and responsibilities (Appendix Four). were clearly defined, the following updates to the NJR’s operating model were considered: 1.3.2 NJR Steering Committee • The creation of a new NJR Executive Committee to In July 2012, the NJR Steering Committee held a provide strategic leadership and decision making in workshop with invited stakeholders to examine how a more effective manner, and tasking the Steering the NJR governance and operating model should Committee to handle operational matters evolve to remain fit for the future. The four key • The creation of a new medical advisory committee components of the workshop were: agree the purpose • The appointment of a new post of NJR Medical of the NJR over the next ten years; agree and prioritise Director the activities needed to support the purpose; assess the capabilities necessary to deliver the activities • Ensuring that the sub-committee structure is fit for and consider the committee structure necessary to purpose and can deliver the objectives support the delivery of those activities. • Ensuring that the relationship between committees and activities is transparent Having agreed the mission statement included at • Increasing committee capacity through the Section 1.1.1, the workshop agreed the recruitment of new members and the involvement following priorities: of NJR RCC Network in what has previously been • Improving data quality steering committee and sub-committee business • Improving data access, data availability and • Ensuring patient involvement in steering committee data linkage and sub-committee business • Supporting research - internal and external • Supporting the NJR’s standing committee structure, by convening ad-hoc working groups to • Effective stakeholder communications and patient work on specific issues and areas of concern of and public engagement NJR development, with clear direction, specified • Extending the benefits of the NJR - new joints, new outcomes, timelines and budgets geographical areas • Ensuring that all committees and working • Greater international collaboration as the largest groups have clear terms of reference, an arthroplasty register in the world agreed composition, and defined purposes and • Supporting cost effectiveness and value for money responsibility, and that members have clearly defined • Outcomes reporting and identifying best and worst roles and terms of office practice and The work to define and agree the final operating • Providing a ‘rapid response’ service - bespoke model will continue in 2013/14 with the final operating reactive analysis of NJR data model being submitted to DH for its consideration and

34 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

endorsement. Once that has been achieved, the final The secondary roles of the Editorial Board include structure will reflect the interactions and relationships identification of special topics and in-depth studies. It between the various committees and will include is recognised that the timelines to produce the Annual detailed terms of reference for the committees and Report are often very tight and more detailed work their members. takes a longer period to obtain the necessary outputs. Recent in-depth topics have included analysis of 1.3.3 NJR Editorial Board - Mr Martyn resurfacing, metal-on-metal hip replacements and this Porter, Chairman year includes a study of mortality. There are plans to study infection and revision operations in more detail. The composition and role of the Editorial Board All these studies are planned and monitored through remains unchanged from 2011/12. Its main role is the Research Sub-committee to prevent duplication, to ensure that the Annual Report is produced on overlap and inconsistency. time, for wider distribution at the British Orthopaedic Association (BOA) Congress in the autumn, and to NJR Fellowship Programme ensure that it is reviewed in detail and quality assured. The first two NJR Fellows, Mr Simon Jameson and In 2013/14, to celebrate the 10th anniversary of the Mr Paul Baker, who were appointed for one year, NJR, there will be an independent event to launch the have worked with the Editorial Board and published report’s findings. a range of papers on aspects of both hip and knee replacement in peer review journals. They have also The overall publication process is managed by HQIP. worked in collaboration with the University of Bristol Parts One and Two, focusing on annual progress but more often with their local mentors and academic and clinical activity, are managed by the NJR Centre supervisors in their places of routine clinical activity. (Northgate Information Solutions), particularly with the They have both been extremely motivated and both help of Mike Swanson and Claire Newell. Part Three, presented and published their work which is listed outcomes after joint replacement is produced by the in the appendices. A further update on the NJR University of Bristol team, particularly with the help of Fellowship programme is included at 1.4.11. Professor Ashley Blom, Michèle Smith and Linda Hunt (statistical lead). Part Four, Trust- and unit-level activity Annual Report Development Strategy is a joint contribution from the NJR Centre and the University of Bristol. It is important that the Editorial Board recognise the changes necessary with emerging technology. We The Editorial Board meets five times a year and has have established a working group which will consider several teleconferences with a significant part of the greater emphasis to online and, where possible, real editorial work carried out by email. We meet shortly time reporting and, possibly, a shorter hard copy of after the BOA Congress to discuss the response to the report. We are also considering more interactive the recently launched report, to consider feedback modes of obtaining information. and start planning for the next report publication. We then have a teleconference in March to review We are also considering holding a Research Day the progress of the previous year’s special topics where the outputs and publications from the NJR can and to plan the special topics for the coming year. be presented and discussed in more detail. The meeting in April is to update the Editorial Board regarding the production of the Annual Report and Public and Patient Guide consider the timelines and actions that need to be In conjunction with the HQIP communications team taken. In May we meet to review the early outputs of and, in particular, Rebecca Beaumont, the Editorial Parts One to Four and to consider the format, tables, Board oversees the production of the Public and content and accuracy. There are then two further Patient Guide which was first launched in 2011. We review stages before we meet in early July to approve will continue to develop this important output and the content before formal sign off at the July NJR the 2012 edition included important patient-friendly Steering Committee meeting. developments (see 1.4.5).

www.njrcentre.org.uk 35 BOA Congress Company B. There have not been any examples of outlier performance when the articulation has been The Editorial Board also work with the British ‘hard on soft’ (such as a metal or ceramic head Orthopaedic Association where they provide at least articulating with a high density polyethylene cup or one instructional session at the Annual Congress. In liner) but there have been two Level 1 reports and nine 2013/14, we will be fortunate to have three sessions Level 2 reports for combinations which involve ‘hard as part of the programme. on hard’ (such as metal-on-metal).

1.3.4 Implant Performance Sub- We keep a close watch on some, usually relatively committee - Mr Keith Tucker, Chairman new prostheses, where concern has been raised in other countries. For example, during the past year, The Implant Performance Sub-committee essentially there have been some concerns expressed by one embraces two committees. The main committee manufacturer with regard to a femoral component includes the members of the Scrutiny Group together which incorporates a modular neck – this was with representatives from the implant manufacturing showing up as an outlier in the Australian Registry. We industry. The composition of the Scrutiny Group have looked carefully at all the NJR data for implants includes representatives from the NJR Steering with modular necks and compared this to use in Committee, the Medicines and Healthcare products some other countries, but surgeons recorded on the Regulatory Agency (MHRA), HQIP, Northgate NJR have not made a lot of use of modular necks Information Solutions and the University of Bristol. prostheses in hip replacement.

It is the Scrutiny Group’s brief to analyse and assess Looking to the future, the committee feels it is time the confidential data that relates to implants that have that we turned our attention to revision prostheses. come under performance review (potential outliers). The NJR has about 76,000 revisions logged into the A performance review can be triggered by verbal or database. Anecdotally we know that some revision written communication, a paper or presentation at a procedures do better than others. The implants are surgical meeting or a report from another joint registry. often (but not always) extremely expensive. There are Most commonly a review results from analysis of our also issues about whether surgical experience and own data where the Patient Time Incidence Rate volume have a part to play in outcomes. In our view, a (PTIR) has been found to be unacceptably high. The fully comprehensive assessment of these implants is PTIR is the revision rate per 100 observed years and now due. is calculated by the statistical team in Bristol from the data prepared by Northgate. The PTIR for every My thanks go to all the committee for their dedication implant is measured twice yearly. and hard work - it is one of the great strengths of the NJR that it has the ability to monitor device We are reviewing our methodology at present but performance in this manner, to facilitate outlier device essentially when an implant has a PTIR of twice the investigation by both the regulator and manufacturers. group PTIR (Level 1) a report is filed with the MHRA. Thanks too must go to Melissa Wright (HQIP) for her When the PTIR is only 1.5 times the group PTIR, a organisation and note keeping. warning letter is sent to the company (Level 2). Over the past year there have been two Level 1 reports and 1.3.5 Surgeon Outliers Sub-committee fourteen Level 2 reports. - Professor Paul Gregg, Chairman Since this committee has been in existence we have The Outliers Sub-committee (surgical data) met picked up outlier performance which has involved three times in 2012/13. The committee’s role is to ‘mix and match’. This is where components of ensure that, at the earliest stage, concerns about a hip replacement are made by more than one potential outlier performance of units and surgeons is manufacturer. Examples include a stem being made highlighted and communicated to those concerned. by Company A and the acetabular component by

36 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

It is important to emphasise that the identification In the case of surgeons known to have retired, of a potential outlier surgeon does not mean that be on long-term sick leave, moved abroad or left the a surgeon’s performance is poor as many other Trust or Local Health Board, a letter would also be contributing factors may have influence, for example, sent to the CEO for notification. type of implant used, case-mix and incomplete provision of data to the NJR. All of these factors For private units, letters are sent to the Group CEO have been seen to contribute to potential outlier who would forward the letter to the appropriate performance in some cases. general manager.

The first Annual Clinical Reports to NHS Trusts, Local Future reports will be sent out in April/May each year Health Boards and independent providers were sent to coincide with the new financial year. out in September 2012 and the reports contained the Six months after the dissemination of the Annual following data: Clinical Reports, the data is scrutinised and if • Compliance rate as a percentage additional potential outlier units or surgeons are identified, the relevant CEO is informed six weeks after • Consent rate as a percentage notifying the surgeon. • Linkability rate as a percentage • A ‘traffic light’ system indicating their performance CEOs are asked to acknowledge receipt of the against these indicators: green, amber, red report and, if there is no response after a second communication, the Care Quality Commission (CQC In addition: in England or NHS Wales) is informed. The latter has only been necessary for two Trusts following the first • Unit mortality ratio (reported as Standardised distribution of Annual Clinical Reports. Mortality Ratio (SMR) with a ‘traffic light’ system) • Unit revision rate for hips and knees (reported as Two CQC representatives attended one of the sub- Standardised Revision Ratio (SRR) with a ‘traffic committee meetings in order for the NJR to share light’ system) with them our current approach to managing potential outlier units and surgeons. The CQC will continue to • Individual surgeon (anonymised) revision rates for work with the NJR to determine how our data can form hips and knees part of the CQC Quality Risk Profiles information and be In addition, the SRR’s are shown on funnel plots for all used in their hospital inspection process in England. units and surgeons with individual funnel plots for all hips, cemented hips, uncemented hips, resurfacing/ In some cases, surgeons had been identified as MoM hips, hybrid hips, all knee replacements, potential outliers because of their use of specific cemented knee replacements, uncemented knee prostheses known to have poor results and which, replacements, unicondylar knee replacements and subsequently, have been removed from the market. patello-femoral knee replacements. It had been agreed that three years after a prosthesis has been removed from the market due to poor In the event of a surgeon being identified as a ‘potential results, those cases would be removed from the outlier’ (SRR > 99.8% control limit), a letter is sent database of all surgeons and their SRR re-calculated. to the surgeon six weeks before the Annual Clinical In the three years interim, funnel plots would be shown Report is sent, informing him/her of the situation and with the withdrawn prosthesis in and a plot with the enclosing the data upon which the potential outlying withdrawn prosthesis out. status had been calculated. This is to allow the surgeon opportunity to validate the data which the NJR holds for The committee also noted for 2012/13 the difficulties them. Separate notification of the identity of a potential encountered in an audit of Trusts to seek further outlying surgeon is made to the CEO at the time of clarification of the performance of the Pinnacle sending the Annual Clinical Report. acetabular cup when used with a metal liner. Despite

www.njrcentre.org.uk 37 a letter being sent to 13 Trusts in July 2012 by Sir survivorship statistics, as part of NHS England’s Bruce Keogh, NHS Medical Director, instructing them transparency agenda. to take part, only eight responded by March 2013 and only four of those had completed the data collection. Thanks are expressed to the members of the The matter was escalated to the CQC to follow up on committee for their time and contribution to this behalf of the NJR. The audit has now been completed important, but sometimes challenging, process. for 12 of the 13 units and a report on the outcomes The Chair wishes to extend particular thanks to the provided to the MHRA and NHS England. significant amount of work undertaken by Mr Peter Howard in connection with many of the analyses. After extensive discussions it has been agreed that, to give greater clarity on the sub-types of joint 1.3.6 NJR Research Sub-committee - replacement, the next Annual Clinical Reports will Professor Alex MacGregor, Chairman show separate funnel plots for units and surgeons as follows: The Research Sub-committee’s remit is to maximise access to NJR for the wider research community Knees whilst upholding the quality of research based on NJR • All types of knee replacements together data. The committee takes formal responsibility for the release of data for research through an impartial and • Total knee replacement cemented objective protocol and has oversight of the use and • Total knee replacement uncemented/hybrid reporting of NJR data by research groups. • Unicondylar knee replacement During the period 2012/13, the committee sanctioned • Patello-femoral knee replacement the release of data to six external research groups. These included data released for analysis that will Hips inform the revised NICE guidance for total hip and • All types of hip replacement together surface replacement. The committee received an unprecedented number of expressions of interest for • Total hip replacement cemented future project work from both the UK and overseas, • Total hip replacement uncemented some of which has been submitted for external • Hybrid and reverse hybrid total hip replacement funding. We encourage any researcher planning • Resurfacing and metal-on-metal total hip externally funded work with the NJR to discuss their replacement proposals in confidence with the NJR in advance of submission. Full details of the NJR’s research data It is important to remember that the success and access policy and protocols, together with detailed validity of this outlier process relies heavily on the guidance notes for applicants, can be found at diligence of Trusts, Local Health Boards, independent www.njrcentre.org.uk. providers and surgeons entering all primary and revision procedures. It is the surgeon’s, independent The research profile of the NJR has continued to provider’s, Local Health Board’s and Trust’s expand over the last year, with over 40 presentations responsibility to make sure that all cases are registered to national and international scientific meetings. A and accurate. Participation in the NJR has been total of 18 scientific papers have been published mandatory since April 2011 for NHS hospitals and in peer reviewed literature for this reporting year, since April 2003 for the independent sector. Surgeons based on work carried out by the NJR statistical are advised to check regularly with the NJR Clinician analysis team at the University of Bristol and by Feedback System to ensure the data held by NJR are academic groups working on data releases that were accurate. This will become increasingly important with sanctioned by the committee. the impending publication of named surgeon implant

38 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Notable publications in this period include a British A letter was sent to provider CEOs explaining that, as Medical Journal (BMJ) paper outlining the most part of their compliance with the NJR, a data quality extensive contemporary analysis of the cost- process was being created and it was therefore being effectiveness of cemented, cementless and hybrid requested that they provide the name and contact prostheses for total hip replacement. Also published details of their nominated data quality returns person. in the BMJ, an analysis using NJR-HES linked data An instruction sheet was sent with the letter outlining of cancer risk in the seven years after metal-on-metal what was required from the nominated person and joint replacement provided the first registry based providing advice regarding any coding issues they evidence of the absence of increased cancer risk may come across. The NJR website would list each following surgery. Findings published in the Lancet provider’s nominated person. showed an inferior implant survivorship for metal-on- metal resurfacing in women when compared to other It is disappointing to report that only 55 responses surgical options. This year also saw the first set of were received from 151 letters which were sent analyses to include NJR data linked to PROMs. to all Trusts in England, Welsh Local Health Boards, and private sector groups. Chase letters were sent A full list of publications related to NJR data is included to all CEOs of non-responding providers, with copies in Appendix Two. The redesigned research library to Medical Directors, and failure to respond will lead section of our website catalogues all NJR research to a third and final letter being sent to provider CEOs data releases with their progress reports, together with to state that any data from their units that appeared links to all publications related to NJR data. in the public domain would be un-validated as the NJR had not received a named representative. A 1.3.7 NJR Data Quality Group - list of providers that have not supplied a data Professor Paul Gregg, Chairman quality representative will appear in future NJR Annual Reports. A Data Quality Sub-committee was established in 2012 and met once during this reporting year In addition to the above, a data accuracy audit is 2012/13. being planned across 20 orthopaedic units under the direction of the Regional Clinical Coordinators (RCCs). Although there has been a gradual and progressive A sample set of 30 patients (15 hips, 15 knees with improvement in compliance, consent and linkability 5 of each set being revisions) will be studied in each since the start of the registry, there was still concern of the units. This study will look at the accuracy of about the compliance with the registration of both data entry onto the NJR Minimum Dataset (MDS) primary and revision joint replacement by some form (comparing with notes, theatre logs etc.) and Trusts despite being made mandatory in April 2011. accuracy of transcription from the NJR MDS form to The terms of reference were to explore methods of the database. improving data capture and quality. It should be recorded that three previous data It was decided to try and establish a named data accuracy studies have shown encouraging results for quality person for each provider who would be the accuracy of the main data fields. responsible on an annual basis, for providing the numbers of procedures performed to the NJR for I would like to thank the members of the sub-committee cross-checking and validation. for their contributions and the RCC Network for undertaking the proposed audit of data accuracy.

www.njrcentre.org.uk 39 1.3.8 NJR Regional Clinical Coordinators’ (RCC) Network – Mr Peter Howard, Chairman

As Chair of the Regional Clinical Co-ordinators Network since 2007, I am aware of the significant achievements this group has made in supporting the raising of the profile of the NJR to fellow clinicians and acting as the key link in communicating NJR activities to hospital managers and other relevant hospital staff.

Working with our respective Regional Coordinators this year, the Network has, as always, been keen to understand the specific concerns of individual hospitals in relation to their NJR responsibilities. Although there continue to be issues with some Trusts, there are clear improvements within this area.

The RCCs have been vital to understanding specific issues regarding data quality by initiating unit-level audits of NJR data. This work feeds into the strategic actions of the Data Quality Group and will become an ongoing function of the Network.

This year also saw a substantial review of the MDS. The RCCs were able to input into the list of proposed changes to the dataset and help create a priority group of actions to be implemented into the relevant NJR forms in the coming months.

Once again, the RCC Network has supported the development of the Patient and Public Guide to the Annual Report by ensuring key messages from the main report are distilled accurately to enable patients to be clear about the key outcomes of joint replacement surgery.

As my time as Chair of the RCC Network comes to an end, I would like to thank all of the surgeons who have put themselves forward for this role and their individual efforts in championing the NJR at a local level.

40 www.njrcentre.org.uk Part 1

1.4 Highlights This section provides a brief summary of the work and according to the perceived risk inherent in the device. activities undertaken by the NJR during 2012/13. Until 2008 joint replacement implants were graded as Class 2b. Total hip, knee, and shoulder replacement 1.4.1 Geographic extension of the NJR have since been re-classified to Class 3. CE marks are awarded by Notified Bodies. Each European State has From February 2013, hospitals in Northern Ireland one or more Notified Body which is supervised by the started to submit data to the NJR. Some system appropriate competent authority in each country, e.g. development was necessary to accommodate a the MHRA in the United Kingdom. patient identifier different to the NHS number used in NHS Trusts in England and Welsh Local Health The CE Mark is “Compliance”. It is illegal for anyone Boards. The Healthcare Number (HCN) allocated to hinder the sale of a product with a CE mark. The to patients in Northern Ireland is retained and used CE mark is, to a very considerable extent, awarded on if those patients subsequently receive treatment in the basis of ‘Equivalence’. Until now manufacturers either England or Wales. The NJR already has many have sometimes claimed equivalence in more than two HCN numbers recorded for patients who have had features of the design. Equivalence could have been joint replacement surgery. This means that it will claimed in the material, shape, lubrication system etc., be possible to get a more accurate assessment of whilst at the same time claiming the product is a new outcomes where patients may have had a primary and important addition to the market. To obtain a Class joint replacement procedure in Northern Ireland and 3 CE Mark, the manufacturer has also to outline a a subsequent revision in either England or Wales. post-market surveillance plan. Despite CE regulations, It is hoped that the benefits that the NJR brings for devices have been introduced into the market that have patients, surgeons, and healthcare management will proven to be poor. Even with NJR, ODEP, MHRA and be recognised by independent providers. the CE Mark, there are examples of implants falling short of expectations. This has frustrated many for a It is hoped that units in both the Isle of Man and Jersey considerable period. will start submitting data to the NJR in 2013/14. Mr Peter Kay, during his time in office as BOA 1.4.2 Extension of the NJR: Elbows President, developed the concept of Beyond and shoulders Compliance. Implicit in Beyond Compliance is the invitation to a manufacturer to go beyond the demands From 1 April 2012, the NJR started to collect of the CE mark and enter into a process that will offer data about elbow and shoulder joint replacement greater rigour and organisation than that currently procedures. The project was supported by the British available through the CE process. It is proposed that Elbow and Shoulder Society (BESS). Beyond Compliance, whilst remaining an entirely voluntary process, could be used with all new implants The NJR Steering Committee has also agreed to a and modifications to established devices. two-year pilot in which PROMs, using Oxford Shoulder Scores, will be collected from patients who have Beyond Compliance embraces two main features undergone a shoulder replacement procedure at and is governed by a Steering Committee. The first six months following surgery. Pre-operative Oxford feature is an Advisory Group service and the second, Shoulder Scores are collected as part of the primary a very sophisticated post-CE mark online surveillance shoulder data set. The follow-up will be with NHS- service. ODEP has been asked to take the lead with funded patients only in England and Wales. Beyond Compliance and the panel agreed to set up the advisory service and the monitoring process last 1.4.3 Beyond Compliance: ‘Protecting year at the invitation of the BOA and MHRA. NHS Patients, Supporting Innovation’ Supply Chain is responsible for the administrative organisation of the process. For a manufacturer to market a joint replacement in Europe, the device must have been awarded a CE mark The Advisory Group’s first duty is to look at the detailed (Conformité Européene). CE marks are categorised product file provided by the manufacturer. The Advisory

42 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Group will then assess the level of residual risk there also receive details of those procedures where the might be with this product. The manufacturer’s post- device has been revised. marketing surveillance plan, including post-market clinical follow-up studies, will also be reviewed. With the The NJR Centre will also be able to add procedure level manufacturer, the group will discuss and agree the rate data from HES, PEDW, and the NHS England national at which the device could enter the market and add any PROMs programme. Surgeons will be able to upload other comments they think appropriate. Following that additional data such as x-rays, operation notes and decision there will also be agreement on a plan for post- video. Suppliers will also be able to upload pre-CE Mark market clinical follow-up, which will be monitored with the data and results from follow-up studies. implanting surgeons and manufacturer representatives. Not only will surgeons, suppliers, and Advisory Group Any implant selected for the process is notified to members be able to access all the data relating to the NJR Centre which will then create a record for a specific implant, a number of automated reports the implant in a central data repository. When the will also be available, enabling them to monitor the record of a procedure is submitted to the NJR in the performance of that implant. Additionally, there will normal way, the system will automatically recognise a be a number of additional services, such as providing Beyond Compliance device and copy the procedure suppliers with the ability to undertake follow-up studies details to that central repository. The repository will with patients. The flow of data and information is shown in the figure below:

Figure 1.2 Beyond Compliance data flow.

Surgeons Advisory Group Suppliers

Reports and Additional Services

Central Data Repository

Operation Notes Operation Video

X-Rays © National Joint Registry 2013

Pre-CE Mark Data Bibliography Follow-Up Study Data

NJR Procedure Data NJR Centre NJR Primary/Revision Linked Data HES, PEDW, PROMs

Hospitals

www.njrcentre.org.uk 43 Should the results of a device going through or interactive version. These ideas and suggestions the Beyond Compliance process prove to be were taken forward from the workshops at the unsatisfactory, the MHRA will be informed immediately. Patient Focus conference in 2012 and came to fruition through the newly established NJR Patient Through this project, NJR data will be used to assist in Network. The Public and Patient Guide has again the monitoring of new devices with a view to detecting been awarded the Information Standard logo, a quality problems earlier, if they occur, and to help promote mark for evidence-based information supported by the good practice with a system of sophisticated post- Department of Health. market surveillance. Presently, with the involvement of NJR, the development of this project is a world first. The NJR Patient Network has grown over the year, Our progress to date has stimulated discussion for the meeting twice to share views and to get to know the use of a Beyond Compliance system for other devices NJR and its work in more detail. Members have helped such as breast implants and heart valves. to shape the recently revised patient information leaflet and new poster for hospital waiting areas and Please visit www.beyondcompliance.co.uk or contact are currently involved in the development of a patient Mr Keith Tucker for further details. implant card pilot project. The NJR would like to thank all members for their thoughts, comments and time It is appropriate to acknowledge the enthusiasm and spent contributing to the registry’s progress. huge amount of preparatory work that Northgate Information Solutions have put into this project. The Network is a significant development for the registry and the group will be a firm feature in the 1.4.4 Publication of NJR’s structure and work programme in the future. consultant-level data All of the registry’s patient-focused initiatives continue to be supported and championed at NJR Steering In December 2012, the NHS England CEO, Sir Committee through patient representatives Mary David Nicholson, announced plans for nine surgical Cowern and Sue Musson. specialties and one medical specialty to start publishing consultant-level activity and outcomes 1.4.6 NJR Feedback Services: Updates information by June 2013. Orthopaedics was one of the specialties included and the NJR will be used as NJR Clinician Feedback the source of outcomes data for hip and knee joint The Steering Committee agreed to a project to replacement and will provide a web-based portal for implement further reports in NJR Clinician Feedback the publication of this information. The NJR is working following the preparation of a set of prioritised closely with the lead specialty association British requirements by the RCC Network. These high level Orthopaedic Association in their discussions with NHS requirements to be delivered are: England and HQIP to ensure that the requirements are understood and in order to obtain consent from those • A plot track for individual surgeons on the Patient consultant surgeons registered with the NJR to publish Time Incidence Rate report (PTIR) enabling their data. The data will be made available at surgeons to identify either improving or worsening www.njrsurgeonhospitalprofile.org.uk. outcomes. The PTIR report will also be able to be run as ‘Lead Surgeon’ enabling either trainee 1.4.5 Patient-focused initiatives surgeons to assess outcomes where they were the lead surgeon or enabling consultant surgeons A second edition of the NJR’s Public and Patient to separate those procedures where they were Guide to the Annual Report was published in 2012/13 the lead surgeon from those procedures where a with notable improvements in content and design as trainee was the lead surgeon. well as a choice of online formats - a standard PDF

44 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

• The current Revision Rate reports will be updated reflect, for example, recent developments in chemical to include revision rates at five and seven years. As thromboprophylaxis, the need to carry out different for the PTIR report, the Revision Rate report will be types of analyses, removing inconsistencies between extended to include ‘Lead Surgeon’ in addition to the data forms, and the requirement to improve data ‘Consultant in Charge’. quality by removing ambiguity and changing system • Surgeons will be provided with further information defaults. The development will take place later in about their patients whose primary procedures have 2013/14 once the changes have been approved by been revised by another surgeon and, potentially, the NHS Information Standards Board. in another hospital. The reasons for revision and time to latest endpoint will be provided along with 1.4.8 Price Benchmarking sufficient information for the primary surgeon to Quality, Innovation, Production, and Protection (QIPP) identify the patient and, if necessary, review patient is a large transformational programme for the NHS notes. Surgeons will also be able to download and which aims to improve the quality of care provided print the primary procedure details submitted to the by the NHS whilst, at the same time, making up to NJR for those patients. £20 billion of efficiency savings by 2014/15 which will • Surgeons will also be provided with an Annual be invested in frontline care. QIPP Procurement is a Summary report for the purposes of the revalidation workstream of the QIPP programme and aims to help process, very similar to the Annual Clinical Report. NHS provider Trusts reduce and optimise non-pay expenditure without compromising quality of patient NJR Management Feedback treatment and care. • NJR Management Feedback is used to produce the Annual Clinical Report which was sent to units Orthopaedic joint replacement implants represent for the first time in September 2012. The report is relatively high cost and high volume procurement available as a download from NJR Management within this category and are, therefore, subject to Feedback and is accessible by anyone in the unit particular focus. Orthopaedic implants are purchased authorised to do so by senior management. locally or regionally across England and Wales. Each purchasing authority negotiates pricing with suppliers NJR Supplier Feedback based upon local conditions and purchase volumes. As a result of this local procurement model, there are • NJR Supplier Feedback is a unique, procedure- variations in pricing for orthopaedic implants across level online data reporting system to enable NHS England and NHS Wales. With funding from manufacturers to perform post-marketing the Department of Health, the Price Benchmarking surveillance on their range of implants. project aims to highlight this price variation in Improvements to the Supplier Feedback system are England and Wales. Following a successful pilot in being considered to broaden reporting to revision 2012/13 with three purchasing authorities, the pilot hips and knees as well as shoulder, elbow and ankle project was widened to other authorities and will joints. Linkage to PROMs data for primary hips and continue to expand in 2013/14. Early indications are knees is planned, along with the development of that there are, in some instances, significant price additional online analysis tools. variations for some prostheses and that the volume of purchases does not necessarily attract appropriate 1.4.7 Proposed changes to the discounts. This information will become more widely Minimum Dataset available in the latter stages of the project roll out via an online reporting tool. The NJR Steering Committee has also approved the development of the Minimum Dataset to the next version, MDSv6. The changes are not major and

www.njrcentre.org.uk 45 1.4.9 International developments with an inaugural conference at their headquarters in Silver Springs, Maryland. The meeting was attended by The NJR is regularly approached and asked to take representatives of most national joint registries. Since part in a conference or symposium to support greater then ICOR has promoted several collaborative audits international collaboration and we are keen to share our and reports. The NJR is represented on the ICOR experiences across the globe. committee by Keith Tucker.

International Society of Arthroplasty The US FDA and the European Commission are driving Registers (ISAR) the Unique Device Identifier (UDI) initiative which the NJR is fully supporting. Eventually all implants will have The NJR was certainly not the first joint registry. The a UDI (usually entrenched in their bar-code) which will Scandinavians, in particular, were many years ahead relate to an internationally agreed component database. with Australia becoming an important member some 13 years ago. New Zealand was not long following and ICOR grant bursaries for research associated there are now about 18 countries with national joint with registry data and are willing to fund travelling registries in varying stages of development. The USA fellowships. For more information, contact Art has a relatively embryonic national registry but also has Sedrakyan whose details can be found on the ICOR a number of registries which tend to be regional. Italy website (www.icor-initiative.org). has three registries and a national registry project. 1.4.10 Patient Reported Outcomes ISAR was set up in 2004 essentially between the Scandinavian and Australian registries. The NJR joined Measures (PROMs) two years ago and are now full members. The idea The NJR reported the results from its first PROMs behind the foundation was to encourage registries to survey in the 2nd Annual Report in 2005. This study communicate with each other and form data linkages – was carried out in a group of 20,000 patients who we know that there is much to learn from each other. had previously undergone replacement of a hip or knee joint. The NJR initiated a second more extensive ISAR had its first Congress in Bergen during May 2012 one year follow-up PROMs programme in 2011. and, following its success, the management board In order to be eligible for the study, patients had to asked the NJR to host ISAR 2013 in the UK. It is a have completed a pre-operative questionnaire for the great honour to showcase the NJR’s progress and NHS England national PROMs programme and have developments, welcome a plethora of experience and, consented to have their personal details held by the of course, have access to Continuing Professional NJR. Over a six-month period, 50,000 questionnaires Development (CPD) on one’s own doorstep. The ISAR were sent to patients, with approximately 42,500 working group chose Stratford-upon-Avon as the host being returned completed. With about 1,500 town, knowing it would produce an excellent ambience patients being lost to follow-up since the start of for the meeting. the programme, approximately 41,000 Year Three The current President of ISAR is Göran Garellick and Mr follow-up questionnaires will be despatched over a Martyn Porter will be President 2014/15 following the six-month period which began in March 2013. It is American Academy of Orthopaedic Surgeons (AAOS) intended to undertake a further follow-up project at meeting in February 2014. five years following surgery.

International Consortium of Orthopaedic The data from the Year One follow-up is currently Registries (ICOR) being examined and will be linked to both pre- and post-operative data from the NHS England PROMs The ICOR initiative was launched in 2011, essentially programme, NJR data and HES data. The team as a response to the problems of metal-on-metal hip from the University of Bristol, along with Steering replacement in the US. It was promoted by the US Committee member Professor Alex MacGregor, will Food and Drugs Administration (FDA) and launched undertake the analysis of the data.

46 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

1.4.11 NJR Fellowships • Evert Smith – Consultant orthopaedic surgeon, North Bristol NHS Trust Following the excellent work undertaken by the • Ian Stockley – Consultant orthopaedic surgeon, previous two NJR Fellows, the NJR advertised the Sheffield Teaching Hospitals NHS Foundation Trust availability of another Fellowship in the latter half of 2012. Following interview, Mr Jeya Palan has been The panel has recently started the assessment of appointed and is due to start his Fellowship in April required criteria for the evaluation of knee implants 2013. His work will be directed by the NJR Research which will follow a similar process of data submission Sub-committee and he will be supported by the team and evaluation to the hip products. In order to gain at the University of Bristol looking at issues in relation the expert knowledge required for this field we have to unicompartmental knee replacement in addition recruited the following members: to revision hip and knee replacements. Another NJR • Colin Esler – Consultant orthopaedic surgeon, Fellowship will be advertised later in 2013. University Hospitals of Leicester NHS Trust 1.4.12 Orthopaedic Data Evaluation • Richard Parkinson – Consultant orthopaedic surgeon, Wirral University Teaching Hospital NHS Panel (ODEP) overview Foundation Trust The ODEP group was set up to monitor NICE • Andrew Porteous – Consultant orthopaedic guidance on primary hip implants in 2002 and hip surgeon, North Bristol NHS Trust resurfacing in 2004. Although ODEP is independent • Andrew Toms – Consultant orthopaedic surgeon, from the NJR, its Chairman Keith Tucker is also a Royal Devon and Exeter NHS Foundation Trust member of the NJR Steering Committee and the NJR • Timothy Wilton – Consultant orthopaedic surgeon, provides data to facilitate the monitoring process. Derby Hospitals NHS Foundation Trust The NJR also uses ODEP outputs in a number of published reports. The panel provides an ongoing The panel members accept data submissions assessment of hip implants to benchmark both hip for hips bi-annually in spring and autumn with femoral stems and hip acetabular cups against the monthly meetings arranged to discuss and review NICE guidance, providing a benchmark rating for improvements to the process scheduled throughout implant survivorship and submitted data quality. The the year. ODEP rating is now a commonly used benchmark not only in the UK but globally. A full list of products The year 2012 saw a marked increase in the level of submitted to the panel and their individual ratings can activity of the panel with the numbers of submissions be found at being provided for benchmark reaching 108 and the www.odep.org.uk. start of three major activities:

The group is hosted by NHS Supply Chain and • Initiation of a full review of the hip benchmarking members, listed below, volunteer their time process and expertise: • Introduction of a benchmarking process for knees • Formation of the Beyond Compliance Advisory • Keith Tucker – Chairman of ODEP - Consultant Group (see 1.4.3) orthopaedic surgeon from Norwich Plans for 2013/14 will see the full launch of the revised • Peter Kay – Consultant orthopaedic surgeon, hip process, submissions from industry to gain a knee Wrightington, Wigan and Leigh NHS Foundation benchmark and the completion of the pilot phase of the Trust. Beyond Compliance initiative. In addition, improvements • Philip Lewis – NHS Supply Chain to access of information with the launch of a new • Martin Pickford – Northgate, orthopaedic advisor to website and, in recognition of the importance placed on the NJR ODEP, introducing further programme support. • Andy Smallwood – Procurement Wales

www.njrcentre.org.uk 47 Part 1 1.5 Finance National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

1.5.1 Income and expenditure Below are some of the 2012/13 costs associated with 2012/13 the major NJR project areas: • Shoulder PROMs The NJR is self financing, funded by a levy raised £46,750 (total cost of project £116,950) on the sale of hip, knee, ankle, shoulder, and elbow implants to NHS and independent healthcare • Hospital Management Feedback providers in England, Wales, and Northern Ireland £14,950 (total cost of project £49,950) (the latter from February 2013). The rate of the levy • Geographical extension of the NJR is recommended by the NJR Steering Committee for £39,083 (total cost of project £146,750) approval by the Department of Health (DH), and is • Clinician Feedback Phase Two subject to a Memorandum of Understanding between £70,500 (total cost of project £141,650) the DH, Welsh Government, Health and Social Care • Price Benchmarking Phase One Board Northern Ireland, Independent Healthcare £49,705 (total cost of project £49,705) Advisory Services and the Association of British Healthcare Industries (ABHI) Orthopaedics Special The NJR’s financial results are included in the audited Interest Section. accounts of HQIP (Healthcare Quality Improvement Partnership) which manages the registry. The full The levy was set at £20.00 per joint from 1 April audited accounts are available on HQIP’s website 2012 to 31 March 2013. Levy income in 2012/13 from September 2013 (www.hqip.org.uk), and also was £3,292,579 (2011/12: £3,131,630). Expenditure from the Charity Commission, and Companies House. for the same period was £2,512,741 (2011/12: £2,831,850).

Total expenditure for members’ expenses during 2012/13 was £8,337 (2011/12: £26,069).

www.njrcentre.org.uk 49 Appendix 1 NJR Steering Committee 2012/13

A1.1 NJR Steering Committee – Composition

As a department expert committee, the composition of the NJRSC by category is:

• Chairman 1 • Orthopaedic surgeons 3 • Patient representatives 2 • Implant manufacturer/supplier industry 2 • Public health/epidemiology 1 • NHS organisation management 1 • Independent healthcare provider 1 • Practitioner with special interest in orthopaedic care who is a GP, nurse or allied 1 health professional (physiotherapist or occupational therapist)

A1.2 Membership from April 2012

Members are appointed as posts become vacant. Laurel Powers-Freeling Chairman (from April 2011) Professor Paul Gregg Vice Chairman Acting Chairman (October 2009 to March 2011) Orthopaedic Surgeon (from October 2002)

Mr Mick Borroff Orthopaedic device industry (from October 2002) Mary Cowern Patient Representative. Patient group – Arthritis Care (from October 2006) Dr Jean-Jacques de Gorter Independent healthcare sector (from October 2011) Professor Alex MacGregor Public health and epidemiology (from October 2002) Sue Musson Patient representative (from October 2011) Carolyn Naisby Practitioner with special interest in orthopaedics (from July 2006) Mr Martyn Porter Orthopaedic surgeon (from January 2003) Dean Sleigh Orthopaedic device industry (from April 2008. Resigned in February 2013) Mr Keith Tucker Orthopaedic surgeon (from May 2007) Andrew Woodhead NHS Trust management (from January 2007)

50 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

A1.3 Observers

The following have regularly attended NJR Steering Committee meetings as observers:

Mr Peter Howard Chairman of the NJR Regional Clinical Coordinators’ Network Dr Crina Cacou Senior Medical Device Specialist, MHRA Andy Smallwood Procurement Wales Elaine Young National Development Lead, HQIP Robin Burgess Chief Executive, HQIP Robin Rice Welsh Government

A1.4 Members’ declarations of interest

Laurel Powers-Freeling No interests to declare Professor Paul Gregg Emeritus Consultant Orthopaedic Surgeon, James Cook University Hospital, Middlesbrough, Professor of Orthopaedic Surgical Science, Orthopaedic Advisor for Ramsay Healthcare Mick Borroff Chair, ABHI Orthopaedics Special Interest Section Employed by DePuy International Ltd, manufacturer of orthopaedic prostheses Mary Cowern Wales Director for the UK charity Arthritis Care Dr Jean-Jacques de Gorter Director of Clinical Services, Spire Healthcare Professor Alex MacGregor Professor of Genetic Epidemiology, University of East Anglia Consultant Rheumatologist, Norfolk and Norwich University Hospital NHS Trust Sue Musson Managing Director, Firecracker Projects Limited (supplying management consultancy to NHS organisations) Non-Executive Director, Bridgewater Community NHS Trust Carolyn Naisby Consultant Physiotherapist, City Hospitals Sunderland NHS Foundation Trust Mr Martyn Porter Consultant Orthopaedic Surgeon, Wrightington, Wigan and Leigh NHS Trust (orthopaedic unit has received financial support from DePuy International for clinical and RSA studies for Elite Plus femoral stem and C-Stem). Has acted as a consultant to DePuy International in relation to the development of a hip femoral stem (C-Stem AMT) and received royalties on this hip stem Mr Dean Sleigh ABHI Council member, ABHI Orthopaedics Special Interest Section Mr Keith Tucker Consultant orthopaedic surgeon, Norwich Royalties received from Johnson & Johnson Orthopaedics more than five years ago for contribution to design of hip prostheses (all royalties paid into an orthopaedic charity) Andrew Woodhead Head of Mergers and Acquisitions, NHS London (until 31 March 2012) Managing Director - Andrew Woodhead Consulting Ltd (from Oct 2012) Associate Consultant - Deloitte (from Oct 2012) Associate Consultant - Verita (from Nov 2012) Trustee - Guideposts Charity (from July 2013)

www.njrcentre.org.uk 51 Appendix 2 List of papers, publications and research requests using NJR data

This appendix provides details of published analyses and data releases that have been sanctioned by the NJR Research Sub-committee between April 2012 and March 2013. NJR data is available for research purposes following approval by the NJR Research Sub-committee. For further details please visit the NJR website at www.njrcentre.org.uk.

Published papers (April 2012 to March 2013) Factors Influencing Revision Risk Following 15,740 Single-Brand Hybrid Hip Arthroplasties: A Cohort Study From a National Joint Registry Jameson SS, Mason JM, Baker PN, Jettoo P, Deehan DJ, Reed MR. J Arthroplasty. 2013 Mar 21. pii: S0883-5403(13)00108-3. doi: 10.1016/j.arth.2012.11.021.

Cemented, cementless and hybrid prostheses for total hip replacement: a cost-effectiveness analysis Pennington M, Grieve R, Sekhon JS, Gregg PJ, Black N, van der Meulen JH. BMJ. 2013 Feb 27;346:f1026. doi: 10.1136/bmj.f1026.

Mid-term survival following primary hinged total knee replacement is good irrespective of the indication for surgery Baker P, Critchley R, Gray A, Jameson S, Gregg P, Port A, Deehan D. Knee Surg Sports Traumatol Arthrosc. 2012 Dec 14.

Mid-term equivalent survival of medial and lateral unicondylar knee replacement: an analysis of data from a National Joint Registry Baker PN, Jameson SS, Deehan DJ, Gregg PJ, Porter M, Tucker K. J Bone Joint Surg Br. 2012 Dec;94(12):1641-8. doi: 10.1302/0301-620X.94B12.29416.

The design of the acetabular component and size of the femoral head influence the risk of revision following 34,721 single-brand cemented hip replacements: a retrospective cohort study of medium-term data from a National Joint Registry Jameson SS, Baker PN, Mason J, Gregg PJ, Brewster N, Deehan DJ, Reed MR. J Bone Joint Surg Br. 2012 Dec;94(12):1611-7. doi: 10.1302/0301-620X.94B12.30040.

Failure rates of metal-on-metal hip resurfacings: analysis of data from the National Joint Registry for England and Wales Smith AJ, Dieppe P, Howard PW, Blom AW; National Joint Registry for England and Wales. Lancet. 2012 Nov 17;380(9855):1759-66. doi: 10.1016/S0140-6736(12)60989-1.

Patient and implant survival following 4323 total hip replacements for acute femoral neck fracture: a retrospective cohort study using National Joint Registry data Jameson SS, Kyle J, Baker PN, Mason J, Deehan DJ, McMurtry IA, Reed MR. J Bone Joint Surg Br. 2012 Nov;94(11):1557-66. doi: 10.1302/0301-620X.94B11.29689.

52 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Revision following patello-femoral arthoplasty Appendix 2 Baker PN, Refaie R, Gregg P, Deehan D. Knee Surg Sports Traumatol Arthrosc. 2012 Oct;20(10):2047-53. List of papers, publications and Revision for unexplained pain following unicompartmental and total knee replacement Baker PN, Petheram T, Avery PJ, Gregg PJ, Deehan DJ. research requests using NJR data J Bone Joint Surg Am. 2012 Sep 5;94(17):e126. The association between body mass index and the outcomes of total knee arthroplasty Baker P, Petheram T, Jameson S, Reed M, Gregg P, Deehan D. J Bone Joint Surg Am. 2012 Aug 15;94(16):1501-8.

Reason for revision influences early patient outcomes after aseptic knee revision Baker P, Cowling P, Kurtz S, Jameson S, Gregg P, Deehan D. Clin Orthop Relat Res. 2012 Aug;470(8):2244-52. doi: 10.1007/s11999-012-2278-7.

The effect of surgical factors on early patient-reported outcome measures (PROMs) following total knee replacement Baker PN, Deehan DJ, Lees D, Jameson S, Avery PJ, Gregg PJ, Reed MR. J Bone Joint Surg Br. 2012 Aug;94(8):1058-66. doi: 10.1302/0301-620X.94B8.28786.

The effect of aspirin and low molecular weight heparin on venous thromboembolism after knee replacement: A non-randomised comparison using National Joint Registry data Jameson SS, Baker PN, Charman SC, Deehan DJ, Reed MR, Gregg PJ, van der Meulen JH. J Bone Joint Surg Br. 2012 Jul;94(7):914-8.

Comparison of patient-reported outcome measures following total and unicondylar knee replacement Baker PN, Petheram T, Jameson SS, Avery PJ, Reed MR, Gregg PJ, Deehan DJ. J Bone Joint Surg Br.2012 Jul;94(7):919-27.

Patient Reported Outcome Measures after revision of the infected Total Knee Replacement: Do patients prefer single or two-stage revision? Baker P, Petheram TG, Kurtz S, Konttinen YT, Gregg P, Deehan D. Knee Surg Sports Traumatol Arthrosc. 2012 June 13.

Independent predictors of revision following metal-on-metal hip resurfacing: A retrospective cohort study using National Joint Registry data Jameson SS, Baker PN, Mason J, Porter ML, Deehan DJ, Reed MR. J Bone Joint Surg Br. 2012 Jun;94(6):746-54.

Risk of cancer in first seven years after metal-on-metal hip replacement compared with other bearings and general population: Linkage study between the National Joint Registry for England and Wales and hospital episode statistics Smith AJ, Dieppe P, Porter M, Blom AW; National Joint Registry of England and Wales. BMJ. 2012 Apr 3;344:e2383. doi: 10.1136/bmj.e2383.

Total hip replacement for the treatment of acute femoral neck fractures: results from the National Joint Registry of England and Wales at 3-5 years after surgery Stafford GH, Charman SC, Borroff MJ, Newell C, Tucker JK. Ann R Coll Surg Engl. 2012 Apr;94(3):193-8. doi: 10.1308/003588412X13171221589720.

www.njrcentre.org.uk 53 Approved requests for NJR data for research (April 2012 to March 2013) Rate of venous Thromboembolism in total ankle arthroplasty Zaidi R Royal National Orthopaedic Hospital NHS Trust, Stanmore February 2013

Does the type of venous thromboembolism (VTE) chemoprophylaxis influence the rate of revision for infection following primary hip and knee replacement? Baker P Health Education North East January 2013

Total hip replacement and surface replacement for the treatment of pain resulting from end stage arthritis of the hip (Review of technology appraisal guidance 2 and 44) Clarke A Warwick Medical School, University of Warwick December 2012

Understanding failure in unicompartmental knee arthroplasty Murray DW Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford November 2012

Revision rates by prosthesis type: parametric survival analysis to inform a model of cost-effectiveness Danielson V Johnson & Johnson Medical November 2012

54 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Appendix 3 Additional information on the NJR website www.njrcentre.org.uk

The following information will also be available on the NJR website:

1. NJR 10th Annual Report - Parts One, Two, Three and Four (annual progress 2012/13, clinical activity 2012, implant survivorship 2003 to 2012, and Trust- and unit-level activity and outcomes 2012) 2. NJR 10th Annual Report - Part One: Annual Progress 2012/13 – Welsh Language 3. NJR 10th Annual Report - NJR Steering Committee and Sub-committee composition and attendance 2012/13 4. NJR 10th Annual Report - NJR Steering Committee Terms of Reference 5. NJR 10th Annual Report - NJR Regional Clinical Coordinators’ Terms of Reference 6. NJR 10th Annual Report - Prostheses Data 7. NJR 10th Annual Report - Tables and Figures 8. NJR 10th Annual Report - Public and Patient Guide

www.njrcentre.org.uk 55 Part 2 Clinical activity 2012

2.1 Introduction National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

This section summarises the number of hip, knee and On average, 218 hip replacements and 235 knee ankle replacement procedures undertaken in England replacements were recorded per orthopaedic unit over and Wales between 1 January and 31 December 2012 the year, although the numbers varied from 1 to 1,490 and entered into the NJR by 28 February 2013. Details procedures for any one joint type. Compared with of elbow and shoulder procedures submitted between 1 previous years, there was a decrease in the number April and 31 December 2012 and submitted into the NJR of units performing more than 400 knee procedures by 28 February 2013 are also included. The information is (Table 2.2) but a large increase in the number of units summarised according to the type of hospital or treatment performing 300-400 knee procedures. Most units centre, procedure type and patient characteristics. performing ankle procedures perform less than 5 in a year but there are four units doing larger volumes. 2.1.1 Hospitals and treatment centres participating in the NJR The nine months of recording elbow and shoulder replacements show that most units do less than 5 A total of 413 orthopaedic units were open and of elbow procedures in nine months whereas more than these 398 (96%) submitted at least one hip, knee, half the units do more than 5 shoulder procedures in ankle, elbow or shoulder procedure to the NJR (Table nine months. 2.1). During the life of the NJR there has been a trend showing a greater proportion of all hip replacements being performed in larger centres. However, in 2012 the proportion of large volume centres has remained constant but the percentage of medium sized centres (200-299) has increased by 3% (Figure 2.1).

www.njrcentre.org.uk 57 Table 2.1 Total number of hospitals and treatment centres in England and Wales able to participate in the NJR and the proportion actually participating in 2012.

Total number of units Number of units submitting Proportion participating Total 413 398 96% NHS hospitals 240 225 94% England 222 207 93% Wales 18 18 100% Independent hospitals 163 163 100% England 157 157 100% Wales 6 6 100% ISTC 10 10 100% © National Joint Registry 2013 England 10 10 100% Wales - - -

Table 2.2 Number of participating hospitals, according to number of procedures performed during 2012.

Number of procedures Total number of hospitals <50 50 - 99 100 - 199 200 - 299 300 - 399 400+ Hip procedures Hospitals entering replacements 396 48 68 109 73 55 43 Hospitals entering primary replacements 396 57 70 119 77 41 32 Knee procedures Hospitals entering replacements 386 43 47 115 64 61 56 Hospitals entering primary replacements 386 49 46 120 69 57 45

Total number of Number of procedures hospitals <5 5 - 9 10 - 14 15 - 19 20 - 24 25+ Ankle procedures Hospitals entering replacements 140 106 20 7 3 0 4

© National Joint Registry 2013 Hospitals entering primary replacements 136 104 20 5 3 2 2 Elbow procedures Hospitals entering replacements 109 95 10 2 2 0 0 Hospitals entering primary replacements 101 93 6 2 0 0 0 Shoulder procedures Hospitals entering replacements 241 114 52 25 17 12 21 Hospitals entering primary replacements 238 115 57 28 11 15 12

Note: Elbow and Shoulder data for nine months only.

58 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.1 Percentage of participating hospitals by number of procedures per annum, 2004 to 2012.

Hip procedures 100%

80%

60%

40%

20% Percentage of hospitals Percentage 0% Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 400+ 3% 4% 7% 9% 10% 8% 10% 11% 11% 300 - 399 7% 8% 8% 10% 9% 12% 11% 14% 14% 200 - 299 13% 18% 14% 18% 17% 17% 19% 16% 18% 100 - 199 27% 31% 29% 25% 31% 31% 27% 29% 28% 50 - 99 23% 21% 22% 23% 17% 18% 19% 18% 17% <50 28% 19% 20% 15% 15% 13% 15% 12% 12% Number of hospitals 392 393 399 391 394 395 400 398 396 Total hip procedures 53,306 63,948 66,703 74,878 78,183 78,918 81,623 84,892 86,488

Knee procedures

100% © National Joint Registry 2013

80%

60%

40%

20% Percentage of hospitals Percentage 0% Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 400+ 4% 5% 8% 12% 13% 12% 14% 16% 15% 300 - 399 4% 8% 9% 11% 10% 14% 12% 10% 16% 200 - 299 11% 14% 12% 14% 15% 17% 17% 18% 17% 100 - 199 25% 33% 24% 25% 31% 29% 26% 29% 30% 50 - 99 24% 20% 25% 19% 15% 15% 18% 14% 12% <50 32% 19% 22% 19% 15% 14% 13% 11% 11% Number of hospitals 393 391 398 389 390 393 392 392 386 Total knee procedures 48,937 64,042 66,207 78,106 82,477 84,148 86,974 89,837 90,842

www.njrcentre.org.uk 59 Part 2 2.2 Hip replacement procedures 2012 National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

The total number of hip procedures entered into the resurfacings has decreased again this year to 1% of NJR during 2012 was 86,488, an increase of 7% over all primary procedures. The percentage of resurfacing 2011. Of these, 76,448 were primary and 10,040 were procedures done in Independent hospitals (3%) is triple revision procedures. The revision ‘burden’ increased that of NHS hospitals (1%), as shown in Figure 2.2. on the 2011 levels by 1% to 12%. Independent units and ISTCs perform more cementless hip primary procedures than NHS hospitals. Table 2.3 shows that 93% of patients at independent hospitals and ISTCs were graded as fit and healthy At NHS hospitals, revision procedures account for or with mild disease according to the ASA system, 14% of all procedures. In comparison, the revision compared with 78% at NHS units. burden at independent hospitals is 7%.

Nearly all procedures (94%) undertaken at ISTCs were primary procedures. The percentage of primary hip

Table 2.3 Procedure details, according to type of provider for hip procedures in 2012.

Independent NHS hospitals hospitals ISTC Total No. % No. % No. % No. % Total hip procedures 61,044 71% 22,415 26% 3,029 4% 86,488 Patient physical status P1 - fit and healthy 6,977 11% 4,703 21% 381 13% 12,061 14% P2 - mild disease not incapacitating 40,955 67% 16,139 72% 2,421 80% 59,515 69% P3 - incapacitating systemic disease 12,565 21% 1,541 7% 227 7% 14,333 17% P4 and P5 547 <1% 32 <1% - 0% 579 <1% Procedure type Primary procedures 52,675 69% 20,927 27% 2,846 4% 76,448 88% Primary total prosthetic replacement 18,756 36% 5,614 27% 946 33% 25,316 33% using cement Primary total prosthetic replacement not 21,985 42% 10,566 50% 1,592 56% 34,143 45% using cement Primary total prosthetic replacement not 11,399 22% 4,203 20% 305 11% 15,907 21% classified elsewhere (e.g. hybrid) Primary resurfacing arthroplasty of joint 535 1% 544 3% 3 <1% 1,082 1% Revision procedures 8,369 83% 1,488 15% 183 2% 10,040 12% Hip single-stage revision 7,236 86% 1,405 94% 171 93% 8,812 88% Hip stage one of two-stage revision 514 6% 39 3% 7 4% 560 6% © National Joint Registry 2013 Hip stage two of two-stage revision 563 7% 43 3% 5 3% 611 6% Hip excision arthroplasty 56 <1% 1 <1% 0 0% 57 <1% Bilateral or unilateral5 Bilateral 214 <1% 152 <1% 2 <1% 368 <1% Unilateral 60,830 100% 22,263 100% 3,027 100% 86,120 100% Funding Independent 697 1% 10,078 45% 5 <1% 10,780 12% NHS 60,347 99% 12,337 55% 3,024 100% 75,708 88%

5 Bilaterals will only be counted as a bilateral if they are entered under the same operation during data entry. If the two procedures are recorded under two different operations they will be counted as two unilateral procedures. Therefore, the count of bilaterals is likely to be an underestimate.

www.njrcentre.org.uk 61 Figure 2.2

Primary hip procedures by type of provider 2012.

100%

80%

60%

40%

20% Percentage of procedure type of procedure Percentage

Type of NHS Independent provider hospitals hospitals ISTCs Cemented 36% 27% 33% Cementless 42% 50% 56% Hybrid 22% 20% 11% Resurfacing 1% 3% <1% Number of procedures 52,675 20,927 2,846

2.2.1 Primary total hip replacement procedures has continued to change in 2012. In 2011, there was a 5% difference and in 2012 this increased procedures (THR) 2012 to 10% more cementless procedures than cemented Of the 76,448 primary hip replacement procedures procedures being carried out. There has been an undertaken in 2012, 33% were cemented THRs, 43% increase in the percentage of hybrid (standard and were cementless, 1% were hip resurfacing procedures reverse hybrid) procedures performed. There has and 2% were large head metal-on-metal (LHMoM) also been a significant decrease in the percentage THRs (Figure 2.3). For the first time the number of of resurfacing procedures and in procedures where operations for the current reporting year available a large metal head is used with a resurfacing cup for this report is greater than the previous year. (LHMoM). This decline is thought to have resulted Normally, this is not the case as not all procedures from the well publicised voluntary recall of one brand performed in the reporting year are submitted in time of resurfacing device (ASR – DePuy) and ongoing to be reported in the Annual Report. It is too soon to concerns regarding the safety of LHMoM procedures tell whether this is due to an increase in activity, an as reflected in MHRA guidance and follow-up. increase in compliance, more timely submissions or a In 2012, 15% of hybrid procedures were reverse hybrid combination of these factors. (cementless stem, cemented socket) and 85% were The ratio between cemented and cementless standard hybrid (cemented stem, cementless socket).

62 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.3

Type of primary hip replacement procedures undertaken between 2005 and 2012.

60%

50%

40%

30%

20%

10% Percentage of procedure type of procedure Percentage 0%

Year 2005 2006 2007 2008 2009 2010 2011 2012 © National Joint Registry 2013 Cemented 54% 46% 43% 38% 36% 36% 36% 33% Cementless 22% 25% 27% 32% 37% 41% 41% 43% Hybrid 12% 14% 14% 15% 15% 16% 18% 20% Resurfacing 9% 10% 9% 8% 6% 3% 2% 1%

Large head metal- 3% 6% 7% 8% 6% 5% 3% 2% on-metal (≥36mm) Number of procedures 56,454 59,967 67,321 70,559 70,908 73,006 75,460 76,448

2.2.1.1 Patient characteristics According to the ASA system, 15% of patients undergoing a primary hip replacement in 2012 were Age and gender were included for those patients graded as fit and healthy prior to surgery, compared who gave consent for their personal identifiers to be with 37% in 2003. Figure 2.6 shows the changes in entered into the NJR and where consent was ‘Not ASA grade over ten years. Patient BMI6 has increased recorded’ (a total of 96%). The average age was 67.4 over the past eight years from 27.4 to 28.7, as shown years, 0.2 years more than last year. Approximately in Figure 2.7. Females undergoing THR have a 60% of the patients were female (Table 2.5) which is consistently lower mean BMI than males; the converse the same as 2011. On average, female patients were is the case for TKR (Figure 2.22). Figure 2.7 shows that older than male patients at the time of their primary hip there has been an increase in the number of patients replacement (69.7 years and 67.2 years respectively, with a BMI of between 30 and 39 and a decrease Table 2.5). Patients undergoing a resurfacing in the number of patients with BMI between 20 and procedure were the youngest, at an average age of 24. The single largest indication recorded for surgery 53.4 years (Table 2.5). Seven times as many males was osteoarthritis, recorded in 92% of procedures have a resurfacing procedure compared with females. (Table 2.4). Figure 2.5 shows that the percentage of patients within the age group bands has not changed

6 BMI: 20-24 normal, 25-29 overweight, 30-39 obese, 40+ morbidly obese.

www.njrcentre.org.uk 63 significantly since 2003, suggesting that the increase in part of MDSv1) and, while this has increased to 67% BMI and reduction in fitness of patients is not due to an in 2012, all BMI data has to be viewed with caution as ageing patient cohort. However, only 16% of entries in surgeons may be more likely to enter BMI data when 2005 had BMI data entered (BMI was not collected as the BMI is high, introducing an element of bias.

Table 2.4 Patient characteristics for primary hip replacement procedures in 2012, according to procedure type.

Primary total Primary total Primary total prosthetic Primary prosthetic prosthetic replacement not resurfacing replacement replacement not classified elsewhere arthroplasty of using cement using cement (e.g. hybrid) joint Total No. % No. % No. % No. % No. % Total hip primaries 25,316 33% 34,143 45% 15,907 21% 1,082 1% 76,448 Patient physical status P1 - fit and healthy 2,433 10% 6,145 18% 2,011 13% 503 46% 11,092 15% P2 - mild disease not 17,730 70% 23,860 70% 11,064 70% 545 50% 53,199 70% incapacitating P3 - incapacitating 4,975 20% 4,019 12% 2,717 17% 34 3% 11,745 15% systemic disease P4 and P5 178 <1% 119 <1% 115 <1% 0 0% 412 <1% BMI Number with BMI data 17,084 67% 22,694 66% 10,775 68% 747 69% 51,300 67% Average 28.35 29.08 28.54 28.09 28.71 SD 5.20 5.34 5.35 4.32 5.29 Indications for surgery Osteoarthritis 23,123 91% 31,837 93% 14,192 89% 1,036 96% 70,188 92% Avascular necrosis 570 2% 828 2% 486 3% 11 1% 1,895 2% Fractured neck of femur 1,035 4% 645 2% 759 5% 1 0% 2,440 3% Congenital dislocation/ © National Joint Registry 2013 171 <1% 758 2% 254 2% 28 3% 1,211 2% Dysplasia of the hip Inflammatory arthropathy 340 1% 409 1% 262 2% 10 <1% 1,021 1% Failed hemiarthroplasty 60 <1% 45 <1% 67 <1% 0 0% 172 <1% Trauma - chronic 370 1% 333 <1% 264 2% 15 1% 982 1% Previous surgery - non 47 <1% 142 <1% 60 <1% 1 <1% 250 <1% trauma related Previous arthrodesis 16 <1% 14 <1% 10 <1% 0 0% 40 <1% Previous infection 22 <1% 25 <1% 17 <1% 3 <1% 67 <1% Other 379 1% 386 1% 288 2% 18 2% 1,071 1% Side Bilateral 42 <1% 228 <1% 80 <1% 14 1% 364 <1% Left, unilateral 11,273 45% 15,487 45% 7,215 45% 529 49% 34,504 45% Right, unilateral 14,001 55% 18,428 54% 8,612 54% 539 50% 41,580 54%

64 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.5 Age and gender for primary hip replacement patients in 2012.

Primary total Primary total Primary total prosthetic Primary prosthetic prosthetic replacement not resurfacing replacement using replacement not classified elsewhere arthroplasty of cement using cement (e.g. hybrid) joint Total No. % No. % No. % No. % No. % Total hip 25,316 33% 34,143 45% 15,907 21% 1,082 1% 76,448 primaries Total hip primaries 24,570 97% 32,827 96% 15,290 96% 958 89% 73,645 96% with patient data Average age 73.1 65.2 70.0 53.4 68.7 SD 9.6 11.4 11.0 9.6 11.4 Interquartile 67.6 – 79.8 58.4 – 73.2 63.9 – 77.7 47.5 – 59.9 62.2 – 76.9 range Average age by gender Female 16,162 66% 18,176 55% 9,682 63% 120 13% 44,140 60% Average 73.66 65.77 70.75 52.69 69.70 SD 9.47 11.35 10.79 9.74 11.10 Interquartile 68.2 - 80.3 59.0 - 73.7 64.7 - 78.4 47.9 - 58.6 63.4 - 77.7 range Male 8,408 34% 14,651 45% 5,608 37% 838 87% 29,505 40% Average 72.14 64.48 68.75 53.52 67.20 SD 9.65 11.40 11.20 9.56 11.60 Interquartile 66.5 - 78.9 57.7 - 72.5 62.4 - 76.1 47.4 - 60.2 60.4 - 75.6 range Age group by gender

Female © National Joint Registry 2013 <30 years 20 <1% 118 <1% 30 <1% 3 3% 171 <1% 30 - 39 years 63 <1% 287 2% 67 <1% 8 7% 425 <1% 40 - 49 years 204 1% 1,176 6% 306 3% 32 27% 1,718 4% 50 - 59 years 983 6% 3,390 19% 1,001 10% 53 44% 5,427 12% 60 - 69 years 3,777 23% 6,523 36% 2,838 29% 22 18% 13,160 30% 70 - 79 years 6,864 42% 5,051 28% 3,561 37% 1 <1% 15,477 35% 80 - 89 years 3,972 25% 1,558 9% 1,729 18% 1 <1% 7,260 16% 90+ 279 2% 73 <1% 150 2% 0 0% 502 1% Male <30 years 12 <1% 116 <1% 30 <1% 11 1% 169 <1% 30 - 39 years 39 <1% 298 2% 61 1% 52 6% 450 2% 40 - 49 years 182 2% 1,146 8% 261 5% 220 26% 1,809 6% 50 - 59 years 595 7% 3,013 21% 761 14% 339 40% 4,708 16% 60 - 69 years 2,278 27% 5,309 36% 1,708 30% 189 23% 9,484 32% 70 - 79 years 3,569 42% 3,761 26% 2,001 36% 24 3% 9,355 32% 80 - 89 years 1,638 19% 976 7% 752 13% 3 <1% 3,369 11% 90+ 95 1% 32 <1% 34 <1% 0 0% 161 <1%

www.njrcentre.org.uk 65 Table 2.6 Indications for hip primary procedure based on age groups.

30-39 40-49 50-59 60-69 70-79 80-89 <30 years years years years years years years 90+ Indication No. % No. % No. % No. % No. % No. % No. % No. % Total hip primaries by 340 875 3,527 10,135 22,644 24,832 10,629 663 age group Osteoarthritis 150 44% 552 63% 2,996 85% 9,348 92% 21,219 94% 23,227 94% 9,756 92% 567 86% Avascular 81 24% 155 18% 237 7% 314 3% 394 2% 363 1% 245 2% 34 5% necrosis Fractured neck of 4 1% 6 <1% 27 <1% 180 2% 598 3% 895 4% 450 4% 31 5% femur Congenital dislocation/ 75 22% 182 21% 333 9% 272 3% 194 <1% 68 <1% 17 <1% 1 <1% Dysplasia of the hip Inflammatory 31 9% 33 4% 100 3% 164 2% 312 1% 249 1% 94 <1% 8 1% arthropathy Failed hemi- 2 <1% 10 1% 7 <1% 20 <1% 43 <1% 67 <1% 16 <1% 0 0% arthroplasty © National Joint Registry 2013 Trauma - 12 4% 26 3% 81 2% 148 1% 236 1% 226 <1% 175 2% 27 4% chronic Previous surgery, 23 7% 41 5% 58 2% 46 <1% 31 <1% 27 <1% 8 <1% 1 <1% non trauma related Previous 5 1% 1 <1% 4 <1% 5 <1% 12 <1% 6 <1% 5 <1% 1 <1% arthrodesis Previous 8 2% 10 1% 7 <1% 9 <1% 13 <1% 14 <1% 4 <1% 0 0% infection Indication 41 12% 63 7% 128 4% 146 1% 215 <1% 248 1% 142 1% 23 3% other

66 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.4

Age and gender for primary hip replacement patients in 2012.

Female 100%

80%

60%

40%

20% Percentage of procedures Percentage 0% Age group <30 30-39 40-49 50-59 60-69 70-79 80-89 90+ Cemented 12% 15% 12% 18% 29% 44% 55% 56% Cementless 69% 68% 68% 62% 50% 33% 21% 15% Hybrid 18% 16% 18% 18% 22% 23% 24% 30% Resurfacing 2% 2% 2% <1% <1% <1% <1% 0% Number of patients 171 425 1,718 5,427 13,160 15,477 7,260 502

Male 100% © National Joint Registry 2013

80%

60%

40%

20% Percentage of procedures Percentage 0% Age group <30 30-39 40-49 50-59 60-69 70-79 80-89 90+ Cemented 7% 9% 10% 13% 24% 38% 49% 59% Cementless 69% 66% 63% 64% 56% 40% 29% 20% Hybrid 18% 14% 14% 16% 18% 21% 22% 21% Resurfacing 7% 12% 12% 7% 2% <1% <1% 0% Number of patients 169 450 1,809 4,708 9,484 9,355 3,369 161

www.njrcentre.org.uk 67 Figure 2.5

Age for primary hip replacement patients between 2003 and 2012.

100%

80%

60%

40%

each age band 20%

Percentage of patients in Percentage 0% Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 <30 <1% <1% <1% <1% <1% <1% <1% <1% <1% <1%

© National Joint Registry 2013 30-39 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 40-49 4% 5% 5% 5% 5% 5% 5% 5% 5% 5% 50-59 16% 15% 16% 15% 15% 15% 14% 14% 14% 14% 60-69 31% 31% 31% 30% 31% 30% 30% 30% 31% 31% 70-79 33% 33% 34% 34% 34% 34% 34% 34% 34% 34% 80-89 13% 13% 13% 13% 13% 13% 14% 14% 14% 14% 90+ <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% Number of patients 14,557 28,222 40,370 47,687 60,637 66,892 67,729 69,900 72,537 73,645

68 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.6

ASA grades for primary hip replacement patients between 2003 and 2012.

100%

80%

60%

40%

20% Percentage of patients Percentage

Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 © National Joint Registry 2013 P1 37% 31% 26% 23% 20% 18% 17% 16% 15% 15% P2 53% 57% 60% 63% 66% 69% 69% 69% 70% 70% P3 9% 11% 13% 13% 13% 13% 14% 15% 15% 15% P4 and P5 <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% Number of patients 26,435 48,061 57,594 59,967 67,322 70,559 70,908 73,006 75,461 76,448

www.njrcentre.org.uk 69 Figure 2.7 BMI for primary hip replacement patients between 2004 and 2012.

30 60%

50% 29

40% 28 30% BMI 27 20%

26 of patients Percentage 10%

25 0% Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 Average BMI 27.36 27.47 27.67 27.91 28.31 28.43 28.53 28.60 28.71 - all patients BMI 15-19 3% 3% 3% 3% 2% 2% 2% 2% 2% BMI 20-24 27% 27% 25% 24% 21% 21% 20% 20% 19% BMI 25-29 40% 40% 40% 40% 41% 40% 40% 40% 40% © National Joint Registry 2013 BMI 30-34 21% 21% 21% 23% 25% 25% 25% 26% 26% BMI 35-39 6% 6% 7% 8% 8% 9% 9% 9% 10% BMI 40-44 2% 2% 2% 2% 2% 2% 2% 2% 3% BMI 45+ <1% <1% <1% <1% <1% <1% <1% <1% <1% Average BMI 27.29 27.15 27.53 27.77 28.12 28.22 28.31 28.40 28.49 - female Average BMI 27.71 27.87 27.94 28.16 28.62 28.72 28.83 28.92 29.04 - male Percentage of procedures 17% 16% 18% 21% 48% 56% 60% 64% 67% with BMI Number of procedures 5,919 9,002 10,473 14,237 33,842 39,431 43,679 48,345 51,300 with BMI

2.2.1.2 Surgical techniques many procedures performed where a lateral (including Hardinge) approach was used. The surgical techniques used in procedures in 2012 are summarised in Table 2.7. Patients were mainly The reduction in the use of cemented stems (from positioned laterally. The lateral position was used more 77% in 2004 to 51% in 2012) and also in the use of frequently in hybrid and resurfacing procedures than in cemented cups (from 56% to 36%) is consistent with cemented and cementless procedures. As would be the reduction seen in the overall number of cemented expected, the most frequently used incision approach procedures and a corresponding increase in hybrid was posterior for all procedure types, though for and cementless surgery. The change in 2012 appears cemented procedure types there were nearly as to have been mainly caused by the rapid decline

70 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

in metal-on-metal procedures rather than a switch increased from 73% in 2004 to 89% in 2012. Use of from one type of conventional hip replacement to minimally-invasive surgery was greatest in cementless another since 2004 (Figure 2.3). The relative usage procedures, although it was used in less than 5% of all of different types of bone cement is shown in Figure procedures (Table 2.7) which is the same as 2011. 2.8 and shows that the use of antibiotic cement has

Table 2.7 Surgical technique for primary hip replacement patients in 2012.

Primary total prosthetic Primary total Primary total replacement Primary prosthetic prosthetic not classified resurfacing replacement replacement not elsewhere arthroplasty using cement using cement (e.g. hybrid) of joint Total No. % No. % No. % No. % No. % Total hip 25,316 33% 34,143 45% 15,907 21% 1,082 1% 76,448 primaries Patient position Lateral 22,965 91% 31,864 93% 15,303 96% 1,077 100% 71,209 93% Supine 2,351 9% 2,279 7% 604 4% 5 <1% 5,239 7% Incision Lateral (inc. 10,131 40% 10,878 32% 4,092 26% 133 12% 25,234 33% Hardinge) Posterior 13,731 54% 21,185 62% 11,157 70% 916 85% 46,989 61% Trochanteric 310 1% 26 <1% 11 <1% 13 1% 360 <1% osteotomy Other 1,144 5% 2,054 6% 647 4% 20 2% 3,865 5% Minimally-invasive surgery Yes 478 2% 2,884 8% 365 2% 14 1% 3,741 5% © National Joint Registry 2013 No 24,838 98% 31,259 92% 15,542 98% 1,068 99% 72,707 95% Image-guided surgery Yes 32 <1% 158 <1% 31 <1% 33 3% 254 <1% No 25,284 100% 33,985 100% 15,876 100% 1,049 97% 76,194 100% Bone graft used - femur Yes 173 <1% 400 1% 44 <1% 13 1% 630 <1% No 25,143 99% 33,743 99% 15,863 100% 1,069 99% 75,818 99% Bone graft used - acetabular Yes 1,014 4% 1,321 4% 841 5% 79 7% 3,255 4% No 24,302 96% 32,822 96% 15,066 95% 1,003 93% 73,193 96%

www.njrcentre.org.uk 71 Figure 2.8

Bone cement types for primary hip replacement procedures undertaken between 2004 and 2012.

90% 80% 70% 60% 50% 40% 30% 20%

Percentage of procedures Percentage 10% 0% Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 Antibiotic-loaded 62% 65% 65% 65% 68% 70% 74% 77% 79% high viscosity Antibiotic-loaded

© National Joint Registry 2013 6% 7% 8% 9% 9% 10% 10% 9% 9% medium viscosity Antibiotic-loaded 5% 4% 3% 3% 3% 2% 2% 1% 1% low viscosity High viscosity 7% 6% 6% 5% 5% 5% 4% 3% 2% Medium viscosity 19% 17% 18% 17% 15% 13% 11% 9% 8% Low viscosity <1% <1% <1% <1% <1% <1% <1% <1% <1% Number of procedures 34,497 40,473 40,544 44,334 43,362 41,514 41,108 43,359 43,789 using cement

2.2.1.3 Thromboprophylaxis (66%), (Table 2.8). There has been a marked decrease over the past years in the use of aspirin, 20% in 2009 Patients may receive more than one chemical and more to 8% in 2012. Direct thrombin inhibitor is now used than one mechanical thromboprophylaxis method. in 13% of hip primary procedures and the use of The most frequently prescribed chemical method of ‘other chemical’ has gone up from 7% in 2009 to 13% thromboprophylaxis for hip replacement patients was in 2012. The number of procedures for which both low molecular weight Heparin (LMWH), at 73%, and chemical and mechanical methods were prescribed has the most used mechanical method was TED stockings continued to rise from 63% in 2007 to 92% in 2012.

72 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.8 Thromboprophylaxis regime for primary hip replacement patients, prescribed at time of operation.

Total

No. % Total hip primaries 76,448 Aspirin 6,124 8% Low molecular weight Heparin (LMWH) 55,713 73% Pentasaccharide 1,148 2% Warfarin 557 <1% Direct thrombin inhibitor 10,203 13% Other chemical 9,583 13% No chemical 2,143 3% Foot pump 20,159 26% Intermittent calf compression 33,052 43% TED stockings 50,713 66% © National Joint Registry 2013 Other mechanical 2,272 3% No mechanical 3,679 5% Both mechanical and chemical 70,700 92% Neither mechanical nor chemical 36 <1%

2.2.1.4 Untoward intra-operative events As would be expected, this occurred more often in cementless than in cemented hips. Furthermore, Untoward intra-operative events were reported in 16% were trochanteric fractures, also more common 1.3% of procedures (Table 2.9). Of the 999 untoward in cementless procedures. 36% of events were of events reported, a decrease of 9 events compared ‘other’ description. with 2011, 33% were attributed to calcar crack.

Table 2.9 Reported untoward intra-operative events for primary hip replacement patients in 2012, according to procedure type.

Shaft penetration 6 14 1 0 21 © National Joint Registry 2013 Trochanteric fracture 56 61 39 0 156 Other 126 157 75 3 361

www.njrcentre.org.uk 73 2.2.1.5 Hip primary components nor the ODEP process. The latest listings for brands currently being used in England and Wales can be This section outlines in more detail the trends in seen on the ODEP website: brand usage for hips. For a full listing of brands used in 2012, please visit the NJR website at www.odep.org.uk www.njrcentre.org.uk. This section also includes an Analysis of primary procedures shows that the use of analysis of usage according to National Institute for products meeting the full 10 year (10A) benchmark, as Health and Care Excellence (NICE) guidelines, as recommended by NICE, is as follows: interpreted by the Orthopaedic Data Evaluation Panel (ODEP). • Cemented stems 88% (using 14 brands out of 47 recorded on the NJR) 2.2.1.5.1 Compliance with ODEP and NICE guidelines • Cementless stems 69% (17 brands out of 91) • Cemented cups 34% (10 brands out of 43) In 2012, 101 brands of acetabular cup, 8 brands of • Cementless cups 3% (7 brands out of 57) resurfacing cup and 146 brands of femoral stem were used in primary and revision procedures and recorded • Resurfacing cups 63% (1 brand out of 9) on the NJR. This is a decrease in the number of These percentages are based on the latest ODEP brands of acetabular cup and stem in use compared ratings from clinical outcomes data already submitted to with 2011. the ODEP committee and published in February 2013. The 2nd NJR Annual Report in 20047 gave a full Manufacturers are expected to submit additional data description of the NICE guidance on the selection to progress through the ratings and this will result in of prostheses for primary THRs and metal-on-metal these percentages changing in the future. hip resurfacing arthroplasty. It also described the Comparison with the 2011 figures shows that use of establishment of ODEP. Its remit is to provide an cemented and cementless stems achieving the 10- independent assessment of clinical evidence, submitted year benchmark has not changed significantly. Usage by suppliers, on the compliance of their implants for of cementless stems achieving the 10-year benchmark THR and hip resurfacing against NICE benchmarks is 69% (a decrease of 3% on last year). However, the for safety and effectiveness. ODEP produced detailed percentage for cementless cups achieving 10-year criteria for this guidance which is currently under review benchmark was 3%. as part of a complete overhaul of the system. 2.2.1.5.2 Hip brand usage in The ODEP committee have reviewed suppliers’ clinical data submissions and ODEP ratings have been given primary procedures to 69 brands of femoral stem (50% of those available) Figures 2.9 to 2.13 show historical trends in the usage and 49 brands of acetabular cup (49%) used in of the most popular brands of cemented stems, primary procedures. However, there are 39 brands of cemented cups, cementless stems, cementless cups acetabular cup (39%) and 46 brands of femoral stem and hip resurfacing cups. (33%) currently being used in England and Wales for which no data has yet been submitted to ODEP. It Figure 2.9 shows that the market is dominated by should be noted however that some of this usage polished collarless tapered stems, with the Exeter relates to stems designed for revision surgery being V40 having a market share of more than 66% and used in primary procedures for unspecified reasons. the CPT stem continuing to consolidate its position in Revision brands are not covered by NICE guidance second place.

7 See pages 86 to 92 of the 2nd NJR Annual Report, available on the NJR website www.njrcentre.org.uk.

74 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.9

Top five cemented hip stem brands, trends 2003 to 2012.

70%

60%

50%

40%

30%

20%

10% Percentage of procedures 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 18,944 31,844 35,608 34,140 35,912 34,004 32,983 33,898 36,222 38,406

Exeter C-Stem AMT C-Stem Charnley CPT V40 Cemented Stem Cemented Stem Cemented Stem

The trend for cemented cups (Figure 2.10) continues significant change being a proportion of Exeter stem to show that sales of different brands are in line with users switching from the Contemporary to the Exeter the popularity of the stem manufacturer. The only Rimfit cup, both manufactured by Stryker.

Figure 2.10

Top five cemented hip cup brands, trends 2003 to 2012.

40% 35% 30% 25% 20% 15% 10% 5% Percentage of procedures 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 15,318 25,540 27,870 25,648 27,051 25,335 24,048 24,276 25,826 26,975

Elite Plus Elite Plus Contemporary Exeter Rim t Marathon Ogee Cemented Cup

www.njrcentre.org.uk 75 The relative sales of cementless stem brands (Figure market. The Corail prosthesis continues to maintain its 2.11) are very similar to the previous year, with press- position as market leader. fit HA coated stems continuing to dominate the

Figure 2.11

Top five cementless hip stem brands, trends 2003 to 2012.

50%

40%

30%

20%

10% Percentage of procedures 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 4,462 10,771 15,163 18,551 23,698 29,411 32,323 34,921 35,885 36,808

Furlong Taperloc M/L Taper Corail Accolade HAC Stem Cementless Stem Cementless

76 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

The cementless stem market share has again reflected Exceed ABT from Biomet. More information on the the sales of the corresponding cementless cups other brands used can be found in the document from the same manufacturers, which means that the ‘Prostheses used in hip, knee, ankle, elbow and Pinnacle cup from DePuy has retained its position shoulder replacement procedures 2012’ which can be as the market leader (Figure 2.12) despite appearing downloaded from the NJR website. to lose some share to other brands, notably the

Figure 2.12

Top five cementless hip cup brands, trends 2003 to 2012.

40% 35% 30% 25% 20% 15% 10% 5% Percentage of procedures 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 7,906 16,417 21,186 24,175 28,826 34,367 39,392 44,296 46,554 47,656

Pinnacle Trident Exceed ABT Trilogy CSF Plus

www.njrcentre.org.uk 77 Figure 2.13 shows the sales evolution of brands of 2012, at the expense of all other brands reflecting its hip resurfacing prostheses in the English and Welsh clinical performance when measured against most of markets. It is evident that the previous trend towards its competitors. However, it should be noted that this a decline in the usage of the original brands has is against a background of an ongoing decline in the continued to reverse. The market share of the BHR overall volume of resurfacing hip replacement. brand increased significantly during the course of

Figure 2.13

Top five resurfacing head brands, trends 2003 to 2012.

90% 80% 70% 60% 50% 40% 30% 20%

Percentage of procedures 10% 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 2,640 5,031 6,230 6,502 6,701 5,798 4,358 2,689 1,883 1,075

BHR Adept Cormet 2000 Accis Recap Resurfacing Resurfacing Resurfacing Resurfacing Resurfacing Head Head Head Head Head

2.2.1.5.3 Trends in head size use 2.2.1.5.4 Trends in hip articulation

Figure 2.14 shows the relative usage of different Figure 2.15 shows the change in hip articulation types femoral head sizes (for all femoral heads used in since the inception of the NJR. Only those procedures conjunction with a femoral stem) each year since the where complete articulation surfaces can be derived inception of the NJR. It is immediately clear that there are included. The most interesting observation being has been a gradual increase in the use of larger head the huge rise in the use of ceramic-on-ceramic sizes (36mm diameter and above). However, between bearings, from small numbers in 2003 to over 17,000 2010 and 2012 this had been reversed slightly. in 2012, though this has remained steady over the This trend accurately reflects the trends in usage of past year. This growth would appear to have been LHMoM prostheses during the lifetime of the NJR. mainly at the expense of metal-on-metal bearings.

78 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.14 Femoral head size, trends 2003 to 2012.

100%

80%

60%

40%

20% Percentage of procedures 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of modular heads components 18,889 36,137 45,078 48,497 56,472 61,674 63,875 67,853 71,247 74,176 used

22.25mm 26mm 28mm 30mm 32mm 34mm

36mm 38-44mm 46+mm

Figure 2.15

Hip articulation, trends 2003 to 2012. 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 Number of procedures 5,000 0 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 © National Joint Registry 2013 Number of procedures with hip articulation 25,344 46,566 55,748 58,293 65,629 68,878 69,457 71,450 73,844 75,897 details

Metal-on Ceramic-on Ceramic-on Metal-on Ceramic-on Metal-on -poly -ceramic -poly -metal -metal -ceramic*

* This combination is contra-indicated

www.njrcentre.org.uk 79 2.2.2 Hip revision procedures 2012 (with the exception of excision arthroplasty). However, the percentage of patients who were graded as being A total of 10,040 hip revision procedures were fit and healthy prior to surgery has decreased from reported in 2012, an increase of 1,401 compared 26% in 2003 to 10% in 2012. with 2011. Table 2.10 shows that of these, 8,812 (88%) were single-stage revision procedures, 560 Aseptic loosening has continued its decrease as a (6%) were stage one of a two-stage process, 611 reason for revision compared with previous years. (6%) procedures were stage two of a two-stage Adverse soft tissue reaction was noted for 13% of revision and 57 (<1%) were excision arthroplasty all revision procedures (Table 2.10). However, this procedures. Infection as an indication for revision has option was added in July 2009 so it is not possible increased to 12% of the total. to tell if this increase is actual or due to a delay in the usage of the new MDS H2 data forms. More than one 2.2.2.1 Patient characteristics indication for revision may be given.

Table 2.10 summarises patient characteristics for the 10,040 hip revision procedures undertaken in 2012. Compared with 2011, the patient demographics have largely remained unchanged, though there was a slight increase in the average age in each procedure type

80 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.10 Patient characteristics for hip revision procedures in 2012, according to procedure type.

Hip stage one Hip stage two Hip single- of two-stage of two-stage Hip excision Total revision stage revision revision revision arthroplasty procedures No. % No. % No. % No. % No. % Total revision procedures 8,812 88% 560 6% 611 6% 57 <1% 10,040 Number with patient data 8,475 96% 551 98% 599 98% 51 89% 9,676 96% Average age 70.0 69.5 68.9 71.3 69.8 SD 12.0 11.3 11.4 13.2 12.0 62.8 - 63.0 - 63.4 - 65.6 - 62.9 - Interquartile range 78.6 77.6 77.3 80.2 78.4 Gender Female 4,926 58% 269 49% 275 46% 27 53% 5,497 57% Male 3,549 42% 282 51% 324 54% 24 47% 4,179 43% Patient physical status P1 - fit and healthy 894 10% 28 5% 45 7% 2 4% 969 10% P2 - mild disease not 5,604 64% 330 59% 362 59% 20 35% 6,316 63% incapacitating P3 - incapacitating systemic 2,180 25% 186 33% 194 32% 28 49% 2,588 26% disease P4 and P5 134 2% 16 3% 10 2% 7 12% 167 2% Indications for surgery Aseptic loosening 3,881 44% 62 11% 54 9% 11 19% 4,008 40% Lysis 1,227 14% 44 8% 32 5% 8 14% 1,311 13% Pain 2,219 25% 72 13% 51 8% 9 16% 2,351 23% Dislocation/subluxation 1,207 14% 22 4% 19 3% 9 16% 1,257 13% Periprosthetic fracture 797 9% 15 3% 27 4% 1 2% 840 8% © National Joint Registry 2013 Infection 288 3% 455 81% 443 73% 36 63% 1,222 12% Malalignment 450 5% 6 1% 4 <1% 0 0% 460 5% Fractured acetabulum 97 1% 3 <1% 0 0% 1 2% 101 1% Fractured stem 151 2% 4 <1% 3 <1% 0 0% 158 2% Fractured femoral head 28 <1% 0 0% 0 0% 0 0% 28 <1% Incorrect sizing head/socket 61 <1% 0 0% 0 0% 1 2% 62 <1% Wear of acetabular 1,177 13% 12 2% 10 2% 0 0% 1,199 12% component Dissociation of liner 133 2% 7 1% 16 3% 2 4% 158 2% Adverse soft tissue reaction 1,286 15% 25 4% 19 3% 0 0% 1,330 13% Other 819 9% 19 3% 57 9% 4 7% 899 9% Side Bilateral 4 <1% 0 0% 0 0% 0 0% 4 <1% Left, unilateral 4,078 46% 285 51% 303 50% 26 46% 4,692 47% Right, unilateral 4,730 54% 275 49% 308 50% 31 54% 5,344 53%

www.njrcentre.org.uk 81 Table 2.11 Indication for surgery for hip revision procedures 2008 to 2012.

2008 2009 2010 2011 2012 Total No. % No. % No. % No. % No. % No. % Indications for single-stage revision Number of procedures 6,421 17% 6,610 18% 7,375 20% 8,201 22% 8,812 24% 37,419 Aseptic loosening 3,809 59% 3,654 55% 3,700 50% 3,776 46% 3,881 44% 18,820 50% Pain 1,746 27% 2,032 31% 2,000 27% 2,114 26% 2,219 25% 10,111 27% Lysis 1,108 17% 997 15% 1,104 15% 1,112 14% 1,227 14% 5,548 15% Adverse soft tissue reaction to particle 1 <1% 86 1% 410 6% 982 12% 1,286 15% 2,765 7% debris Infection 173 3% 192 3% 236 3% 264 3% 288 3% 1,153 3% Indications for stage one of a two-stage revision Number of © National Joint Registry 2013 procedures 456 17% 554 21% 529 20% 549 21% 560 21% 2,648 Infection 366 80% 440 79% 422 80% 448 82% 455 81% 2,131 80% Pain 87 19% 103 19% 110 21% 78 14% 72 13% 450 17% Aseptic loosening 89 20% 84 15% 72 14% 56 10% 62 11% 363 14% Lysis 58 13% 50 9% 52 10% 43 8% 44 8% 247 9%

2.2.2.2 Components removed and components were more likely to be removed during components used a two-stage revision process than during a single- stage revision. This is expected since the majority Both the acetabular and femoral components were of two-stage revisions are carried out for reasons of removed in approximately half of all revision procedures infection, where all components are routinely removed. (Table 2.12). However, comparison of the different The components used during single-stage revision types of revision procedures indicates that both procedures are shown in Table 2.13.

Table 2.12 Components removed during hip revision procedures in 2012.

Hip single-stage Hip stage one of a Hip excision revision two-stage revision arthroplasty Total No. % No. % No. % No. % Total 8,812 560 57 9,429 Both cup and stem 3,986 45% 470 84% 40 70% 4,496 48% Acetabular cup only 2,662 30% 23 4% 3 5% 2,688 29% Femoral stem only 1,275 14% 31 6% 10 18% 1,316 14% Neither cup nor stem 889 10% 36 6% 4 7% 929 10% © National Joint Registry 2013

82 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.13 Components used during single-stage hip revision procedures in 2012.

Hip single-stage revision

Number of procedures % Total 8,812 Femoral prosthesis Cemented 2,482 28% Cementless 2,569 29% Not revised 3,761 43% Acetabular prosthesis Cemented 1,569 18% © National Joint Registry 2013 Cementless 5,090 58% Not revised 2,153 24%

www.njrcentre.org.uk 83 Part 2 2.3 Knee replacement procedures 2012 National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

The total number of knee replacement procedures As a percentage of their activity, ISTCs performed entered into the NJR during 2012 was 90,842, more unicondylar knee replacement procedures an increase of 7.3% compared with 2011. Of the (Figure 2.16) than any other type of provider and NHS 90,842 procedures submitted, 84,833 were primary hospitals performed more cemented bicondylar knee procedures and 6,009 were revision procedures. procedures than any other provider. The revision Table 2.14 summarises the patient characteristics and procedures undertaken at NHS hospitals comprised details of knee replacement procedures according to 85% of all revision procedures performed. type of provider.

Table 2.14 Procedure details, according to type of provider for knee procedures in 2012.

Independent NHS hospitals hospitals ISTC Total No. % No. % No. % No. % Total knee procedures 63,679 70% 23,501 26% 3,662 4% 90,842 Patient physical status P1 - fit and healthy 5,320 8% 3,873 16% 369 10% 9,562 11% P2 - mild disease not 45,825 72% 17,784 76% 3,015 82% 66,624 73% incapacitating P3 - incapacitating systemic 12,217 19% 1,829 8% 276 8% 14,322 16% disease P4 and P5 317 <1% 15 <1% 2 <1% 334 <1% Procedure type Primary procedures 58,556 69% 22,749 27% 3,528 4% 84,833 93% Total prosthetic replacement 51,703 88% 18,597 82% 3,040 86% 73,340 86% using cement Total prosthetic replacement not 1,823 3% 887 4% 1 <1% 2,711 3% using cement Hybrid total knee 335 <1% 103 <1% 8 <1% 446 <1% Patello-femoral replacement 815 1% 420 2% 36 1% 1,271 1% Unicondylar knee replacement 3,880 7% 2,742 12% 443 13% 7,065 8% Revision procedures 5,123 85% 752 13% 134 2% 6,009 7% Knee single-stage revision 3,900 76% 657 87% 118 88% 4,675 78% Knee stage one of two-stage © National Joint Registry 2013 548 11% 42 6% 11 8% 601 10% revision Knee stage two of two-stage 649 13% 52 7% 5 4% 706 12% revision Knee conversion to arthrodesis 20 <1% 1 <1% 0 0% 21 <1% Amputation 6 <1% 0 0% 0 0% 6 <1% Bilateral or unilateral8 Bilateral 668 1% 374 2% 46 1% 1,088 1% Unilateral 63,011 99% 23,127 98% 3,616 99% 89,754 99% Funding Independent 511 <1% 8,995 38% 16 <1% 9,522 10% NHS 63,168 99% 14,506 62% 3,646 100% 81,320 90%

8 Bilaterals will only be counted as a bilateral if they are entered under the same single operation during data entry. If the two procedures are recorded under two different operations they will be counted as two unilateral procedures. Therefore, the count of bilaterals is likely to be an underestimate.

www.njrcentre.org.uk 85 Figure 2.16

Primary knee procedures by type of provider 2012.

100%

80%

60%

40%

20% Percentage of procedure type of procedure Percentage

Patello 1% 2% 1% -femoral Unicondylar 7% 12% 13% Number of procedures 58,556 22,749 3,528

2.3.1 Primary knee replacement replacements where the meniscal implant has been specified. The usage of unconstrained fixed implants procedures 2012 has increased gradually over the past six years at the Of the 84,883 primary knee replacements undertaken in expense of unconstrained mobile constructs. Figure 2012, 76,497 (90%) were bicondylar procedures (TKR), 2.19 shows that the usage of fixed constraint implants 7,065 (8%) were unicondylar knee replacements and has increased since 2006. 1,271 (1%) were patello-femoral replacements (Table The single largest indication recorded for surgery 2.15). Compared with previous years, these proportions was osteoarthritis, recorded in 98% of all primary have largely remained the same (Figure 2.17) though procedures (Table 2.15). All other indications were there has been a slight increase in cemented TKR at recorded at 1% or less. the expense of cementless TKR over the past four years. Figure 2.18 is based on total condylar knee

86 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.15 Patient characteristics for primary knee replacement procedures in 2012, according to procedure type.

Primary total Primary total prosthetic Primary total prosthetic replacement prosthetic replacement not classified Patello- replacement not using elsewhere femoral Unicondylar using cement cement (e.g. hybrid) replacement replacement Total No. % No. % No. % No. % No. % No. % Total knee primaries 73,340 86% 2,711 3% 446 <1% 1,271 1% 7,065 8% 84,833 Patient physical status P1 - fit and healthy 6,881 9% 280 10% 46 10% 309 24% 1,574 22% 9,090 11% P2 - mild disease not 54,368 74% 2,092 77% 324 73% 864 68% 4,947 70% 62,595 74% incapacitating P3 - incapacitating 11,840 16% 333 12% 74 17% 97 8% 536 8% 12,880 15% systemic disease P4 and P5 251 <1% 6 <1% 2 <1% 1 <1% 8 <1% 268 <1% BMI Number with BMI 51,084 70% 1,577 58% 306 69% 865 68% 5,282 75% 59,114 70% data Average 30.92 31.02 31.25 30.02 30.00 30.83 SD 5.54 5.13 5.35 5.17 4.81 5.47 Indications for surgery Osteoarthritis 71,384 97% 2,683 99% 428 96% 1,238 97% 6,987 99% 82,720 98% © National Joint Registry 2013 Avascular necrosis 223 <1% 3 <1% 3 <1% 3 <1% 55 <1% 287 <1% Inflammatory 500 <1% 11 <1% 2 <1% 2 <1% 4 <1% 519 <1% arthropathy Previous infection 64 <1% 2 <1% 0 0% 0 0% 1 <1% 67 <1% Rheumatoid arthritis 1,108 2% 15 <1% 4 <1% 1 <1% 5 <1% 1,133 1% Previous trauma 413 <1% 9 <1% 8 2% 5 <1% 25 <1% 460 <1% Other 507 <1% 9 <1% 9 2% 32 3% 45 <1% 602 <1% Side Bilateral 667 <1% 24 <1% 7 2% 90 7% 292 4% 1,080 1% Left, unilateral 34,293 47% 1,269 47% 203 46% 504 40% 3,373 48% 39,642 47% Right, unilateral 38,380 52% 1,418 52% 236 53% 677 53% 3,400 48% 44,111 52%

www.njrcentre.org.uk 87 Figure 2.17

Type of primary knee replacement procedures undertaken between 2006 and 2012.

100%

80%

60%

40%

20% Percentage of procedures Percentage

0% Year 2006 2007 2008 2009 2010 2011 2012 © National Joint Registry 2013 TKR using cement 83% 83% 83% 83% 85% 86% 86% TKR not using 7% 6% 6% 6% 5% 4% 3% cement TKR Hybrid 1% 1% 1% 1% <1% <1% <1% Patello-femoral 1% 1% 1% 1% 1% 1% 1% Unicondylar 8% 8% 8% 8% 8% 8% 8% Number of procedures 62,430 73,767 77,754 79,071 81,427 84,230 84,833

88 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.18

Implant constraint for bicondylar primary knee replacement procedures between 2006 and 2012.

80%

60%

40%

20% Percentage of procedures Percentage

0% Year 2006 2007 2008 2009 2010 2011 2012

Constrained <1% <1% <1% <1% <1% <1% <1% condylar

Posterior 24% 24% 24% 25% 25% 25% 23% © National Joint Registry 2013 stabilised, xed

Posterior 2% 2% 2% 2% 2% 1% 1% stabilised, mobile Unconstrained, 62% 63% 64% 65% 66% 67% 70% xed Unconstrained, 11% 11% 10% 8% 7% 6% 5% mobile Hinged/ <1% <1% <1% <1% <1% <1% <1% linked

Number of 69,852 71,956 74,787 76,140 procedures 54,968 64,884 68,502

www.njrcentre.org.uk 89 Figure 2.19 Bearing type for unicondylar implant used in primary knee replacement procedures undertaken between 2006 and 2012.

80% 70% 60% 50% 40% 30% 20%

Percentage of procedures Percentage 10% © National Joint Registry 2013 0% Year 2006 2007 2008 2009 2010 2011 2012 Fixed 25% 23% 24% 26% 30% 31% 36% Mobile 75% 77% 76% 74% 70% 69% 64% Number of procedures* 5,723 6,639 7,017 7,105 7,460 7,260 7,095

* This is the number of procedures using a unicondylar meniscal implant regardless of patient procedure selected

2.3.1.1 Patient characteristics number of patients assessed as being fit and healthy at the time of operation and an increase in P2 and P3 According to the ASA grade system, 11% of patients status of patients. Figure 2.22 shows the increase in BMI9 undergoing a primary knee replacement procedure were over the past nine years for patients having primary knee graded as fit and healthy (Table 2.15). procedures. This figure has progressively increased from 29.2 to 30.8 over the period. There has been a slight The average age of patients was 69.3 years and 57% decrease in BMI for female patients whereas the BMI of were female. Patients undergoing a patello-femoral male patients has continued to increase. It also shows replacement were the youngest, at an average age of that there has been a steady increase in the number of 59.7 years and 73% of these were female (Table 2.16). patients within the BMI range 30 to 34 and 35 to 39 and On average, female patients were of a similar age to male a decrease within the ranges 20 to 24 and 25 to 29. patients at the time of their primary knee replacement The average knee replacement patient in 2012, by BMI (69.5 years and 68.9 years respectively), see Table measurement, was clinically obese. It is interesting to 2.16. However, female patients were, on average, note that the profile of Figure 2.22 is significantly different older than male patients for cementless, cemented and to the equivalent chart for hips, Figure 2.7. hybrid procedures but younger for patello-femoral and unicondylar procedures. However, only 16% of entries in 2004 had BMI data entered and while this has increased to 70% in 2012 all Figure 2.21 shows the trend in ASA grade over the BMI data has to be viewed with caution as surgeons may past ten years. Since 2003, similar to the data trend for be more likely to enter BMI data when the BMI is high, total hip replacement, there has been a reduction in the introducing an element of bias.

9 BMI: 20-24 normal, 25-29 overweight, 30-39 obese, 40+ morbidly obese.

90 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.16 Age and gender for primary knee replacement patients in 2012.

Primary total Primary total prosthetic Primary total prosthetic replacement prosthetic replacement not classified Patello- replacement not using elsewhere femoral Unicondylar using cement cement (e.g. hybrid) replacement replacement Total No. % No. % No. % No. % No. % No. % Total knee primaries 73,340 86% 2,711 3% 446 <1% 1,271 1% 7,065 8% 84,833 Total knee primaries 71,022 97% 2,674 99% 429 96% 1,207 95% 6,659 94% 81,991 97% with patient data Average Age 70.01 68.51 68.24 59.74 63.59 69.30 SD 9.36 9.51 10.57 11.74 9.92 9.68 63.9 - 62.0 - 61.1 - 51.3 - 56.5 - 63.0 - Interquartile Range 76.8 75.6 75.9 67.8 70.2 76.3 Average age by gender Female 41,007 58% 1,318 49% 215 50% 888 74% 3,064 46% 46,492 57% Average age 70.26 68.91 68.60 59.05 62.95 69.50 SD 9.49 9.66 11.34 11.60 10.13 9.87 64.0 - 62.1 - 60.0 - 50.6 - 55.6 - 63.0 - Interquartile range 77.2 76.3 76.7 67.1 70.0 76.7 Male 30,015 42% 1,356 51% 214 50% 319 26% 3,595 54% 35,499 43% Average age 69.69 68.15 67.98 61.64 64.16 68.90 SD 9.18 9.35 9.80 11.98 9.72 9.45 63.8 - 61.9 - 61.9 - 53.7 - 57.3 - 63.0 - Interquartile range 76.2 75.1 74.9 69.3 70.0 75.7 Age group by gender

Female © National Joint Registry 2013 <30 years 16 <1% 1 <1% 0 0% 2 <1% 3 <1% 22 <1 % 30 - 39 years 60 <1% 1 <1% 1 <1% 27 3% 15 <1% 104 <1 % 40 - 49 years 805 2% 38 3% 11 5% 170 19% 307 10% 1,331 3% 50 - 59 years 5,025 12% 192 15% 39 18% 296 33% 887 29% 6,439 14% 60 - 69 years 13,503 33% 464 35% 66 31% 226 25% 1,083 35% 15,342 33% 70 - 79 years 15,145 37% 455 35% 61 28% 127 14% 622 20% 16,410 35% 80 - 89 years 6,187 15% 163 12% 33 15% 40 5% 146 5% 6,569 14% 90+ 266 <1% 4 <1% 4 2% 0 0% 1 <1% 275 <1% Male <30 years 6 <1% 0 0% 0 0% 2 <1% 0 0% 8 <1% 30 - 39 years 36 <1% 1 <1% 2 <1% 6 2% 16 <1% 61 <1% 40 - 49 years 653 2% 44 3% 6 3% 43 13% 248 7% 994 3% 50 - 59 years 3,665 12% 209 15% 28 13% 98 31% 922 26% 4,922 14% 60 - 69 years 10,756 36% 508 37% 91 43% 94 29% 1,449 40% 12,898 36% 70 - 79 years 11,023 37% 456 34% 63 29% 54 17% 767 21% 12,363 35% 80 - 89 years 3,737 12% 136 10% 24 11% 19 6% 187 5% 4,103 12% 90+ 139 <1% 2 <1% 0 0% 3 <1% 6 <1% 150 <1%

www.njrcentre.org.uk 91 Figure 2.20

Age and gender for primary knee replacement patients in 2012.

Female 100%

80%

60%

40%

20% Percentage of procedures Percentage

0% Age group <30 30-39 40-49 50-59 60-69 70-79 80-89 90+ TKR using cement 73% 58% 60% 78% 88% 92% 94% 97% TKR not using cement 5% <1% 3% 3% 3% 3% 2% 1% TKR hybrid 0% <1% <1% <1% <1% <1% 1% 1% Patello-femoral 9% 26% 13% 5% 1% <1% <1% 0% Unicondylar 14% 14% 23% 14% 7% 4% 2% <1% Number of patients 22 104 1,331 6,439 15,342 16,410 6,569 275

Male © National Joint Registry 2013 100%

80%

60%

40%

20% Percentage of procedures Percentage

0% Age group <30 30-39 40-49 50-59 60-69 70-79 80-89 90+ TKR using cement 75% 59% 66% 74% 83% 89% 91% 93% TKR not using cement 0% 2% 4% 4% 4% 4% 3% 1% TKR hybrid 0% 3% <1% <1% <1% <1% <1% 0% Patello-femoral 25% 10% 4% 2% <1% <1% <1% 2% Unicondylar 0% 26% 25% 19% 11% 6% 5% 4% Number of patients 8 61 994 4,922 12,898 12,363 4,103 150

92 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.21

ASA grades for primary knee replacement patients between 2003 and 2012.

100%

80%

60%

40%

20% Percentage of procedures Percentage 0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 P1 31% 25% 20% 17% 15% 13% 13% 12% 11% 11% P2 58% 62% 65% 68% 71% 73% 72% 73% 73% 74% P3 10% 13% 14% 14% 14% 13% 15% 15% 15% 15% P4 and P5 <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% Number of procedures 24,665 46,596 60,767 62,430 73,767 77,754 79,071 81,427 84,230 84,833

www.njrcentre.org.uk 93 Figure 2.22 BMI for primary knee replacement patients between 2004 and 2012.

32 50% 45% 31 40% 30 35% 30% 29 25% BMI 28 20% 27 15%

10% of patients Percentage 26 5% 25 0% Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 Average BMI - all patients 29.25 29.45 29.53 29.85 30.28 30.54 30.66 30.81 30.83 BMI 15-19 1% 1% 1% <1% <1% <1% <1% <1% <1% BMI 20-24 17% 14% 15% 14% 12% 11% 10% 10% 10%

© National Joint Registry 2013 BMI 25-29 38% 40% 38% 37% 36% 36% 35% 35% 34% BMI 30-34 28% 29% 29% 30% 31% 32% 32% 32% 33% BMI 35-39 11% 12% 11% 13% 14% 15% 15% 16% 16% BMI 40-44 3% 3% 3% 4% 4% 5% 5% 5% 5% BMI 45+ 1% 1% 1% 1% 1% 2% 2% 2% 2% Average BMI 30.01 29.96 29.97 30.32 30.73 31.04 31.16 31.34 31.29 - female Average BMI 28.77 28.91 28.96 29.25 29.69 29.88 30.03 30.14 30.23 - male Percentage of procedures 16% 15% 17% 22% 49% 57% 61% 66% 70% with BMI Number of procedures 5,509 9,101 10,579 16,110 38,207 45,200 49,930 55,409 59,114 with BMI

2.3.1.2 Surgical techniques Compared with previous years, the surgical techniques used in primary knee replacements have largely The most common surgical approach was the medial remained unchanged. However, there has been parapatellar, used in 93% of procedures (Table 2.17). an increase in the use of MIS in unicondylar knee Minimally-invasive surgery (MIS) was used in 46% of replacements, from 37% in 2004 to 46% in 2012. unicondylar knee replacement procedures, reflecting the popularity of the Oxford Partial Knee, but was The use of bone cement in primary knee procedures is used in only 2% of all other types of knee replacement summarised in Figure 2.23. intervention. For cemented knee procedures, 38% had the patella replaced at the time of the primary procedure whereas only 7% of patellas were replaced during primary cementless knee procedures.

94 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.17 Characteristics of surgical practice for primary knee replacement procedures in 2012, according to procedure type.

Primary total Primary total prosthetic Primary total prosthetic replacement prosthetic replacement not classified replacement not using elsewhere Patello-femoral Unicondylar using cement cement (e.g. hybrid) replacement replacement Total No. % No. % No. % No. % No. % No. % Total knee 73,340 86% 2,711 3% 446 <1% 1,271 1% 7,065 8% 84,833 primaries Surgical approach Lateral 590 <1% 20 <1% 8 2% 14 1% 242 3% 874 1% parapatellar Medial 68,359 93% 2,583 95% 352 79% 1,178 93% 6,193 88% 78,665 93% parapatellar Mid-Vastus 2,317 3% 54 2% 19 4% 35 3% 268 4% 2,693 3% Sub-Vastus 1,281 2% 10 <1% 64 14% 22 2% 209 3% 1,586 2% Other 793 1% 44 2% 3 <1% 22 2% 153 2% 1,015 1% Patella Patella 27,819 38% 190 7% 123 28% 1,231 97% 60 1% 29,423 35% implanted Patella not 45,521 62% 2,521 93% 323 72% 40 3% 7,005 99% 55,410 65% implanted © National Joint Registry 2013 Minimally-invasive surgery Yes 1,750 2% 99 4% 7 2% 158 12% 3,222 46% 5,236 6% No 71,590 98% 2,612 96% 439 98% 1,113 88% 3,843 54% 79,597 94% Image-guided surgery Yes 2,169 3% 206 8% 25 6% 3 <1% 67 1% 2,470 3% No 71,171 97% 2,505 92% 421 94% 1,268 100% 6,998 99% 82,363 97% Bone graft used - femur Yes 521 <1% 15 <1% 3 <1% 0 0% 22 <1% 561 <1% No 72,819 99% 2,696 99% 443 99% 1,271 100% 7,043 100% 84,272 99% Bone graft used - tibia Yes 267 <1% 12 <1% 4 <1% 1 <1% 7 <1% 291 <1% No 73,073 100% 2,699 100% 442 99% 1,270 100% 7,058 100% 84,542 100%

www.njrcentre.org.uk 95 Figure 2.23 Bone cement types for primary knee replacement procedures undertaken between 2003 and 2012.

90% 80% 70% 60% 50% 40% 30% 20%

Percentage of procedures Percentage 10% 0% Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Antibiotic-loaded high viscosity 82% 84% 86% 87% 89% 91% 91% 91% 94% 95%

© National Joint Registry 2013 Antibiotic-loaded 3% 3% 3% 2% 3% 2% 2% 3% 2% 3% medium viscosity Antibiotic-loaded 2% 3% 3% 2% 1% <1% <1% <1% <1% <1% low viscosity High viscosity 6% 7% 4% 5% 4% 4% 4% 3% 2% <1% Medium viscosity 6% 4% 3% 4% 3% 2% 3% 2% 2% 1% Low viscosity <1% <1% <1% <1% <1% <1% <1% <1% <1% 0% Number of procedures 18,396 36,472 50,062 52,656 64,584 69,710 71,979 75,702 78,929 80,164 using cement

2.3.1.3 Thromboprophylaxis to 92% in 2012. There has been a marked decrease over the past two years in the use of aspirin (from Table 2.18 shows that the most frequently prescribed 20% in 2009 to 8% in 2012). Direct thrombin inhibitor chemical method of thromboprophylaxis for knee is now used in 13% of knee primary procedures and replacement patients was LMWH (72%), while TED the use of ‘other chemical’ has gone up from 7% in stockings were the most used mechanical method 2009 to 12% in 2012. This change was also seen (70%). Compared with previous years, there has in hip primary procedures. Less than 1% of patients been an increase in the prescription of a combined had neither mechanical nor chemical-prescribed chemical and mechanical regime, from 49% in 2004 thromboprophylaxis.

96 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.18 Thromboprophylaxis regime for primary knee replacement patients, prescribed at time of operation.

Total

No. % Totals 84,833 Aspirin 6,568 8% Low molecular weight Heparin 61,347 72%

Pentasaccharide 1,265 1%

Warfarin 559 <1% Direct thrombin inhibitor 10,731 13% Other chemical (all) 10,039 12% No chemical 3,118 4% Foot pump 22,482 27% Intermittent calf compression 36,027 42% TED stockings 59,351 70% © National Joint Registry 2013 Other 1,177 1% No mechanical 3,867 5% Both mechanical and chemical 77,926 92% Neither mechanical nor chemical 47 <1%

2.3.1.4 Untoward intra-operative events

Table 2.19 shows that untoward intra-operative events were rare, reported in less than 1% of knee procedures. There were fewer instances reported in 2012 compared to 2011.

www.njrcentre.org.uk 97 Table 2.19 Reported untoward intra-operative events for primary knee replacement patients in 2012, according to procedure type.

2.3.1.5 Knee primary components marketed by DePuy, continues to dominate the market. The Triathlon, Genesis 2 and Vanguard appear to be Figure 2.24 shows the leading brands of total condylar increasing in popularity. knees in England and Wales. The PFC Sigma knee,

Figure 2.24

Top five total condylar knee brands, trends 2003 to 2012.

40%

30%

20%

10% Percentage of procedures

PFC Sigma Bicondylar Knee Nexgen Triathlon Genesis 2 Vanguard

98 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Likewise, the market for unicondylar knees is dominated More information on the other brands used can be by one product, the Oxford Partial Knee (Figure 2.25). found in the document ‘Prostheses used in hip, knee, The market share of the Oxford Partial has decreased ankle, elbow and shoulder replacement procedures gradually since 2003 and the Sigma HP, still relatively 2012’ which can be downloaded from the NJR website. new to the market, and the Zimmer Uni continue to be the next most used brands of unicondylar knee system.

Figure 2.25

Top five unicondylar knee brands, trends 2003 to 2012.

80% 70% 60% 50% 40% 30% 20%

Percentage of procedures 10% © National Joint Registry 2013 0% Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 2,226 4,261 5,439 5,750 6,659 7,083 7,181 7,458 7,269 7,078

Oxford Partial Knee Zimmer Uni Sigma HP AMC/Uniglide Triathlon Uni

The brand usage for patello-femoral prostheses are highly constrained and hinged revision knees is shown shown in Figure 2.26 and the equivalent graph for in Figure 2.27.

www.njrcentre.org.uk 99 Figure 2.26

Top five patello-femoral knee brands, trends 2003 to 2012.

80% 70% 60% 50% 40% 30% 20% 10% Percentage of procedures

Avon Zimmer PFJ FPV Sigma HP Journey PFJ Oxinium

100 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 2.27

Top five fixed hinged knee brands, trends 2003 to 2012.

60%

50%

40%

30%

20%

10% Percentage of procedures

0% © National Joint Registry 2013 Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Number of components used 70 129 195 213 262 282 307 305 306 294

Endo Rotating Nexgen Hinge MRH RT-Plus Noiles Hinge Type

2.3.2 Knee revision procedures 2012 and 706 (12%) were stage two of a two-stage revision (Table 2.20). A further 27 procedures were A total of 6,009 knee revision procedures were recorded, comprising 21 conversions of previous knee reported, an increase of 17% on 2011. Of these, replacements to arthrodesis and six knee amputations. 4,675 (78%) were single-stage revision procedures, Compared with previous years, there has been no 601 (10%) were stage one of a two-stage revision change in the types of revision procedures carried out.

www.njrcentre.org.uk 101 Table 2.20 Patient characteristics for knee revision procedures in 2012, according to procedure type.

Knee stage one Knee stage Knee Knee single- of two-stage two of two- conversion to Knee stage revision revision stage revision arthrodesis Amputation Total No. % No. % No. % No. % No. % No. % Total 4,675 78% 601 10% 706 12% 21 <1% 6 <1% 6,009 Total with patient data 4,546 97% 583 97% 687 97% 19 90% 6 100% 5,841 97% Average age 68.76 69.74 69.87 71.71 70.62 68.97 SD 10.47 10.28 9.49 9.80 12.25 10.64 61.77 - 63.41 - 63.60 - 67.55 - 68.53 - 62 .09 - Interquartile range 76.33 77.20 77.12 76.96 78.45 76.48 Gender Female 2,466 54% 242 42% 295 43% 10 53% 5 83% 3,018 52% Male 2,080 46% 341 58% 392 57% 9 47% 1 17% 2,823 48% Patient physical status P1 - fit and healthy 408 9% 31 5% 33 5% 0 0% 0 0% 472 8% P2 - mild disease not incapacitating 3,217 69% 366 61% 434 61% 10 48% 2 33% 4,029 67% P3 - incapacitating systemic disease 1,005 21% 192 32% 230 33% 11 52% 4 67% 1,442 24% P4 and P5 45 1% 12 2% 9 1% 0 0% 0 0% 66 1% Indications for surgery Aseptic loosening 1,778 38% 69 11% 84 12% 1 5% 0 0% 1,932 32% Pain 846 18% 27 4% 31 4% 2 10% 0 0% 906 15% Lysis 405 9% 59 10% 40 6% 1 5% 0 0% 505 8% Wear of polyethylene component 563 12% 14 2% 14 2% 0 0% 0 0% 591 10% © National Joint Registry 2013 Instability 792 17% 24 4% 31 4% 2 10% 0 0% 849 14% Infection 255 5% 512 85% 541 77% 14 67% 4 67% 1,326 22% Malalignment 368 8% 7 1% 9 1% 0 0% 0 0% 384 6% Stiffness 297 6% 12 2% 9 1% 2 10% 0 0% 320 5% Progressive arthritis remaining 540 12% 3 <1% 10 1% 0 0% 0 0% 553 9% Dislocation/subluxation 188 4% 5 <1% 14 2% 0 0% 0 0% 207 3% Periprosthetic fracture 187 4% 8 1% 4 <1% 1 5% 0 0% 200 3% Component dissociation 86 2% 2 <1% 4 <1% 0 0% 0 0% 92 2% Implant fracture 49 1% 2 <1% 2 <1% 0 0% 0 0% 53 <1% Other 493 11% 24 4% 54 8% 2 10% 4 67% 577 10% Side Bilateral 8 <1% 0 0% 0 0% 0 0% 0 0% 8 <1% Left, unilateral 2,168 46% 299 50% 347 49% 13 62% 2 33% 2,829 47% Right, unilateral 2,499 53% 302 50% 359 51% 8 38% 4 67% 3,172 53%

2.3.2.1 Patient characteristics single-stage revision (38%) and infection was the most common indication for two-stage revision, conversion The mean age of knee revision patients was 69 years to arthrodesis and amputation. (Table 2.20). The average has decreased by 0.5 years compared with 2011. There were more female (52%) Compared with previous years, the patient than male patients (48%) which is similar to 2011. characteristics described above have largely remained Aseptic loosening was the most common indication for the same.

102 www.njrcentre.org.uk Part 2

2.4 Ankle replacement procedures 2012 The NJR started recording primary and revision total procedures performed were reported as being ankle replacements on 1 April 2010. By reviewing uncemented but the use of cement was reported in five submitted procedures against ankle levy submissions, cases, three of which pertained to a hybrid procedure. we have calculated compliance of ankle joint NJR submissions to be 77%, which is improving on previous 2.4.1.1 Patient characteristics years but further work is required. The average age of female patients was 64.7 years A total of 590 ankle replacement proformas, comprising whereas the average age for a male patient was 540 primary and 50 revision procedures carried out 68.8 years. 58% of patients were male (Table 2.22). between 1 January and 31 December 2012, were The BMI average was 29.0, which is higher than for submitted to the NJR by 28 February 2013. Due to the hip primary procedures but lower than knee primary small number collected so far the procedures tables in procedures. No bilateral procedures were submitted this section are displayed at a summary level only. Of to the NJR and 55% of procedures were performed all the ankle procedures carried out 86% were funded on the right ankle. 84% of patients had their procedure by the NHS. 84% of patients were classified as P1 performed due to osteoarthritis and 13% due to – fit and healthy (15%) or P2 – had mild disease not inflammatory arthritis. The pre-operative range of incapacitating (69%). (Table 2.21) movement and degrees of deformity can be seen in Table 2.21. 2.4.1 Primary ankle replacement procedures 2012

Of the 540 primary procedures, 414 (77%) were performed in the NHS, 107 (20%) in the independent sector and 19 (3%) in ISTCs. Almost all of the primary

104 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.21 Patient characteristics for primary ankle replacement procedures in 2012.

Primary procedures No. % Total ankle primaries 540 Patient physical status P1 - fit and healthy 81 15% P2 - mild disease not incapacitating 371 69% P3 - incapacitating systemic disease 87 16% P4 and P5 1 <1% Indications for surgery Osteoarthritis 455 84% Rheumatoid arthritis 56 10% Other inflammatory arthropathy 17 3% Other 20 4% Tibia-hindfoot alignment Physiological neutral 247 46% 5-15° Varus 126 23% 16-30° Varus 26 5% >30° Varus 2 <1% 5-15° Valgus 71 13% 16-30° Valgus 15 3% © National Joint Registry 2013 >30° Valgus 3 <1% Not available 50 9% Pre-operative range of movement ankle dorsiflexion 5-20° 229 42% Neutral 208 39% Fixed equinus 54 10% Not available 49 9% Pre-operative range of movement ankle plantarflexion 5-15° 292 54% 16-45° 185 34% Not available 63 12%

www.njrcentre.org.uk 105 Table 2.22 Age and gender for primary ankle replacement patients in 2012.

Primary procedures No. % Total ankle primaries 540 Total ankle primaries with patient data 519 96% Female age 218 42% Average 64.7 SD 12.1 Interquartile range 57.2 - 73.3 Male age 301 58% Average 68.8 SD 9.4 Interquartile range 63.5 - 76.1 Female age groups <45 years 18 8% 45 - 54 years 27 12% 55 - 64 years 51 23% © National Joint Registry 2013 65 - 74 years 79 36% 75 - 84 years 40 18% >85 years 3 1% Male age groups <45 years 6 2% 45 - 54 years 20 7% 55 - 64 years 64 21% 65 - 74 years 130 43% 75 - 84 years 75 25% >85 years 6 2%

2.4.1.2 Surgical techniques Achilles tendon lengthening was performed in 11% of procedures and subtalar joint fusion in 6%. Bone graft Table 2.23 details the surgical technique used during was used in 15% of procedures. ankle primary procedures. Additional ankle related procedures were performed in 36% of procedures.

106 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.23 Characteristics of surgical practice for primary ankle replacement procedures in 2012.

Primary procedures No. % Total ankle primaries 540 Incision Anterior 523 97% Anterolateral 8 1% Lateral (transfibular) 2 <1% Other 7 1% Associated procedures at time of surgery Subtalar joint fusion 30 6% Talonavicular fusion 19 4% Calcaneal displacement osteotomy 17 3% Achilles tendon lengthening 59 11% Fusion distal tibiofibular joint 1 <1% Fibula osteotomy 3 <1% Medial malleolar osteotomy 4 <1% © National Joint Registry 2013 Lateral ligament reconstruction 4 <1% Medial ligament reconstruction 2 <1% Other 77 14% None 348 64% Image-guided surgery Yes 6 1% No 534 99% Bone graft used Yes 79 15% No 461 85%

www.njrcentre.org.uk 107 2.4.1.3 Thromboprophylaxis only 2% used neither mechanical nor chemical regimes. LMWH was the most popular chemical Table 2.24 shows that 80% of primary ankle thromboprophylaxis regime used in 80% of total replacement procedures used both chemical and ankle replacement (TAR) procedures. mechanical thromboprophylaxis regimes and

Table 2.24 Thromboprophylaxis regime for primary ankle replacement patients, prescribed at time of operation.

Total No. % Total ankle primaries 540 Aspirin 26 5% Low molecular weight Heparin 430 80% Pentasaccharide 0 0% Warfarin 8 1% Direct thrombin inhibitor 16 3% Other chemical 62 11% No chemical 32 6% Foot pump 70 13% Intermittent calf compression 154 29% © National Joint Registry 2013 TED stockings 303 56% Other mechanical 17 3% No mechanical 87 16% Both mechanical and chemical 432 80% Neither mechanical nor chemical 11 2%

2.4.1.4 Untoward intra-operative events was ‘Fracture of the medial malleolus’ which occurred in 15 primary procedures (3%). In 4% of procedures an untoward intra-operative event was reported. Of those reported the most common

Table 2.25 Reported untoward intra-operative events for primary ankle replacement patients in 2012.

Total ankle primaries 540 Total events specified 24 Fracture of medial malleolus 15 Fracture of lateral malleolus 2 Fracture (other) 3 Ligament Injury 0 Other 4 © National Joint Registry 2013

108 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

2.4.1.5 Ankle primary components shoulder replacement procedures 2012’ which can be downloaded from the NJR website. The DePuy Mobility ankle prosthesis was used in 52% of all primary procedures recorded in 2012. The next 2.4.2 Ankle revision procedures 2012 most commonly used prosthesis was Corin’s Zenith ankle replacement at 20% followed by MatOrtho’s Of the 50 revision procedures, 46 were performed in BOX prosthesis at 8%. More information on the the NHS (92%). 66% were single-stage revisions and other brands used can be found in the document 22% were conversion to arthrodesis. ‘Prostheses used in hip, knee, ankle, elbow and

Table 2.26 Details for ankle revision procedures in 2012.

Procedure type Patient procedure No. 50 Single-stage revision Prosthetic replacement not classified elsewhere (e.g. hybrid) 4 Single-stage revision Prosthetic replacement not using cement 29 Stage one of two-stage revision 2 Stage two of two-stage revision Prosthetic replacement not classified elsewhere (e.g. hybrid) 1 Stage two of two-stage revision Prosthetic replacement not using cement 2 Stage two of two-stage revision Prosthetic replacement using cement 1 Conversion to arthrodesis Conversion to ankle fusion (subtalar joint not fused) 3 Conversion to arthrodesis Conversion to ankle & subtalar fusion (not using ttc nail) 4 © National Joint Registry 2013 Conversion to arthrodesis Conversion to ankle & subtalar fusion (using ttc nail) 4

2.4.2.1 Patient characteristics

The average age for a patient having a revision procedure was 64.7 years. Only 12% were fit and healthy and 24% had severe systemic disease (P3). 36% of revisions were for undiagnosed pain and 20% due to infection – low suspicion.

www.njrcentre.org.uk 109 Table 2.27 Patient characteristics for ankle revision procedures in 2012.

Revision procedures No. % Total 50 Patient physical status P1 - fit and healthy 6 12% P2 - mild disease not incapacitating 32 64% P3 - incapacitating systemic disease 12 24% P4 and P5 0 0% Indications for surgery Infection high suspicion 3 6% Infection low suspicion 10 20% Aseptic loosening - tibial 9 18% Aseptic loosening - talar 8 16% Lysis - tibia 4 8% Lysis - talus 2 4% Malalignment 9 18% Implant fracture - tibia 1 2% Implant fracture - talar 1 2% Implant fracture - meniscal 5 10% © National Joint Registry 2013 Wear of polyethylene component 6 12% Meniscal insert dislocation 0 0% Component migration/dissociation 2 4% Pain (undiagnosed) 18 36% Stiffness 4 8% Soft tissue impingement 3 6% Other 6 12% Side Bilateral 0 0% Left, unilateral 21 42% Right, unilateral 29 58%

110 www.njrcentre.org.uk Part 2 2.5 Elbow replacement procedures 2012 The NJR started recording primary and revision elbow independent sector. 88% of the primary procedures replacements on 1 April 2012. This Annual Report were total elbow replacement procedures and therefore deals with nine months of operation data (1 9% were radial head replacements only (Table April to 31 December 2012). 2.28). Primary procedures were mainly performed due to other inflammatory arthropathy (33%) and A total of 288 elbow replacement procedures, osteoarthritis (32%). comprising 214 primary and 74 revision procedures carried out between 1 April and 31 December 2.5.1.1 Patient characteristics 2012, were submitted to the NJR by 28 February 2013. Due to the small number collected so far the The average age of female patients was 66.9 years procedures tables in this section are displayed at a whereas the average age for a male patient was 65.6 summary level only. years. 67% of patients were female (Table 2.29). 83% of patients were right handed and 3% were 2.5.1 Primary elbow replacement ambidextrous. Table 2.29 shows that 54% have procedures 2012 (nine months) ASA grade of P2 and 34% have P3 – incapacitating systemic disease. Of the 214 primary procedures, 200 (93%) were performed in the NHS, and the remaining in the

Table 2.28 Details for primary elbow procedures in 2012 (nine months).

Primary procedures No. % Total elbow primaries 214 Procedure type Primary radial head replacement 20 9% Primary total prosthetic replacement 188 88% Primary lateral resurfacing 6 3% Primary indications Osteoarthritis 69 32% Other inflammatory arthropathy 70 33% Essex Lopresti 0 0% Avascular necrosis 0 0% Acute trauma 50 23% Trauma sequelae 29 14% Failed hemiarthroplasty 1 <1%

112 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.29 Patient characteristics for primary elbow replacement procedures in 2012 (nine months).

Primary procedures No. % Total elbow primaries 214 Patient physical status P1 - fit and healthy 25 12% P2 - mild disease not incapacitating 115 54% P3 - incapacitating systemic disease 73 34% P4 and P5 1 <1% Handedness Ambidextrous 6 3% Left 31 14% Right 177 83% Total elbow primaries with patient data 196 92% Female age 131 67% Average 66.9 SD 14.7 Interquartile range 59.6 - 77.4

Male age 65 33% © National Joint Registry 2013 Average 65.6 SD 13.9 Interquartile range 58.0 - 74.5 Age groups 196 <45 years 18 9% 45 - 54 years 20 10% 55 - 64 years 40 20% 65 - 74 years 61 31% 75 - 84 years 41 21% >85 years 16 8%

2.5.1.2 Surgical techniques less than 1% required an ulnar bone graft. 8% of procedures resulted in an untoward intra-operative Table 2.30 details the surgical technique used during event with fracture of the humerus occurring in 39% of elbow primary procedures. 89% of elbow procedures these and shaft penetration of the ulna in 33%. were performed using a posterior approach and less than 1% used image-guided surgery. 16% of replacements required humeral bone grafting but

www.njrcentre.org.uk 113 Table 2.30 Characteristics of surgical practice for primary elbow replacement procedures in 2012 (nine months).

Primary procedures No. % Total 214 Incision Kocher 24 11% Posterior 190 89% Image-guided surgery Yes 2 <1% No 212 99% Humeral bone graft used Yes 34 16% No 180 84% Ulnar bone graft used Yes 2 <1% No 212 99%

© National Joint Registry 2013 Untoward intra-operative events Procedures with none specified 198 Procedures with events specified 16 7% No. of events specified 18 Shaft penetration humerus 1 6% Shaft penetration ulna 6 33% Fracture humerus 7 39% Fracture ulna 1 6% Nerve injury 0 0% Vascular injury 0 0% Other 3 17%

2.5.1.3 Thromboprophylaxis both chemical and mechanical thromboprophylaxis regimes. 4% used neither mechanical nor chemical Table 2.31 shows that 90% of elbow replacement regimes. LMWH was the most popular chemical procedures used mechanical methods and 66% used thromboprophylaxis regime used in 59% of procedures. chemical methods; 60% of primary procedures used

114 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.31 Thromboprophylaxis regime for primary elbow replacement patients, prescribed at time of operation.

Total no. of primary procedures No. % Total 214 Aspirin 1 <1% Low molecular weight Heparin 126 59% Pentasaccharide 0 0% Warfarin 1 <1% Direct thrombin inhibitor 2 <1% Other chemical 11 5% No chemical 73 34% Foot pump 50 23% Intermittent calf compression 81 38%

TED stockings 116 54% © National Joint Registry 2013 Other mechanical 3 1% No mechanical 21 10% Both mechanical and chemical 129 60% Neither mechanical nor chemical 9 4%

2.5.2 Elbow revision procedures 2012 67% of the patients were female and their average age was 65.3. The average age for male patients Of the 74 revision procedures, 74% were single- undergoing revision was 63.9. stage revisions (Table 2.32). Aseptic loosening accounted for 50% of the reasons for revision.

www.njrcentre.org.uk 115 Table 2.32 Details for elbow revision procedures in 2012 (nine months).

Procedure type Patient procedure No. % Total 74 Single-stage revision Revision to lateral resurfacing 1 1% Single-stage revision Revision total prosthetic replacement 54 73% Stage one of two-stage revision 6 8% Stage two of two-stage revision Revision total prosthetic replacement 10 14% Excision arthroplasty 3 4% Indications for surgery Aseptic loosening 37 50% Infection 13 18% Instability 9 12% Periprosthetic fracture 14 19% Other 15 20% Side Left 32 43% Right 42 57% © National Joint Registry 2013 Total elbow revision procedures with patient data 73 99% Female age 49 67% Average 65.3 SD 13.6 Interquartile range 56.6 - 75.4 Male age 24 33% Average 63.9 SD 13.1 Interquartile range 55.0 - 72.4

2.5.3 Elbow components used in where only the radial head was replaced, the most used brand was the Anatomic Radial head primary and revision procedures manufactured by Acumed. More information on the The Zimmer Coonrad Morrey had 45% of the market other brands used can be found in the document share in total elbow replacement prostheses used ‘Prostheses used in hip, knee, ankle, elbow and in primary and revision procedures. For procedures shoulder replacement procedures 2012’ which can be downloaded from the NJR website.

116 www.njrcentre.org.uk Part 2 2.6 Shoulder replacement procedures 2012 The NJR started recording primary and revision shoulder independent sector. The most frequently performed replacements on 1 April 2012. This Annual Report primary procedure was the reverse polarity total shoulder therefore deals with nine months of operation data (1 replacement at 597 (30%) followed by the standard April to 31 December 2012). polarity total shoulder replacement with 525 (27%) procedures (Table 2.33). Primary procedures were mainly A total of 2,225 shoulder replacement procedures, performed due to osteoarthritis (61%) and cuff tear comprising 1,968 primary and 257 revision procedures arthropathy (24%). carried out between 1 April and 31 December 2012, were submitted to the NJR by 28 February 2013. Due to 2.6.1.1 Patient characteristics the small number collected so far the procedures tables in this section are displayed at a summary level only. The average age of female patients was 73.2 years whereas the average age for a male patient was 68.8 2.6.1 Primary shoulder replacement years. 72% of patients were female (Table 2.34). 84% of procedures 2012 (nine months) patients were right handed and 4% were ambidextrous. Table 2.34 shows that 62% had ASA grade of P2 and Of the 1,968 primary procedures, 1,630 (83%) were 28% have P3 – incapacitating systemic disease. performed in the NHS sector and the remaining in the

Table 2.33 Details for primary shoulder procedures in 2012 (nine months).

Primary procedures No. % Total shoulder primaries 1,968 Procedure type Primary total prosthetic replacement 525 27% Primary hemi-arthroplasty of joint 296 15% Primary resurfacing arthroplasty of joint 122 6% Primary resurfacing hemi-arthoplasty of joint 428 22% Primary reverse polarity total prosthetic replacement 597 30% Primary indications Osteoarthritis 1,202 61% Cuff tear arthropathy 470 24% Other inflammatory arthropathy 112 6% Avascular necrosis 81 4% Acute trauma 92 5% Trauma sequelae 111 6% Other 54 3%

118 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.34 Patient characteristics for primary shoulder replacement procedures in 2012 (nine months).

Primary procedures No. % Total shoulder primaries 1,968 Patient physical status P1 - fit and healthy 173 9% P2 - mild disease not incapacitating 1,214 62% P3 - incapacitating systemic disease 558 28% P4 and P5 23 1% Previous surgery (not arthroplasty) None 1,623 82% Cuff repair 85 4% Excision ACJ 31 2% Stabilisation 28 1% Subacromial decompression 132 7% Other 134 7% Handedness Ambidextrous 69 4% Left 254 13% Right 1,645 84% Total shoulder primaries with patient data 1,894 96% Female age 1,359 72% © National Joint Registry 2013 Average 73.2 SD 9.3 Interquartile range 68.1 - 79.8 Male age 535 28% Average 68.8 SD 11.3 Interquartile range 63.1 - 77.1 Age groups 1,894 <45 years 38 2% 45 - 54 years 93 5% 55 - 64 years 256 14% 65 - 74 years 688 36% 75 - 84 years 698 37% >85 years 121 6%

2.6.1.2 Surgical techniques graft and 4% required glenoid bone graft. Long head biceps tenotomy was required in 45% of procedures Table 2.35 details the surgical technique used during and 43% had normal rotator cuff condition. 2% of shoulder primary procedures. 75% of shoulder procedures resulted in an untoward intra-operative procedures were performed using a delto-pectoral event with vascular injury occurring in 44% of these. approach and less than 1% used image-guided surgery. 12% of replacements required humeral bone

www.njrcentre.org.uk 119 Table 2.35 Characteristics of surgical practice for primary shoulder replacement procedures in 2012 (nine months).

Primary procedures No. % Total 1,968 Approach Deltoid detachment 11 1% Deltoid split 112 6% Delto-pectoral 1,474 75% Posterior 9 <1% Superior (MacKenzie) 362 18% Image-guided surgery Yes 2 <1% No 1,966 100% Humeral bone graft used Yes 243 12% No 1,725 88% Glenoid bone graft used Yes 79 4% No 1,889 96% Soft tissues Long head biceps tenotomy 882 45% Rotator cuff condition © National Joint Registry 2013 Normal 848 43% Attenuated 470 24% Absent/torn 590 30% Repaired 60 3% Untoward intra-operative events Procedures with none specified 1,920 Procedures with events specified 48 2% No. of events specified 48 Shaft penetration 1 2% Fracture humerus 15 31% Fracture glenoid 10 21% Nerve injury 0 0% Vascular injury 21 44% Other 1 2%

2.6.1.3 Thromboprophylaxis regimes. 4% used neither mechanical nor chemical regimes. LMWH was the most popular chemical Table 2.36 shows that 93% of shoulder replacement thromboprophylaxis regime used in 57% of procedures procedures used mechanical methods and 65% used and TED stockings, the most popular mechanical chemical methods; 61% of primary procedures used thromboprophylaxis, were used in 60% of procedures. both chemical and mechanical thromboprophylaxis

120 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 2.36 Thromboprophylaxis regime for primary shoulder replacement patients, prescribed at time of operation.

Total no. of primary procedures No. % Total 1,968 Aspirin 52 3% Low molecular weight Heparin 1,118 57% Pentasaccharide 7 <1% Warfarin 22 1% Direct thrombin inhibitor 35 2% Other chemical 71 4% No chemical 695 35% Foot pump 376 19% Intermittent calf compression 802 41%

TED stockings 1,180 60% © National Joint Registry 2013 Other mechanical 18 <1% No mechanical 142 7% Both mechanical and chemical 1,202 61% Neither mechanical nor chemical 71 4%

2.6.2 Shoulder components used in 2.6.3 Shoulder revision procedures primary procedures 2012 (nine months)

The Delta Xtend reverse shoulder from DePuy had Of the 257 revision procedures, 86% were single- 22% of the total shoulder replacement market for stage revisions (Table 2.37). Conversion from a primary procedures. The Copeland shoulder system hemi to a total was the cause for 30% of revision from Biomet has 27% of market share for primary procedures performed. 68% of the patients were resurfacing. More information on the other brands female and their average age was 69.9. The average used can be found in the document ‘Prostheses used age for male patients undergoing revision was 68.5. in hip, knee, ankle, elbow and shoulder replacement procedures 2012’ which can be downloaded from the NJR website.

www.njrcentre.org.uk 121 Table 2.37 Details for shoulder revision procedures in 2012 (nine months).

Procedure type Patient procedure No. % Total 257 Single-stage revision Revision hemi-arthroplasty of joint 27 11% Single-stage revision Revision resurfacing arthoplasty of joint 13 5% Revision resurfacing hemi-arthoplasty Single-stage revision 9 4% of joint Revision reverse polarity total prosthetic Single-stage revision 109 42% replacement Single-stage revision Revision total prosthetic replacement 63 25% Stage one of two-stage revision 14 5% Stage two of two-stage revision Revision hemi-arthroplasty of joint 4 2% Stage two of two-stage revision Revision resurfacing arthoplasty of joint 1 <1% Revision reverse polarity total prosthetic Stage two of two-stage revision 11 4% replacement Stage two of two-stage revision Revision total prosthetic replacement 6 2% Indications for surgery Aseptic loosening 36 14% Conversion hemi to total 77 30% Conversion total to hemi 1 <1% Cuff insufficiency 65 25% Infection 25 10%

© National Joint Registry 2013 Instability 34 13% Periprosthetic fracture 11 4% Other 60 23% Side Left 128 50% Right 129 50% Total shoulder revision procedures with patient data 241 94% Female age 164 68% Average 69.9 SD 9.1 Interquartile range 64.9 - 75.5 Male age 77 32% Average 68.5 SD 10.3 Interquartile range 62.8 - 75.4

122 www.njrcentre.org.uk Part 3 Outcomes after joint replacement 2003 to 2012

3.1 Summary of data sources and linkage The outcome analyses have been based on all Amongst the patients with person-level identifiers, 4.7% patients with at least one primary joint replacement had solely revision operations recorded within the time carried out between 1 April 2003 and 31 December frame, i.e. they had no primary operation record in the 2012, inclusive, and whose record had been NJR. This was either because the primary had taken submitted to the NJR by the end of February 2013. place at an earlier point in time (before the NJR data collection period began in 2003) or it had not been Documentation of implant survivorship and mortality included for other reasons; these cases were excluded. require a person-level identifier to relate operations This left 945,196 patients with at least one record of carried out on the same individual. Starting with a a primary joint replacement within the NJR, i.e. hips, total of 1,419,738 NJR source file records, around knees, ankles, elbows or shoulders. 12.9% were lost because no suitable person-level identifier was found. In around half of these (48.7%), At the joint level, some further revisions were excluded if the patient had declined to give consent for details to they could not be matched to primary joint replacements. be held, the remainder being attributable to tracing For example, if a primary operation was recorded only for and linkage difficulties. A person-level identifier was one side and there was a documented revision for the available for 95.9% of operations from 2008 to other side, the latter was excluded. 2012 but, in earlier years, the proportion had been much lower, for example, it was 58.9% in 2004. The The resulting data sets are shown in Table 3.1. Shoulder implication of this is that the subset of patients with and elbow replacements are not shown here but will be longer follow-up may be less representative than included in a later Report. Separate sets of analyses have those patients with shorter follow-up. been carried out for hips, knees and ankles, although the numbers for the latter are still quite small.

Table 3.1 Summary description of datasets used for survivorship analysis.

NJR data Summary of data All NJR procedure-level data restructured to person-level Time period 1 April 2003 – 31 December 2012 - Excludes data where patient-level identifier is not present Data exclusions - Excludes patients where no primary operation is recorded in the NJR - Excludes any revisions after the first revision Number of primary operations 539,372 hips; 589,028 knees; 1,417 ankles NJR identified primary-linked first revisions: Number of revisions linked to a primary operation © National Joint Registry 2013 11,780 hips; 11,666 knees; 9 ankles

Table 3.2 opposite shows the composition of the three operation to any subsequent one. Table 3.2 shows the data sets. Of the 478,730 patients with primary hip number of subsequent revision procedures and these operations 12.6% had operations documented for both numbers are small. hips; likewise 19.5% of the 499,015 patients with knee operations had operations on both sides. The unit of observation for all sets of survivorship analysis has been taken as the individual primary joint Implant survivorship is later described with respect to replacement. A patient with left and right replacements the lifetime of the primary joint only, i.e. we have looked of a particular type, therefore, will have two entries, only at time to first revision, not the time from a revision and an assumption is made that the survivorship of a

124 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

replacement on one side is independent of the other. In Figure 3.1 shows the overlap amongst patients in the practice it would be difficult to validate this, particularly dataset. For example, 443,190 patients had only hip given that some patients had had prior replacements operations in the data base, 35,422 had both hip and not recorded in the NJR. Risk factors such as age knee replacements and a further 24 had hip, knee and are recorded at the time of primary operation and will ankles. We stress this is only the composition within the therefore be different for the two procedures unless the NJR; ankle operations have only recently been included two operations are performed at the same time. and so the figures are not generalisable.

A further complication is that patients may have more than one type of implant.

Table 3.2 Composition of person-level datasets for survivorship analysis.

Hips Knees Ankles

Number Number Number Number of patients 478,730 499,015 1,388 Number (%) of patients with only one primary joint operation 418,088 (87.3%) 409,002 (82.0%) 1,359 (97.9%) Number (%) of patients with different operation dates for left and 57,550 (12.0%) 82,440 (16.5%) 27 (1.9%) right sides of the same primary joint Number (%) of patients with bilateral operations (both sides of the 3,092 (0.6%) 7,573 (1.5%) 2 (0.1%) same primary joint replaced at the same time) Total number of primary joints 539,372 589,028 1,417 Number with at least one revision operation linked to the primary 11,780 11,666 9 © National Joint Registry 2013 Number with more than one revision procedure 1,524 1,821 0

Figure 3.1 Patients with hip, knee and ankle primary operations within the survivorship data sets.

KNEES 463,368

35,422 201 24 HIPS ANKLES © National Joint Registry 2013 443,190 94 1,069

www.njrcentre.org.uk 125 Part 3 3.2 Outcomes after primary hip replacement National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

This section looks at revision and mortality for all Details of the patient cohort are given in Tables 3.1 primary hip operations performed between 1 April and 3.2 of the preceding section; a total of 539,372 2003 and 31 December 2012. Patients operated on at hips were included. the beginning of the registry therefore had a potential for nearly 10 years follow up.

Methodological note very small, the cumulative probability and cumulative hazard are numerically very similar, but with more Survival analyses have been used throughout this extended follow up, they have started to diverge. section, first looking at the need for revision and then looking at mortality. For revision, only the first revision Cumulative hazard estimates have continued to has been considered here. The majority of implants be used for graphs that compare the survival did not require revision and survival analysis made experiences of different sub-groups. Although use of the information that was available on them, either Kaplan-Meier or cumulative hazards could be i.e. that they had not been revised up to the end of used for visual comparison, the latter have some the follow up period (the end of 2012) or prior to their advantage - again see our Annexe. death; these observations were regarded as being ‘censored’ at those times. For mortality, the event The cumulative hazard plots now also include tables was death, censoring only those cases that were of the numbers at risk at each anniversary. These still alive at the end of 2012 (and not for any revision are particularly useful where a particular group has procedure). appeared to ‘plateau’ it may simply be because the number of cases fell so low that the occurrence of The survival tables below show ‘Kaplan-Meier’ further revisions/deaths became unlikely. estimates of the cumulative chance (probability) of revision, or death, at different times from the All the Kaplan-Meier estimates shown have been primary operation. This is instead of the Nelson- multiplied by 100, therefore estimate cumulative Aalen estimates of ‘cumulative hazard’ used in earlier percentage probability. annual reports, a change that brings us more into line with other registries. Please see the Annexe In the case of revisions, no attempt has been made to Survival Analysis at the end of the chapter for a to adjust for the competing risk of death. fuller explanation of the change. Where rates are

Terminology note to patients with a stemmed prosthesis and metal bearing surfaces (a monobloc metal acetabular The six main categories of bearing surfaces cup or a metal acetabular cup with a metal liner). for hip replacements are ceramic-on-ceramic Although they have metal-on-metal bearing (CoC), ceramic-on-metal (CoM), ceramic-on- surfaces, resurfacing procedures, which have a polyethylene (CoP), metal-on-metal (MoM), metal- surface replacement femoral prosthesis combined on-polyethylene (MoP) and resurfacing procedures. with a metal acetabular cup, are treated as a The metal-on-metal group in this section refers separate category.

www.njrcentre.org.uk 127 3.2.1 Overview of primary hip surgery The most commonly used type overall remained cemented metal-on-polyethylene (33%). Table 3.3 below shows the breakdown of cases by method of fixation and, within each fixation sub-group, by bearing surface.

Table 3.3 Numbers (%) of primary hip replacements by fixation and, within each fixation sub-group, by bearing surface.

Bearing surface within Fixation Number (%) Number (%) fixation group All cases 539,372 (100.0%) 539,372 (100.0%) MoP 178,077 (33.0%) MoM 1,229 (0.2%) All cemented 201,580 (37.4%) CoP 17,202 (3.2%) Others/unsure 5,072 (0.9%) MoP 74,762 (13.9%) MoM 28,367 (5.3%) CoP 27,964 (5.2%) All uncemented 205,317 (38.1%) CoC 67,922 (12.6%) CoM 2,017 (0.4%) Others/unsure 4,285 (0.8%) MoP 55,585 (10.3%) MoM 2,297 (0.4%)

© National Joint Registry 2013 All hybrid 84,671 (15.7%) CoP 9,720 (1.8%) CoC 15,239 (2.8%) Others/unsure 1,830 (0.3%) MoP 8,397 (1.6%) All reverse hybrid 12,296 (2.3%) CoP 3,816 (0.7%) Others/unsure 83 (<0.1%) All resurfacing 35,470 (6.6%) (MoM) 35,470 (6.6%) Unsure 38 (<0.1%) Unsure 38 (<0.1%)

128 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.4 shows the distribution of fixation/bearing time whereas the percentages of both the uncemented surface groups for each year of primary operation. and hybrid metal-on-polyethylene have been increasing.

The percentage of cemented metal-on-polyethylene, The proportions of metal-on-metal and resurfacing the most popular type of implant, has been falling with implants have been falling since 2008.

Table 3.4 Percentage of primary hip replacements performed each year by type of hip fixation and constraint.

Percentage of hip replacements by year of primary operation 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 n= n= n= n= n= n= n= n= n= n= Fixation/bearing 14,410 27,974 40,125 47,455 60,375 66,586 67,384 69,649 72,248 73,166 All cemented 60.5 54.2 48.6 42.8 39.8 34.3 31.9 31.4 32.4 33.2 Cemented by bearing surface: MoP 55.3 49.0 43.9 38.4 35.7 30.4 28.2 27.2 27.6 28.6 MoM 0.2 0.4 0.4 0.4 0.4 0.4 0.1 0.1 0.1 0.1 CoP 3.0 3.5 3.1 3.0 2.6 2.7 2.9 3.3 3.5 4.1 Others/unsure 2.0 1.3 1.2 1.0 1.1 0.9 0.6 0.9 1.2 0.4 All uncemented 16.8 21.4 25.6 30.0 33.3 39.3 43.1 45.7 44.9 44.8 Uncemented by bearing surface: MoP 6.2 9.0 9.8 10.3 10.8 13.1 15.1 16.9 17.1 17.8 MoM 1.3 2.2 5.4 8.3 10.3 10.9 8.0 3.2 0.5 0.1 CoP 5.0 5.1 5.1 4.3 4.0 3.9 4.7 5.6 6.1 7.3 CoC 3.5 4.2 4.4 6.2 7.3 10.1 13.6 18.1 20.1 19.2 CoM 0.0 <0.05 <0.05 <0.05 0.1 0.4 0.9 1.0 0.4 0.1 Others/unsure 0.9 0.8 0.9 0.9 0.9 1.0 0.9 0.9 0.7 0.3 All hybrid 12.3 13.3 14.1 15.2 15.0 15.1 15.8 16.2 17.2 17.6 Hybrid by bearing surface:

MoP 8.2 9.1 9.2 9.7 9.8 9.7 10.4 10.8 11.5 11.5 © National Joint Registry 2013 MoM 0.7 0.5 0.5 0.7 0.8 0.9 0.4 0.2 0.1 <0.05 CoP 1.6 1.5 1.2 1.3 1.1 1.4 1.8 2.0 2.3 3.1 CoC 1.2 1.9 2.7 3.2 2.9 2.7 2.9 3.0 3.1 2.8 Others/unsure 0.7 0.4 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.1 All reverse hybrid 0.6 0.9 1.1 1.2 1.8 2.5 2.7 2.8 3.1 3.1 Reverse hybrid by bearing surface: MoP 0.4 0.6 0.7 0.8 1.1 1.7 1.9 1.9 2.2 2.0 CoP 0.3 0.2 0.3 0.3 0.6 0.7 0.8 0.9 0.9 1.1 Others/unsure <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 All resurfacing 9.8 10.2 10.7 10.8 10.2 8.8 6.5 3.8 2.5 1.3 (MoM) All types 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

www.njrcentre.org.uk 129 The primary operations were carried out by a total procedures of that type, together with the median and of 2,772 consultant surgeons across 456 units. The IQR of the number of procedures they carried out. median number of primary procedures per consultant was 64, inter-quartile range (IQR) 10-244 and the Surgeons performing cemented THR carried out a median number of procedures per unit was 910.5, median of 43 operations over the period they were IQR 419.5-1677. observed with IQR 13 to 126 procedures. This means that 25% of surgeons performed fewer than 13 Table 3.5 below shows the distributions of consultant cemented hip replacements and 25% more than 126. surgeon and unit caseloads for each type of fixation; Additionally, 10% of surgeons carrying out cemented consultants and units with fewer than 10 cases in hip replacements had performed between 782 and the database have been excluded (683 of 2,772 1,390 procedures (not shown in table). Similarly, consultant surgeons and 11 of 456 units). the surgeons with the highest 10% of caseloads performing uncemented THR carried out between 917 The table shows, for each fixation type, the and 2,614 procedures over the period of observation. percentage of surgeons or units that carried out

Table 3.5 Distribution of consultant surgeon and unit caseload for each fixation type.

Number of procedures carried out by Number of procedures carried consultant surgeons (n=2,089): out by units (n=445): % performing % performing this fixation this fixation Fixation type Median (IQR) type Median (IQR) Cemented 95.2 43 (13-126) 99.6 252 (83-645) Uncemented 84.4 38 (8-147) 98.0 296 (120.5-622.5) Hybrids 71.3 11 (3-50) 94.6 69 (14-208) Reverse hybrids 36.6 2 (1-8.5) 74.4 6 (2-25)

The median age at operation was 69 (IQR 61-76; sub-group. On the whole, resurfacing and ceramic- range 7-105) years.10 on-ceramic bearings tended to be used in younger patients although the age ranges were wide. Table 3.6 gives the breakdown of ages by fixation and by the main bearing surfaces within each fixation

10 15 ages were 0 and assumed missing.

130 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.6 Age (in years) at primary hip replacement, by fixation and main bearing surface.

By bearing surface within Median (IQR*; Fixation fixation group** range) of ages Minimum age Maximum age All cases 69 (61-76) 7 105 All cemented 74 (68-79) 7*** 103 Cemented and MoP 74 (69-79) 15 103 MoM 65 (58-73) 25 98 CoP 65 (59-71) 16 101 All uncemented 65 (58-72) 9 105 Uncemented and MoP 71 (65-77) 14 101 MoM 64 (57-70) 13 105 CoP 65 (59-71) 13 100 CoC 61 (53-67) 9 100 CoM 63 (56-69) 20 92 All hybrid 70 (62-76) 12 100 Hybrid and MoP 73 (67-78) 12 100 © National Joint Registry 2013 MoM 63 (56-71) 15 93 CoP 65 (59-72) 16 96 CoC 60 (54-66) 13 92 All reverse hybrid 71 (64-77) 13 100 Reverse hybrid and MoP 73 (68-78) 13 100 CoP 64 (59-70) 16 94 All resurfacing 55 (49-60) 12 95 Resurfacing and MoM 55 (49-60) 12 95

*IQR=inter-quartile range; **Excludes all sub-groups with ‘other/unsure’ bearing surfaces, ***The bearing surface for the 7 year old was other or unsure so not shown in sub groups presented. 3.2.2 Revisions after primary hip surgery was highest for metal-on-metal. This is further exemplified by Figure 3.2, where the cumulative hazard for the three bearing Table 3.7 shows Kaplan-Meier estimates of the cumulative surface groups are plotted together. The shaded bands in percentage probability of first revision, for any cause, for all these figures indicate point-wise 95% CIs for the estimates. cases combined and then subdivided by fixation and bearing surface within each fixation group. Estimates are shown, A similar picture exists for uncemented hips, see Figure together with 95% Confidence Intervals (95% CI), at 30 and 3.3, namely that the metal-on-metal revision rates were the 90 days after the primary operation and at each anniversary highest, higher even than those for resurfacing. up to the ninth year. These do not adjust for other factors Amongst the hybrid hips, again the revision rates for metal- such as age and gender. on-metal were highest – see Figure 3.4. At time points where the estimates are shown in italics, fewer For the cumulative hazard curves for metal-on-metal in than 100 cases remained at risk; if the numbers at risk are Figures 3.2 to 3.4, the tangents appeared to increase over small, revisions, as they occur, may appear to have greater the 9 years of follow up, suggesting that the hazard rate (the impact on the failure rate estimates, i.e. the step upwards rate of revision amongst the unrevised cases) increased with may appear steeper. time from primary operation. Amongst the cemented hips, the revision rate at nine years

www.njrcentre.org.uk 131 © National Joint Registry 2013 Registry Joint National © - 4.08 1.84 3.67 3.03 2.99 3.02 3.42 2.62 3.31 4.30 5.75 2.71 4.07 6.71 5.06 2.34 3.24 7.94 2.72 14.86 21.43 17.66 12.31 9 years (3.59-4.64) (1.51-2.24) (2.65-3.45) (2.66-3.37) (2.47-2.78) (3.82-4.84) (1.81-3.02) (2.30-4.57) (1.88-3.95) (2.09-6.40) (2.32-3.91) (2.52-4.33) (4.36-7.57) (2.54-6.50) (3.10-3.76) (2.57-2.87) (6.40-7.05) (4.92-5.21) (2.65-22.50) (15.93-19.56) (12.05-18.25) (15.41-29.36) (11.62-13.04) 1.97 3.85 1.84 3.67 2.89 2.65 3.02 2.99 2.33 3.31 3.80 4.96 2.41 3.39 6.23 5.51 4.53 2.12 2.83 7.94 13.92 14.38 15.55 10.54 8 years (3.46-4.27) (1.51-2.24) (2.56-3.25) (2.40-2.93) (2.32-3.91) (2.22-2.46) (3.48-4.14) (3.93-6.26) (2.25-5.10) (1.73-2.60) (2.15-3.72) (1.50-2.60) (2.09-6.40) (3.59-8.42) (2.77-3.22) (2.52-4.33) (2.29-2.52) (5.99-6.48) (4.42-4.63) (2.65-22.50) (11.61-16.63) (14.57-16.59) (11.40-18.06) (10.05-11.05) 1.75 3.41 1.67 2.80 2.66 2.35 3.02 2.66 2.02 2.99 3.21 4.28 2.09 2.93 5.59 9.02 5.51 3.97 1.87 2.55 7.94 12.67 12.43 13.04 7 years (1.35-2.27) (3.12-3.72) (1.39-2.01) (1.83-4.26) (2.39-2.97) (2.16-2.57) (2.32-3.91) (1.92-2.11) (3.00-3.43) (3.51-5.21) (2.04-4.21) (1.58-2.26) (1.96-3.32) (3.59-8.42) (2.49-2.85) (2.41-3.72) (2.00-2.18) (5.30-5.68) (8.63-9.43) (3.88-4.05) (9.93-15.51) (2.65-22.50) (10.64-15.06) (12.39-13.72) 1.52 7.94 2.96 9.34 1.41 2.11 2.40 2.08 9.20 2.55 2.27 1.75 2.46 2.92 4.08 1.79 2.49 4.77 7.32 5.51 3.38 1.67 2.18 10.65 6 years (1.19-1.94) (2.74-3.19) (1.18-1.68) (1.52-2.92) (2.16-2.66) (1.91-2.25) (2.02-3.20) (1.67-1.83) (2.74-3.11) (3.35-4.96) (1.75-3.53) (1.41-1.97) (1.68-2.84) (3.59-8.42) (2.13-2.42) (2.04-2.96) (1.72-1.87) (4.62-4.92) (7.00-7.65) (3.31-3.45) (7.87-11.06) (7.32-11.52) (2.65-22.50) (10.16-11.16) 1.37 4.10 2.58 6.29 1.21 1.94 2.16 1.78 6.32 2.20 7.65 1.88 1.50 2.13 2.50 3.31 1.52 2.06 3.76 5.82 5.51 2.75 1.44 1.60 5 years (1.08-1.73) (2.41-2.76) (5.27-7.51) (1.02-1.45) (1.41-2.65) (1.95-2.40) (1.65-1.93) (4.94-8.08) (1.77-2.72) (7.30-8.02) (1.43-1.57) (2.35-2.65) (2.72-4.02) (1.44-2.92) (1.23-1.70) (1.22-2.10) (3.59-8.42) (1.77-2.00) (1.78-2.53) (1.45-1.58) (3.65-3.88) (5.55-6.10) (2.69-2.81) (1.34-12.24) 1.14 4.10 2.20 4.68 1.02 1.79 1.86 1.46 4.39 1.71 5.24 1.54 1.24 1.76 2.19 2.63 1.24 1.85 2.88 4.39 4.15 2.15 1.27 1.20 4 years (0.90-1.44) (2.06-2.34) (3.85-5.69) (0.85-1.21) (1.33-2.42) (1.68-2.07) (1.35-1.59) (3.31-5.80) (1.40-2.10) (4.97-5.52) (1.18-1.30) (2.07-2.32) (2.14-3.21) (1.29-2.65) (3.06-5.62) (1.08-1.50) (0.90-1.60) (1.44-1.64) (1.49-2.08) (1.19-1.30) (2.79-2.98) (4.17-4.63) (2.11-2.20) (1.34-12.24) 0.93 4.10 1.87 2.98 0.88 1.54 1.55 1.23 2.87 1.42 3.38 1.23 1.01 1.48 1.93 2.18 1.01 1.58 2.19 3.14 2.94 1.66 0.99 1.05 3 years (0.73-1.18) (1.76-2.00) (2.34-3.78) (0.73-1.06) (1.13-2.10) (1.39-1.72) (1.13-1.34) (2.04-4.04) (1.16-1.74) (3.17-3.60) (0.96-1.06) (1.82-2.05) (1.76-2.69) (1.08-2.30) (2.19-3.95) (0.84-1.18) (0.78-1.42) (1.15-1.31) (1.25-1.75) (0.97-1.06) (2.12-2.26) (2.96-3.33) (1.62-1.69) (1.34-12.24) 0.69 2.61 1.50 1.60 0.64 1.01 1.17 1.02 1.65 1.13 1.99 0.98 0.74 1.11 1.51 1.70 0.74 1.37 1.56 2.18 1.47 1.20 0.86 0.75 2 years (0.53-0.90) (1.40-1.60) (1.16-2.21) (0.52-0.78) (0.71-1.43) (1.04-1.32) (0.93-1.11) (1.06-2.58) (0.91-1.40) (1.84-2.17) (0.70-0.79) (1.42-1.61) (1.35-2.15) (0.92-2.03) (1.01-2.12) (0.72-1.03) (0.54-1.04) (0.91-1.05) (0.92-1.33) (1.50-1.62) (2.03-2.34) (1.17-1.23) (0.70-0.78) (0.66-10.04) 0.47 0.51 1.23 0.96 0.66 0.35 0.63 0.81 0.69 0.67 0.81 1.02 0.65 0.45 0.76 1.06 1.25 0.44 1.12 0.99 1.27 0.65 0.75 0.54 1 year (0.35-0.64) (0.89-1.04) (0.40-1.09) (0.27-0.45) (0.42-0.96) (0.71-0.93) (0.62-0.77) (0.33-1.33) (0.63-1.03) (0.91-1.14) (0.42-0.48) (0.98-1.14) (0.96-1.64) (0.72-1.73) (0.38-1.12) (0.43-0.67) (0.34-0.75) (0.17-8.44) (0.59-0.71) (0.61-0.93) (0.41-0.47) (0.95-1.03) (1.16-1.39) (0.72-0.77) 0.22 0.24 1.23 0.53 0.44 0.14 0.21 0.47 0.38 0.24 0.35 0.48 0.36 0.23 0.31 0.62 0.87 0.22 0.55 0.55 0.66 0.20 0.40 0.32 90 days (0.14-0.34) (0.14-0.42) (0.48-0.59) (0.23-0.81) (0.09-0.20) (0.11-0.43) (0.40-0.56) (0.33-0.44) (0.08-0.76) (0.24-0.50) (0.41-0.57) (0.20-0.25) (0.56-0.68) (0.63-1.19) (0.30-1.02) (0.07-0.53) (0.24-0.42) (0.17-8.44) (0.23-0.43) (0.20-0.24) (0.52-0.58) (0.58-0.75) (0.38-0.42) (0.32-0.40) 0.11 0.18 0.00 0.34 0.35 0.07 0.11 0.30 0.24 0.16 0.18 0.31 0.22 0.13 0.16 0.42 0.77 0.12 0.22 0.36 0.25 0.15 0.24 0.22 30 days (0.06-0.21) (0.09-0.34) (0.29-0.38) (0.17-0.70) (0.04-0.12) (0.04-0.28) (0.24-0.37) (0.20-0.28) (0.04-0.65) (0.11-0.30) (0.26-0.39) (0.11-0.14) (0.37-0.47) (0.55-1.08) (0.08-0.58) (0.05-0.46) (0.16-0.31) (0.19-0.26) (0.10-0.24) (0.11-0.14) (0.34-0.39) (0.20-0.31) (0.23-0.25) Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI), at different times after the primary operation, (95% CI), at different of first revision probability Kaplan-Meier estimates of the cumulative percentage

CoM CoP Others/unsure Others/unsure** CoC MoM CoP CoP CoP Hybrids by bearing surface: MoP MoM Reserve hybrids by bearing surface: MoP MoM All hybrids Cemented by bearing surface: MoP All reverse hybrids All reverse Uncemented by bearing surface: MoP Others/unsure All cemented Others/unsure All uncemented All resurfacing (MoM) All resurfacing Fixation/bearing types All cases* CoC Table 3.7 Table small group size (n=83) *Including the 38 with unsure fixation/bearing surface; ** Wide CIs because based on a for each fixation/bearing surface sub-group. For the estimates in blue italics, fewer than 100 cases remain at risk. for each fixation/bearing surface sub-group.

132 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.2 Comparison of cumulative hazard of first revision for cemented hips with different bearing surfaces (with 95% CI).

10 Cumulative hazard x 100

5 © National Joint Registry 2013

0 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk Cemented MoP 178,077 152,9811130,206 2108,931 388,203 467,448 547,008 630,708 7 16,134 5,59498 Cemented MoM 1,229 1,151 1,072 1,002 880 612 397 216 96 15 Cemented CoP 17,202 14,022 11,325 8,961 6,984 5,172 3,652 2,319 1,164 344

www.njrcentre.org.uk 133 Figure 3.3 Comparison of cumulative hazard of first hip revision for uncemented hips with different bearing surfaces (with 95% CI).

5 Cumulative hazard x 100

0 © National Joint Registry 2013 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Uncemented MoP 74,762 60,095 47,032 34,874 24,669 16,385 10,371 6,010 2,689 670 Uncemented MoM 28,367 27,716 26,841 24,022 18,424 11,517 5,880 2,476 648 142 Uncemented CoP 27,964 22,234 17,670 13,639 10,443 7,877 5,538 3,617 1,832 602 Uncemented CoC 67,922 53,029 38,276 25,689 16,673 10,147 5,918 3,134 1,512 445 Uncemented CoM 2,017 1,942 1,621 921 310 53 10 1 1 0 Resurfacing 35,470 33,989 31,820 28,841 24,276 18,459 12,547 7,759 3,881 1,273

134 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.4 Comparison of cumulative hazard of a first hip revision for hybrid (not including reverse hybrid) hips with different bearing surfaces (with 95% CI).

5 Cumulative hazard x 100 © National Joint Registry 2013 0 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Hybrid MoP 55,585 45,908 37,004 29,075 21,909 15,596 10,122 6,095 2,957 911 Hybrid MoM 2,297 2,247 2,145 1,951 1,627 1,073 610 338 190 79 Hybrid CoP 9,720 7,357 5,673 4,295 3,092 2,189 1,559 993 548 187 Hybrid CoC 15,239 13,020 10,672 8,584 6,592 4,802 3,104 1,690 671 163

www.njrcentre.org.uk 135 3.2.3 Revisions for different causes after primary hip surgery

Methodological note time incidence rates (PTIRs); the total number of revisions for that reason has been divided by the The previous section looked at revisions for any total of the individual patient-years at risk. The reason. A number of reasons may be associated figures shown are the numbers per 1,000 years at with any revision; for example, pain and osteolysis. risk. This method is appropriate if the hazard rate This means the reasons are not mutually exclusive (the rate at which revisions occur in the unrevised and therefore cannot be regarded as representing cases) remains constant. The latter is explored ‘competing risks’. Here we have calculated further later in this section. incidence rates for each reason using patient-

Table 3.8 shows the revision rates for each case, for contender. Resurfacings also appeared to have the all cases and broken down by fixation and, within each highest incidence of adverse soft tissue reactions. The fixation group, by bearing surface. In the initial years incidence of dislocation/subluxation, however, was of the formation of the registry, adverse soft tissue lowest for resurfacings. reaction was not included in the clinical assessment forms. Therefore, there were fewer ‘patient-years at After further subdivision by bearing surface, amongst risk’ for this reason. the cemented, uncemented and hybrid fixation groups, the metal-on-metal sub-groups had the The main reasons for revision were pain and aseptic highest incidences of pain and aseptic loosening loosening; resurfacing seemed to have the highest and also appeared to have the highest incidences of incidence of each, with uncemented hips a close adverse soft tissue reactions.

136 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.8 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first hip revision. Rates shown are for all revised cases and by fixation and bearing surface.

Number of revisions per 1,000 patient-years for: Patient- Fixation/ years at risk Dislocation/ Aseptic Periprosthetic Implant bearing types (x1,000) Pain subluxation Infection loosening Lysis fracture fracture 1.37 1.03 0.83 1.44 0.29 0.64 0.17 All cases* 2,026.4 (1.32-1.42) (0.99-1.08) (0.80-0.88) (1.39-1.49) (0.27-0.31) (0.61-0.68) (0.15-0.19) 0.48 0.89 0.79 0.93 0.18 0.30 0.06 All Cemented 824.5 (0.44-0.53) (0.83-0.96) (0.73-0.85) (0.87-1.00) (0.15-0.21) (0.27-0.34) (0.05-0.08) Cemented by bearing surface: 0.45 0.92 0.78 0.93 0.17 0.31 0.05 MoP 735.1 (0.40-0.50) (0.85-0.99) (0.72-0.84) (0.86-1.00) (0.15-0.21) (0.27-0.35) (0.04-0.07) 4.13 0.83 1.16 4.13 1.65 1.16 0.83 MoM 6.1 (2.79-6.11) (0.34-1.99) (0.55-2.43) (2.79-6.11) (0.89-3.07) (0.55-2.43) (0.34-2.00) 0.43 0.69 0.87 0.67 0.11 0.10 0.05 CoP 62.4 (0.30-0.63) (0.51-0.93) (0.66-1.13) (0.50-0.91) (0.05-0.24) (0.04-0.21) (0.02-0.15) 0.62 0.62 0.96 0.91 0.14 0.34 0.24 Others/unsure 20.9 (0.36-1.07) (0.36-1.07) (0.62-1.48) (0.58-1.43) (0.05-0.45) (0.16-0.70) (0.10-0.58) All 1.85 1.29 0.97 1.94 0.31 0.82 0.28 680.8 uncemented (1.75-1.95) (1.21-1.38) (0.89-1.04) (1.84-2.05) (0.27-0.35) (0.75-0.89) (0.24-0.32) Uncemented by bearing surface: 0.99 1.62 0.89 1.53 0.20 1.01 0.12 MoP 239.6 (0.87-1.12) (1.47-1.79) (0.78-1.02) (1.38-1.70) (0.15-0.27) (0.89-1.15) (0.08-0.17) 5.32 1.10 1.60 3.44 0.82 0.64 0.16 MoM 132.1 (4.94-5.73) (0.93-1.29) (1.40-1.83) (3.14-3.77) (0.68-0.99) (0.51-0.79) (0.10-0.24) 0.76 1.20 0.64 1.44 0.16 0.51 0.14 CoP 97.2 (0.61-0.96) (1.00-1.44) (0.50-0.82) (1.22-1.70) (0.10-0.27) (0.39-0.68) (0.09-0.24) 1.10 1.05 0.80 1.64 0.13 0.83 0.63 CoC 188.2 (0.96-1.26) (0.92-1.21) (0.68-0.94) (1.46-1.83) (0.09-0.19) (0.71-0.97) (0.53-0.76) 2.37 1.19 1.52 2.71 0.68 0.34 0.17 CoM 5.9 (1.40-4.00) (0.57-2.49) (0.79-2.93) (1.66-4.42) (0.25-1.80) (0.08-1.35) (0.02-1.20) 1.40 1.34 0.56 2.13 0.50 1.18 0.39 Others/unsure 17.9 (0.95-2.07) (0.90-2.00) (0.30-1.04) (1.55-2.93) (0.26-0.97) (0.77-1.80) (0.19-0.82) 0.73 1.19 0.74 0.83 0.25 0.65 0.12 All hybrid 304.0 (0.64-0.84) (1.07-1.32) (0.65-0.84) (0.73-0.94) (0.20-0.31) (0.56-0.75) (0.08-0.16) © National Joint Registry 2013 Hybrids by bearing surface: 0.52 1.35 0.77 0.80 0.23 0.68 0.09 MoP 196.9 (0.43-0.63) (1.20-1.52) (0.66-0.91) (0.69-0.94) (0.17-0.31) (0.57-0.80) (0.06-0.15) 4.54 1.84 0.96 2.97 1.49 1.66 0.17 MoM 11.4 (3.46-5.96) (1.20-2.82) (0.53-1.74) (2.12-4.16) (0.92-2.39) (1.06-2.60) (0.04-0.70) 0.62 0.98 0.65 0.52 0.13 0.46 0.03 CoP 30.6 (0.40-0.97) (0.69-1.40) (0.42-1.01) (0.32-0.85) (0.05-0.35) (0.27-0.77) (0.005-0.23) 0.69 0.65 0.55 0.65 0.11 0.47 0.25 CoC 56.9 (0.50-0.94) (0.47-0.90) (0.38-0.78) (0.47-0.90) (0.05-0.23) (0.33-0.69) (0.15-0.42) 1.34 0.85 1.34 0.85 0.49 0.49 Others/unsure 8.2 0 (0.74-2.41) (0.41-1.78) (0.74-2.41) (0.41-1.78) (0.18-1.29) (0.18-1.29) All reverse 0.86 1.19 1.08 1.38 0.11 0.61 0.08 36.2 hybrid (0.60-1.22) (0.88-1.60) (0.79-1.48) (1.05-1.82) (0.04-0.29) (0.40-0.92) (0.03-0.26) Reverse hybrids by bearing surface: 0.57 1.30 1.14 1.22 0.08 0.69 0.08 MoP 24.6 (0.34-0.96) (0.92-1.84) (0.79-1.65) (0.85-1.75) (0.02-0.33) (0.43-1.11) (0.02-0.33) 1.51 0.98 0.98 1.60 0.18 0.44 0.09 CoP 11.3 (0.94-2.43) (0.54-1.76) (0.54-1.76) (1.01-2.54) (0.04-0.71) (0.18-1.07) (0.01-0.63) Others/ 6.26 0.3 0 0 0 0 0 0 unsure** (1.57-25.03) All 4.78 0.43 0.65 2.92 0.82 1.57 0.35 Resurfacing 180.6 (4.47-5.11) (0.34-0.53) (0.55-0.78) (2.68-3.18) (0.70-0.97) (1.39-1.76) (0.28-0.45) (MoM) *Including the 38 with unknown fixation/bearing. **Based on a small group size (n=83), therefore estimates are unreliable. Continued >

www.njrcentre.org.uk 137 Table 3.8 (continued)

Number of revisions per 1,000 patient-years for: Adverse Fixation/bearing Patient-years Head/socket Other Patient-years soft tissue types at risk (x1,000) Implant failure mismatch Malalignment indication at risk (x1,000) reaction*** 0.28 0.06 0.50 0.66 0.68 All cases* 2,026.40 797.5 (0.26-0.31) (0.05-0.07) (0.47-0.53) (0.63-0.70) (0.63-0.74) 0.11 0.03 0.27 0.19 0.04 All Cemented 824.5 260.4 (0.09-0.14) (0.02-0.04) (0.24-0.31) (0.16-0.22) (0.02-0.07) Cemented by bearing surface: 0.10 0.03 0.28 0.17 0.02 MoP 735.1 227.6 (0.08-0.13) (0.02-0.04) (0.24-0.32) (0.14-0.20) (0.01-0.05) 1.16 0.17 0.33 2.97 3.10 MoM 6.1 1.6 (0.55-2.43) (0.02-1.20) (0.08-1.32) (1.87-4.72) (1.30-7.40) 0.14 0.02 0.22 0.14 CoP 62.4 24.4 0 (0.08-0.28) (0.002-0.11) (0.13-0.38) (0.08-0.28) 0.10 0.24 0.24 Others/unsure 20.9 0 6.8 0 (0.02-0.38) (0.10-0.58) (0.10-0.58) 0.50 0.10 0.72 0.96 1.02 All uncemented 680.8 340.5 (0.45-0.55) (0.08-0.13) (0.66-0.78) (0.89-1.03) (0.91-1.13) Uncemented by bearing surface: 0.43 0.09 0.71 0.46 0.18 MoP 239.6 120.0 (0.35-0.52) (0.06-0.13) (0.61-0.83) (0.38-0.55) (0.12-0.28) 0.95 0.09 0.95 2.57 5.50 MoM 132.1 51.7 (0.79-1.13) (0.05-0.16) (0.80-1.14) (2.31-2.86) (4.90-6.20) 0.35 0.09 0.54 0.55 0.10 CoP 97.2 38.8 (0.25-0.49) (0.05-0.18) (0.41-0.70) (0.42-0.71) (0.04-0.27) 0.35 0.12 0.63 0.72 0.19 CoC 188.2 117.1 (0.28-0.45) (0.08-0.18) (0.53-0.76) (0.61-0.85) (0.12-0.29) 0.68 0.17 1.19 0.85 1.60 CoM 5.9 5.6 (0.25-1.80) (0.02-1.20) (0.57-2.49) (0.35-2.03) (0.84-3.10) 0.34 0.17 0.73 0.56 0.67 Others/unsure 17.9 7.4 (0.15-0.75) (0.05-0.52) (0.42-1.25) (0.30-1.04) (0.28-1.60) 0.24 0.03 0.40 0.34 0.14 All hybrid 304.0 128.6 (0.19-0.31) (0.01-0.05) (0.33-0.48) (0.28-0.41) (0.09-0.22) © National Joint Registry 2013 Hybrids by bearing surface: 0.24 0.03 0.42 0.23 0.04 MoP 196.9 84.2 (0.18-0.32) (0.01-0.06) (0.34-0.52) (0.17-0.31) (0.01-0.11) 0.35 0.17 0.96 2.62 3.70 MoM 11.4 3.5 (0.13-0.93) (0.04-0.70) (0.53-1.74) (1.83-3.75) (2.10-6.30) 0.20 0.16 0.29 CoP 30.6 0 14.4 0 (0.09-0.44) (0.07-0.39) (0.15-0.57) 0.23 0.02 0.37 0.30 0.09 CoC 56.9 23.5 (0.13-0.39) (0.00(2)-0.12) (0.24-0.57) (0.19-0.48) (0.02-0.34) 0.36 0.24 0.36 Others/unsure 8.2 0 3.0 0 (0.12-1.13) (0.06-0.97) (0.12-1.13) 0.14 0.06 0.41 0.39 All reverse hybrid 36.2 20.4 0 (0.06-0.33) (0.01-0.22) (0.25-0.69) (0.23-0.65) Reverse hybrids by bearing surface: 0.12 0.08 0.45 0.37 MoP 24.6 13.8 0 (0.04-0.38) (0.02-0.33) (0.25-0.81) (0.19-0.70) 0.18 0.36 0.36 CoP 11.3 0 6.4 0 (0.04-0.71) (0.13-0.95) (0.13-0.95) 3.13 Others/unsure** 0.3 0 0 0 0.1 0 (0.44-22.00) All Resurfacing 0.36 0.12 0.93 2.27 3.55 180.6 47.6 (MoM) (0.28-0.46) (0.08-0.18) (0.80-1.08) (2.06-2.50) (3.05-4.13)

*Including the 38 with unknown fixation/bearing. **Based on a small group size (n=83), therefore estimates are unreliable. ***This reason was not asked in the early versions of the clinical assessment forms MDSv1 and MDSv2 for joint replacement/revision surgery and hence, for this reason, there are fewer patient-years at risk.

138 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.9 illustrates how the risk of revision (for any Note the maximum follow-up for any implant was reason) changed with time from primary operation. 9.75 years.

The number of revisions per 1,000 patient-years The rate was higher initially, with a small decrease was calculated for different periods after the primary between one and three years, after which the rate operation, within the first year from primary operation, began to rise. The rates were less reliable after seven and between 1-3, 3-5, 5-7 and 7+ years after surgery. years because there were fewer patient-years at risk.

Table 3.9 Revision rates, expressed as numbers per 1,000 patient-years, for any reason, according to time interval from primary operation.

Time from primary operation Patient-years at risk (x1,000) Revision rate per 1,000 patient-years (95% CI) <1 year 497.6 7.71 (7.47-7.96) 1-3 years 756.8 4.60 (4.45-4.75) 3-5 years 469.0 5.51 (5.30-5.72) 5-7 years 227.8 6.36 (6.04-6.69) >7 years* 75.1 5.75 (5.23-6.32)

*current maximum observed follow up is 9.75 years © National Joint Registry 2013

In Table 3.10 the breakdown of Table 3.9 has been subluxation/dislocation, infection and periprosthetic repeated, this time showing separate breakdowns fracture were highest early on, then fell. Incidence for each reason. There were trends of increasing of adverse soft tissue reaction seemed to increase risk of pain and aseptic loosening with time. Risk of with time.

Table 3.10 Revision rates (95% CI), expressed as numbers per 1,000 patient-years, for each reason, by time interval from primary operation.

Time Patient- Number of revisions per 1,000 patient-years for: from years primary at risk Dislocation/ Aseptic Periprosthetic Implant operation (x1,000) Pain subluxation Infection loosening Lysis fracture fracture 0.85 2.32 1.28 1.28 0.10 1.49 0.27 <1 year 497.6 (0.77-0.94) (2.19-2.45) (1.18-1.38) (1.18-1.38) (0.07-0.13) (1.38-1.60) (0.22-0.31) 1.25 0.66 0.87 1.38 0.21 0.29 0.14 1-3 years 756.8 (1.17-1.33) (0.61-0.72) (0.80-0.93) (1.30-1.47) (0.18-0.25) (0.26-0.33) (0.12-0.17) 1.77 0.55 0.58 1.48 0.40 0.39 0.12 3-5 years 469.0 (1.66-1.90) (0.48-0.62) (0.51-0.65) (1.38-1.60) (0.34-0.46) (0.34-0.46) (0.09-0.16) 1.96 0.58 0.44 1.75 0.56 0.54 0.17 5-7 years 227.8 (1.79-2.15) (0.48-0.68) (0.36-0.53) (1.59-1.93) (0.47-0.66) (0.46-0.65) (0.12-0.23) © National Joint Registry 2013 >7 1.72 0.75 0.39 1.88 0.84 0.48 0.15 75.1 years*** (1.44-2.04) (0.57-0.97) (0.27-0.56) (1.59-2.21) (0.65-1.07) (0.35-0.66) (0.08-0.26)

Continued >

www.njrcentre.org.uk 139 Table 3.10 (continued)

Number of revisions per 1,000 patient-years for: Patient- Adverse Patient- years at risk soft tissue Time from years at risk Implant Head/socket Other (x1,000) reaction** primary operation (x1,000) failure mismatch Malalignment indication 0.26 0.12 0.91 0.81 0.10 <1 year 497.6 301.0 (0.22-0.31) (0.10-0.16) (0.83-0.99) (0.74-0.90) (0.07-0.15) 0.16 0.05 0.40 0.57 0.71 1-3 years 756.8 376.3 (0.13-0.19) (0.03-0.07) (0.36-0.45) (0.52-0.63) (0.63-0.80) 0.36 0.03 0.35 0.69 2.04 3-5 years 469.0 116.2 (0.31-0.42) (0.01-0.05) (0.30-0.41) (0.62-0.77) (1.80-2.32) 0.51 0.03 0.31 0.61 2.36 5-7 years 227.8 3.4 (0.42-0.61) (0.01-0.06) (0.25-0.39) (0.52-0.73) (1.18-4.72) © National Joint Registry 2013 0.52 0.04 0.39 0.48 1.80 >7 years*** 75.1 0.6 (0.38-0.71) (0.01-0.12) (0.27-0.56) (0.35-0.66) (0.25-1.30)

*Includes the 38 with unknown fixation/bearing. ** This reason was not asked in the early versions of the clinical assessment forms MDSv1 and MDSv2 for joint replacement/revision surgery and hence, for these reasons, there are fewer patient-years at risk. ***Current maximum observed follow up is 9.75 years.

Note on dislocations/subluxation: There were 2097 first revisions performed where dislocations/subluxation was given as a reason (out of all 539,372 operations); in 1,361 of these (64.9%) dislocation/subluxation was the only given reason. From tables 3.9-3.10 above, revisions due to dislocation/subluxation were more common in uncemented and hybrid/reverse hybrid hips than cemented hips and were more common within the first year after primary operation.

3.2.4 Revisions after primary hip The figures in italics were at time points where fewer surgery for the main stem-cup than 100 cases remained at risk. brand combinations Note it is possible that these sub-groups may differ in composition with respect to other factors that might Table 3.11 show Kaplan-Meier estimates of the influence revision, such as age and gender; no attempt cumulative percentage probability of revision (for any here has been made to adjust for such factors. reason) for the main stem-cup brands. Some further subdivisions by bearing type, however, are As in our previous reports, we have only included shown in the table that immediately follows, Table 3.12. those stem-cup combinations with more than 2,500 procedures in the case of cemented, uncemented, hybrid and reverse hybrid hips, and more than 1,000 in the case of resurfacings.

140 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.11 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at different times after the primary operation, for cup-stem brand combinations with large group sizes (>2,500 or >1,000 in the case of resurfacings). Blue italics indicate that fewer than 100 cases remain at risk.

Cumulative percentage probability of first revision (95% CI) at time shown if time since primary operation is: Stem/cup No. of brand patients 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Cemented: Charnley Cemented 0.37 0.85 1.18 1.46 1.84 2.16 2.48 2.92 3.17 9,099 Stem/ (0.26-0.52) (0.67-1.06 (0.97-1.44) (1.22-1.75) (1.56-2.17) (1.84-2.54) (2.12-2.91) (2.48-3.45) (2.66-3.78) Charnley Ogee Charnley Cemented Stem/ 0.29 0.58 0.80 1.03 1.34 1.64 1.88 2.33 2.49 10,307 Charnley (0.20-0.41) (0.45-0.76) (0.64-1.01) (0.85-1.27) (1.12-1.61) (1.38-1.94) (1.59-2.23) (1.96-2.75) (2.09-2.96) Cemented Cup C-Stem 0.44 0.70 0.90 1.07 1.17 1.28 1.55 1.97 1.97 Cemented/ 3,749 (0.27-0.73) (0.47-1.04) (0.63-1.29) (0.76-1.49) (0.84-1.62) (0.93-1.77) (1.11-2.16) (1.38-2.81) (1.38-2.81) Elite Plus Ogee Stanmore Modular Stem/ 0.38 0.58 0.98 1.10 1.45 1.61 1.61 1.61 1.61 4,135 Stanmore- (0.23-0.63) (0.38-0.89) (0.69-1.38) (0.79-1.54) (1.05-1.98) (1.17-2.21) (1.17-2.21) (1.17-2.21) (1.17-2.21) Arcom Cup 0.69 0.94 1.22 1.50 1.92 2.31 2.73 2.88 2.88 CPT/ZCA 8,234 (0.53-0.90) (0.74-1.18) (0.99-1.51) (1.22-1.84) (1.58-2.33) (1.90-2.81) (2.22-3.36) (2.31-3.58) (2.31-3.58) Exeter V40/ 0.43 0.71 0.93 1.14 1.39 1.65 1.84 2.11 2.30 55,656 Contemporary (0.38-0.49) (0.63-0.78) (0.85-1.03) (1.04-1.26) (1.26-1.52) (1.51-1.82) (1.66-2.03) (1.88-2.37) (1.99-2.64) Exeter V40/ 0.29 0.51 0.74 0.86 1.04 1.31 1.50 1.57 1.57 17,261 Elite Plus Ogee (0.22-0.39) (0.41-0.63) (0.61-0.89) (0.72-1.03) (0.87-1.24) (1.11-1.56) (1.26-1.79) (1.30-1.90) (1.30-1.90) Exeter V40/ 0.58 0.91 1.20 1.51 1.74 2.12 2.54 3.07 3.23 Exeter 13,781 (0.47-0.73) (0.76-1.09) (1.02-1.42) (1.30-1.76) (1.51-2.01) (1.84-2.44) (2.22-2.94) (2.61-3.60) (2.73-3.82) Duration Exeter V40/

Elite Plus 0.43 0.69 0.88 0.95 0.99 1.14 1.33 1.50 1.50 © National Joint Registry 2013 6,124 Cemented (0.29-0.64) (0.50-0.95) (0.65-1.18) (0.71-1.28) (0.74-1.33) (0.83-1.56) (0.95-1.85) (1.03-2.16) (1.03-2.16) Cup Exeter V40/ 0.67 0.91 4,024 ------Exeter Rimfit (0.42-1.05) (0.55-1.51) C-Stem AMT Cemented Stem/ 0.44 0.54 0.75 0.92 1.12 1.12 2,658 - - - Elite Plus (0.24-0.79) (0.31-0.92) (0.45-1.25) (0.57-1.49) (0.70-1.78) (0.70-1.78) Cemented Cup Uncemented: Accolade/ 0.97 1.62 2.18 2.60 3.22 3.55 3.88 3.88 3.88 16,940 Trident (0.83-1.13) (1.43-1.85) (1.94-2.46) (2.31-2.93) (2.83-3.67) (3.05-4.12) (3.12-4.81) (3.12-4.81) (3.12-4.81) Corail/Duraloc 0.72 1.21 1.66 1.88 2.28 2.64 3.11 3.78 4.66 Cementless 4,023 (0.50-1.04) (0.92-1.61) (1.30-2.12) (1.49-2.37) (1.83-2.83) (2.13-3.27) (2.51-3.84) (3.02-4.73) (3.51-6.17) Cup 0.82 1.32 1.86 2.49 3.14 4.03 4.89 5.89 8.07 Corail/Pinnacle 67,724 (0.75-0.89) (1.23-1.42) (1.74-1.98) (2.34-2.65) (2.94-3.34) (3.75-4.33) (4.48-5.34) (5.11-6.79) (5.70-11.36) Corail/ASR 1.00 3.29 7.47 14.25 23.28 33.82 39.04 43.30 Resurfacing 2,597 - (0.68-1.47) (2.67-4.05) (6.52-8.56) (12.94-15.69) (21.56-25.12) (31.50-36.26) (36.12-42.10) (38.30-48.66) Cup Furlong HAC 0.91 1.30 1.61 1.85 2.02 2.27 2.58 2.89 2.97 14,852 Stem/CSF (0.77-1.08) (1.12-1.50) (1.42-1.84) (1.63-2.09) (1.79-2.27) (2.01-2.55) (2.29-2.91) (2.54-3.28) (2.59-3.40) Continued >

www.njrcentre.org.uk 141 Table 3.11 (continued)

Cumulative percentage probability of first revision (95% CI) at time shown if time since primary operation is: Stem/cup No. of brand patients 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Furlong HAC 1.10 1.54 1.70 1.87 2.01 2.01 Stem/CSF 11,391 - - - (0.92-1.31) (1.32-1.81) (1.46-1.99) (1.58-2.21) (1.63-2.47) (1.63-2.47) Plus SL-Plus Cementless 1.27 2.22 2.85 3.37 4.26 4.69 5.49 5.75 5.75 4,186 Stem/ EP/Fit (0.97-1.67) (1.78-2.70) (2.37-3.43) (2.83-4.01) (3.59-5.04) (3.92-5.60) (4.45-6.76) (4.62-7.14) (4.62-7.14) Plus Taperloc Cementless 1.15 1.44 1.70 1.93 2.34 2.72 9,893 - - - Stem/Exceed (0.95-1.39) (1.21-1.73) (1.42-2.04) (1.59-2.34) (1.85-2.96) (2.06-3.59) ABT Hybrid: 0.83 1.11 1.21 1.62 1.94 2.26 2.48 2.63 3.04 CPT/Trilogy 8,987 (0.66-1.05) (0.90-1.37) (0.99-1.49) (1.33-1.97) (1.60-2.35) (1.85-2.75) (2.01-3.04) (2.10-3.29) (2.19-4.21) Exeter V40/ 0.48 0.79 1.02 1.13 1.29 1.68 1.68 1.68 2,803 - Pinnacle (0.27-0.85) (0.50-1.26) (0.66-1.59) (0.73-1.75) (0.82-2.03) (0.95-2.98) (0.95-2.98) (0.95-2.98) Exeter V40/ 0.59 0.90 1.09 1.25 1.48 1.73 2.06 2.31 2.65 28,367 Trident (0.51-0.69) (0.79-1.03) (0.96-1.24) (1.10-1.41) (1.30-1.67) (1.51-1.97) (1.75-2.42) (1.91-2.80) (1.96-3.58) Exeter V40/ 0.57 0.80 1.02 1.20 1.38 1.64 1.80 2.12 2.30 10,375 Trilogy (0.44-0.73) (0.64-1.00) (0.83-1.25) (0.98-1.45) (1.13-1.67) (1.35-1.99) (1.47-2.20) (1.66-2.71) (1.75-3.01) Reverse hybrid: Corail/ Elite Plus 0.54 0.93 1.26 1.52 2.04 2.04 2.04 2.04 2,559 - Cemented (0.31-0.93) (0.60-1.42) (0.85-1.86) (1.04-2.21) (1.36-3.04) (1.36-3.04) (1.36-3.04) (1.36-3.04) Cup Corail/ 0.38 0.65 0.93 0.93 0.93 © National Joint Registry 2013 3,140 - - - - Marathon (0.20-0.70) (0.39-1.09) (0.57-1.53) (0.57-1.53) (0.57-1.53) Resurfacing: Adept 1.23 1.85 2.57 3.54 5.00 6.67 8.70 8.70 Resurfacing 3,293 - (0.90-1.67) (1.44-2.39) (2.06-3.20) (2.90-4.31) (4.13-6.03) (5.45-8.15) (6.91-10.93) (6.91-10.93) Cup ASR 1.59 3.44 5.91 9.44 13.71 18.57 24.07 29.69 36.40 Resurfacing 3,026 (1.20-2.10) (2.85-4.16) (5.12-6.81) (8.44-10.55) (12.48-15.05) (17.06-20.19) (22.18-26.11) (27.03-32.55) (30.94-42.48) Cup BHR 1.05 1.66 2.38 3.15 3.85 4.69 5.61 6.61 8.11 Resurfacing 18,280 (0.92-1.21) (1.49-1.86) (2.17-2.62) (2.89-3.43) (3.56-4.17) (4.35-5.06) (5.21-6.04) (6.12-7.13) (7.40-8.88) Cup Cormet 2000 1.44 2.67 3.55 5.27 7.39 8.96 11.41 13.85 16.34 Resurfacing 3,629 (1.10-1.88) (2.19-3.25) (2.99-4.21) (4.57-6.07) (6.52-8.36) (7.96-10.08) (10.14-12.83) (12.23-15.65) (14.19-18.78) Cup Durom 1.40 2.78 3.83 4.69 5.89 7.08 9.45 9.80 9.80 Resurfacing 1,660 (0.93-2.10) (2.08-3.70) (2.99-4.90) (3.74-5.87) (4.79-7.24) (5.80-8.63) (7.74-11.53) (7.99-12.01) (7.99-12.01) Cup Recap 1.85 2.77 3.56 4.36 6.04 7.43 8.66 8.66 1.680 - Magnum (1.31-2.63) (2.07-3.69) (2.75-4.60) (3.43-5.55) (4.78-7.62) (5.83-9.45) (6.51-11.49) (6.51-11.49) Conserve Plus 1.98 3.62 5.03 6.83 8.23 10.12 10.63 12.22 12.22 Resurfacing 1,320 (1.35-2.89) (2.73-4.78) (3.96-6.38) (5.54-8.41) (6.74-10.02) (8.28-12.36) (8.59-13.11) (8.93-16.62) (8.93-16.62) Cup

The cemented, uncemented and hybrid stem-cup Table 3.12 shows the estimated cumulative brand sub-groups with more than 10,000 procedures percentage probabilities for the resulting fixation/ in the preceding table (Table 3.11) have been further bearing sub-groups in which there were more than divided by bearing surface. 1,000 procedures.

142 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.12 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at different times after the primary operation for cup-stem brand combinations with large group sizes (>10,000) with further subdivision by main bearing surface (provided the sub-group size >1,000). Blue italics indicate that fewer than 100 cases remain at risk.

Cumulative percentage probability of first revision (95% CI) at time shown if time since primary operation is: Stem/cup Bearing No. of brand surface patients 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Cemented: Exeter V40/ 0.42 0.70 0.92 1.14 1.39 1.65 1.83 2.12 2.33 MoP 51,912 Contemporary (0.36-0.48) (0.63-0.78) (0.83-1.02) (1.03-1.25) (1.26-1.52) (1.49-1.82) (1.65-2.03) (1.88-2.40) (2.00-2.70) 0.48 0.68 1.05 1.18 1.30 1.63 1.87 1.87 1.87 CoP 3,298 (0.29-0.80) (0.44-1.06) (0.71-1.54) (0.81-1.71) (0.88-1.92) (1.07-2.46) (1.20-2.90) (1.20-2.90) (1.20-2.90) Exeter V40/ 0.30 0.50 0.73 0.85 1.02 1.27 1.47 1.55 1.55 Elite Plus MoP 16,026 (0.23-0.40) (0.40-0.63) (0.60-0.90) (0.70-1.02) (0.85-1.22) (1.06-1.52) (1.22-1.77) (1.27-1.89) (1.27-1.89) Ogee 0.21 0.68 0.96 1.14 1.41 1.41 1.41 1.41 1.41 CoP 1,042 (0.05-0.85) (0.31-1.51) (0.48-1.92) (0.59-2.20) (0.74-2.71) (0.74-2.71) (0.74-2.71) (0.74-2.71) (0.74-2.71) Uncemented: Accolade/ 0.97 1.56 2.25 2.71 3.24 3.39 3.39 3.39 MoP 8,281 - Trident (0.77-1.21) (1.29-1.88) (1.89-2.66) (2.28-3.23) (2.63-3.98) (2.73-4.20) (2.73-4.20) (2.73-4.20) 0.85 1.83 2.22 2.79 3.43 3.43 3.43 3.43 3.43 CoP 2,171 (0.52-1.39) (1.25-2.67) (1.53-3.22) (1.86-4.20) (2.26-5.21) (2.26-5.21) (2.26-5.21) (2.26-5.21) (2.26-5.21) 1.02 1.69 2.16 2.49 3.16 3.60 4.10 4.10 4.10 CoC 6,319 (0.80-1.31) (1.38-2.07) (1.79-2.60) (2.07-2.99) (2.61-3.83) (2.89-4.47) (3.02-5.57) (3.02-5.57) (3.02-5.57) Corail/ 0.86 1.21 1.47 1.69 1.82 2.19 2.41 2.62 2.62 MoP 23,949 Pinnacle (0.75-0.99) (1.07-1.37) (1.30-1.66) (1.50-1.92) (1.60-2.08) (1.89-2.55) (2.03-2.87) (2.10-3.29) (2.10-3.29) 0.83 1.55 2.38 3.51 4.84 6.51 8.45 11.10 19.89 MoM 11,711 (0.68-1.01) (1.34-1.79) (2.12-2.68) (3.18-3.89) (4.41-5.32) (5.91-7.17) (7.52-9.50) (9.06-13.58) (11.37-33.48) 0.65 0.96 1.25 1.60 1.84 1.84 1.84 1.84 CoP 6,134 - (0.47-0.89) (0.72-1.28) (0.94-1.66) (1.19-2.14) (1.34-2.51) (1.34-2.51) (1.34-2.51) (1.34-2.51) 0.83 1.37 1.83 2.26 2.53 2.91 3.05 3.05 CoC 23,496 - (0.72-0.96) (1.21-1.55) (1.63-2.05) (2.01-2.54) (2.23-2.86) (2.50-3.39) (2.57-3.62) (2.57-3.62) 0.46 1.28 2.86 3.72 5.06 5.06 5.06 5.06 CoM 1,742 - (0.23-0.93) (0.83-1.95) (2.07-3.95) (2.64-5.22) (2.85-8.90) (2.85-8.90) (2.85-8.90) (2.85-8.90) Furlong HAC 0.98 1.41 1.82 2.05 2.23 2.76 2.91 3.48 3.48 MoP 6,729 © National Joint Registry 2013 Stem/CSF (0.77-1.25) (1.14-1.73) (1.51-2.19) (1.71-2.44) (1.87-2.65) (2.32-3.27) (2.45-3.46) (2.84-4.26) (2.84-4.26) 0.74 1.07 1.28 1.53 1.63 1.68 1.93 1.99 2.13 CoP 6,228 (0.55-0.98) (0.84-1.37) (1.02-1.60) (1.24-1.89) (1.33-2.00) (1.37-2.06) (1.57-2.37) (1.62-2.45) (1.69-2.68) 1.25 1.70 2.03 2.30 2.59 2.68 3.70 4.44 4.44 CoC 1,608 (0.81-1.93) (1.17-2.47) (1.44-2.85) (1.66-3.18) (1.90-3.51) (1.98-3.63) (2.74-4.98) (3.25-6.06) (3.25-6.06) Furlong HAC 1.35 1.73 1.87 2.40 3.02 Stem/CSF MoP 2,626 - - - - (0.97-1.89) (1.27-2.35) (1.38-2.52) (1.70-3.40) (1.86-4.89) Plus 1.11 1.58 1.73 1.73 1.73 CoP 1,281 - - - - (0.64-1.91) (0.98-2.55) (1.08-2.76) (1.08-2.76) (1.08-2.76) 1.00 1.47 1.65 1.70 1.70 1.70 CoC 7,422 - - - (0.79-1.27) (1.20-1.81) (1.35-2.01) (1.39-2.09) (1.39-2.09) (1.39-2.09) Hybrid: Exeter V40/ 0.64 0.98 1.18 1.30 1.50 1.68 2.18 2.36 2.36 MoP 15,113 Trident (0.52-0.78) (0.82-1.17) (1.00-1.40) (1.10-1.54) (1.26-1.78) (1.39-2.04) (1.66-2.86) (1.76-3.17) (1.76-3.17) 0.55 0.77 1.02 1.02 1.68 1.68 2.16 2.16 2.16 CoP 3,301 (0.33-0.89) (0.50-1.18) (0.68-1.52) (0.68-1.52) (1.00-2.80) (1.00-2.80) (1.19-3.90) (1.19-3.90) (1.19-3.90) 0.53 0.82 0.99 1.19 1.42 1.74 1.98 2.31 2.89 CoC 9,375 (0.40-0.70) (0.65-1.04) (0.80-1.23) (0.97-1.47) (1.16-1.74) (1.41-2.14) (1.57-2.49) (1.74-3.07) (1.84-4.53) Exeter V40/ 0.54 0.75 0.96 1.10 1.33 1.61 1.75 2.07 2.07 MoP 8,210 Trilogy (0.40-0.73) (0.58-0.97) (0.76-1.22) (0.88-1.39) (1.07-1.67) (1.28-2.01) (1.38-2.21) (1.54-2.76) (1.54-2.76) CoP 1,876 0.55 0.79 1.08 1.42 1.42 1.69 1.93 2.32 2.32 (0.29-1.01) (0.47-1.34) (0.68-1.72) (0.93-2.17) (0.93-2.17) (1.11-2.55) (1.24-2.98) (1.42-3.77) (1.42-3.77)

www.njrcentre.org.uk 143 3.2.5 Revisions after primary hip In the case of (i), metal-on-polyethylene monobloc- cup, there were significant differences between the surgery: Effect of head sizes for five head sizes shown (P<0.001). Rates of revision for polyethylene liners 22.25mm head size did not differ significantly from the baseline of 28mm heads (Hazard rate ratio (HRR) Four sub-groups were identified with polyethylene liners, 1.05 (95% CI 0.95-1.17) P=0.31). Revision rates for i. metal-on-polyethylene monobloc-cup (n=185,027) 26mm heads were significantly lower (0.80 (0.69- 0.93) P=0.003) than for 28mm heads and rates for ii. (metal-on-polyethylene - metal shell with both 32mm and 36mm were higher (1.19 (1.02-1.40) polyethylene liners (n=129,152) P=0.031 & 2.08 (1.27-3.41) P=0.004, respectively). iii. ceramic-on-polyethylene – polyethylene monobloc-cup (n=20,841) For (ii) metal-on-polyethylene with metal shell/ iv. ceramic-on-polyethylene - metal shell polyethylene liner, the differences between the head with polyethylene liners (n=37,278) sizes were not significant overall (P=0.055), however the revision rates were higher for 44mm heads than The Graphs (i)-(iv) in Figure 3.5 show the respective the baseline 28mm heads (HRR 1.84, 95% CI 1.17- cumulative hazards broken down by head size. The 2.90, P=0.008) whereas other head sizes did not differ same scale has been used for all graphs to facilitate significantly (minimum P=0.14). comparison between them. Only head sizes where there were more than 500 procedures are shown. For the other groups, (iii) and (iv), there were no significant differences amongst the sets of head sizes A Cox ‘proportional hazards’ regression analysis was shown (P=0.69, P=0.15, respectively), although there performed to compare the head sizes in each case. were too few cases for larger head sizes to make The most frequently used head size, 28mm, was used worthwhile comparisons. as the baseline for comparison in each analysis.

144 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.5 (i) Comparison of the cumulative hazard for revision for different head sizes for metal-on-polyethylene monobloc cup. (NB only head sizes where n>500 are shown)

Cumulative hazard x 100 1

0 © National Joint Registry 2013 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Head size 22.5mm 29,582 27,420 25,021 22,410 19,585 16,573 13,107 9,701 5,922 2,312 Head size 26mm 16,642 15,416 14,049 12,742 10,951 8,990 6,943 4,945 2,772 965 Head size 28mm 120,152 103,028 86,715 71,059 55,970 40,629 26,182 15,606 7,224 2,260 Head size 32mm 16,784 11,586 7,925 5,258 3,317 2,056 1,172 639 291 78 Head size 36mm 1,331 746 393 165 30 3 0 0 0 0

www.njrcentre.org.uk 145 Figure 3.5 (ii) Comparison of the cumulative hazard for revision for different head sizes for metal-on-polyethylene with metal shell/polyethylene liners. (NB only head sizes where n>500 are shown)

Cumulative hazard x 100 1

0 © National Joint Registry 2013 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Head size 22.5mm 850 722 608 527 474 404 333 239 139 45 Head size 26mm 737 680 615 534 464 386 298 208 108 23 Head size 28mm 68,789 60,999 52,854 44,150 35,350 26,202 17,662 10,739 5,039 1,462 Head size 32mm 32,691 23,357 16,037 10,190 5,840 3,051 1,512 614 219 20 Head size 36mm 22,394 16,227 10,769 6,272 2,997 1,317 476 171 66 14 Head size 40mm 2,928 2,462 1,923 1,382 838 286 25 10 5 1 Head size 44mm 698 571 452 320 193 58 2 1 1 0

146 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.5 (iii) Comparison of the cumulative hazard for revision for different head sizes for ceramic-on-polyethylene with polyethylene monobloc cup. (NB only head sizes where n>500 are shown)

Cumulative hazard x 100 1 © National Joint Registry 2013

0 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Head size 22.25mm 2,314 2,082 1,814 1,599 1,376 1,104 801 485 190 0 Head size 28mm 15,115 12,558 10,310 8,134 6,259 4,506 3,040 1,958 1,032 364 Head size 32mm 3,073 2,074 1,339 829 447 210 120 62 23 9

www.njrcentre.org.uk 147 Figure 3.5 (iv) Comparison of the cumulative hazard for revision for different head sizes for ceramic-on-polyethylene with metal shell/polyethylene liners. (NB only head sizes where n>500 are shown)

Cumulative hazard x 100 1 © National Joint Registry 2013

0 0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

Head size 28mm 18,790 16,743 14,724 12,686 10,569 8,443 6,161 4,125 2,199 748 Head size 32mm 11,265 8,077 5,591 3,497 2,096 1,184 725 406 151 37 Head size 36mm 6,883 4,229 2,657 1,524 711 313 120 21 0 0

3.2.6 Mortality after primary Table 3.13 shows Kaplan-Meier estimates of cumulative percentage mortality at 30 days, 90 days hip surgery and at each anniversary up to the ninth, for all cases This section describes the mortality of the cohort up and by age and gender. to nine years from primary operation, according to Note the cases were not ‘censored’ when further gender and age group. Deaths were updated at the revision surgery was undertaken. Such surgery may end of February 2013 using data from the Patient have contributed to the overall mortality as shown Demographic Service. Two hundred cases were here. The impact of this has not been quantified. excluded because the NHS number was not traceable and therefore no death date could be ascertained. A further three were excluded because of uncertainty in gender (n=1) and age (n=2), leaving 539,169, with a total of 42,805 reported deaths. Amongst these were 3,091 bilateral operations, with the left and right side operated on the same day; here the second of the two has been excluded, leaving 536,078 procedures and 42,655 deaths.

148 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report 7.97 4.39 7.02 4.29 6.31 21.41 16.52 25.18 11.75 41.92 18.54 65.67 29.13 10.22 52.16 9 years (7.34-8.65) (3.88-4.96) (6.27-7.86) (3.79-4.86) (5.62-7.07) (21.11-21.71) (9.42-11.09) (15.58-17.52) (24.00-26.40) (11.03-12.51) (40.13-43.76) (17.70-19.41) (63.62-67.72) (27.99-30.30) (50.87-53.47) 7.02 9.68 3.76 5.97 3.81 8.81 5.39 18.14 14.03 21.34 34.49 15.15 57.38 23.75 45.07 8 years (6.54-7.54) (3.39-4.17) (5.42-6.58) (3.44-4.22) (8.23-9.44) (4.91-5.92) (17.94-18.35) (9.19-10.19) (13.34-14.76) (20.51-22.21) (33.30-35.72) (14.58-15.75) (55.89-58.89) (23.00-24.52) (44.14-46.01) 5.77 7.92 3.23 4.78 3.31 7.27 4.60 14.97 11.29 17.23 28.15 12.30 49.14 19.21 37.35 7 years (5.40-6.16) (7.55-8.31) (2.93-3.55) (4.37-5.24) (3.02-3.63) (6.83-7.73) (4.22-5.01) (14.82-15.13) (10.77-11.84) (16.61-17.87) (27.24-29.07) (11.86-12.76) (47.94-50.35) (18.65-19.79) (36.62-38.08) 4.65 6.39 2.64 9.78 3.84 2.79 6.21 3.77 8.94 12.17 14.00 22.94 41.01 15.40 30.17 6 years (8.53-9.37) (4.37-4.96) (6.09-6.69) (2.40-2.89) (3.51-4.20) (2.56-3.05) (5.85-6.60) (3.46-4.10) (9.43-10.13) (12.05-12.30) (13.51-14.52) (22.21-23.70) (40.01-42.03) (14.95-15.86) (29.58-30.77) 9.54 3.79 5.02 2.08 7.50 3.19 2.33 4.87 3.04 7.03 11.05 17.55 32.76 11.89 23.36 5 years (9.43-9.64) (3.55-4.04) (4.78-5.27) (1.89-2.28) (7.22-7.79) (2.91-3.49) (2.13-2.54) (4.58-5.18) (2.79-3.32) (6.70-7.38) (10.64-11.47) (16.97-18.15) (31.91-33.62) (11.53-12.26) (22.88-23.85) 7.15 2.85 8.41 3.73 1.70 5.38 2.62 8.64 2.01 3.85 2.37 5.26 12.93 24.90 17.26 4 years (7.06-7.23) (2.66-3.05) (8.08-8.76) (3.54-3.93) (1.54-1.87) (5.16-5.61) (2.38-2.89) (8.35-8.94) (1.84-2.20) (3.60-4.10) (2.16-2.60) (4.99-5.55) (12.47-13.41) (24.19-25.62) (16.87-17.66) 4.93 1.99 5.77 2.57 9.06 1.28 3.61 1.78 5.83 1.61 2.71 1.69 3.72 17.16 11.63 3 years (4.86-5.00) (1.84-2.14) (5.51-6.04) (2.42-2.73) (8.69-9.44) (1.15-1.43) (3.45-3.79) (1.60-1.99) (5.61-6.06) (1.46-1.77) (2.52-2.92) (1.52-1.87) (3.50-3.95) (16.60-17.75) (11.32-11.94)

3.06 1.26 3.52 1.59 5.38 0.87 2.22 1.05 3.56 1.10 1.71 7.16 1.20 2.27 10.85 2 years (3.01-3.11) (1.15-1.38) (3.33-3.73) (1.48-1.71) (5.11-5.67) (0.77-0.98) (2.10-2.35) (0.92-1.20) (3.40-3.74) (0.99-1.23) (1.57-1.87) (6.93-7.40) (1.06-1.35) (2.12-2.44) (10.41-11.30) Cumulative percentage probability of first revision (95% CI) at time shown if since primary operation is: of first revision probability Cumulative percentage 1.52 0.61 1.69 0.77 2.63 0.49 1.02 5.58 0.57 1.64 0.68 0.87 3.60 0.63 1.16 1 year (1.49-1.56) (0.53-0.69) (1.57-1.83) (0.70-0.85) (2.45-2.83) (0.42-0.57) (0.94-1.11) (5.28-5.90) (0.48-0.68) (1.53-1.76) (0.59-0.77) (0.77-0.97) (3.44-3.77) (0.54-0.74) (1.06-1.28) 0.51 0.17 0.50 0.23 0.84 0.16 0.32 2.18 0.17 0.53 0.21 0.25 1.31 0.20 0.41 90 days (0.50-0.53) (0.14-0.22) (0.44-0.58) (0.19-0.28) (0.74-0.95) (0.12-0.21) (0.28-0.37) (1.99-2.38) (0.12-0.23) (0.47-0.59) (0.16-0.27) (0.20-0.31) (1.22-1.41) (0.15-0.26) (0.35-0.48) 0.25 0.08 0.23 0.09 0.47 0.08 0.14 1.18 0.05 0.28 0.06 0.13 0.59 0.09 0.20 30 days (0.24-0.26 (0.04-0.09 (0.06-0.11) (0.19-0.29) (0.07-0.12) (0.39-0.55) (0.06-0.12) (0.12-0.18) (1.05-1.33) (0.03-0.09) (0.24-0.33) (0.09-0.17) (0.53-0.66) (0.06-0.14) (0.16-0.25) No. of 41,432 38,136 53,092 30,152 32,565 59,720 22,579 22,505 53,608 31,939 33,235 54,214 26,083 36,818 Kaplan-Meier estimates of the cumulative percentage mortality (95% CI), at different times after primary hip operation, for all cases and by mortality (95% CI), at different Kaplan-Meier estimates of the cumulative percentage 536,078 patients

All cases By gender and age group By gender and age group 60-64 years 70-74 years 65-69 years 75-79 years Males: <55 years 70-74 years 80+ years 55- 59 years 75-79 years Females: <55 years

60-64 years 80+ years 55- 59 years 65-69 years Table 3.13 Table age/gender. © National Joint Registry 2013 Registry Joint National ©

www.njrcentre.org.uk 149 3.2.7 In-depth study: Metal-on-metal There were substantial differences between the bearing groups in terms of patient characteristics. To reduce the hip resurfacing effects of confounding, analysis was based on the large Since the last annual report we have examined metal- majority with an American Society of Anaesthesiologists on-metal hip resurfacing in greater depth. A research (ASA) grade of 1 or 2 at time of primary surgery and paper comparing the failure rates of metal-on-metal hip those whose surgery was undertaken for osteoarthritis resurfacings with alternative forms of hip replacement only (Table 1). The now-withdrawn ASR implants was published in the Lancet in November 2012. (2,829, 8.9% of the total) were excluded from the resurfacing analysis so as not to distort the results. Smith AJ, Dieppe P, Howard PW, Blom AW; National Furthermore, sensitivity analysis confirmed that the Joint Registry for England and Wales. Failure rates of results for the newer technologies of resurfacing and metal-on-metal hip resurfacings: Analysis of data from ceramic-on-ceramic were not distorted by a ‘learning the National Joint Registry for England and Wales. curve’ effect of less experienced surgeons (see web Lancet. 2012 Nov 17;380(9855):1759-66 appendix published in Lancet).

We have analysed data from the NJR to test the Multivariable analysis has been used to adjust for following hypotheses: larger head size is related to patient age and ASA grade and to measure the effect higher failure rates; metal-on-metal has significantly of head size. As head size and gender were highly worse results in women; and hip resurfacing gives correlated, separate models were estimated for men better implant survivorship than stemmed THR in and women to avoid multicollinearity. Separate models younger patients. were also specified for the different bearing groups.

Methods Flexible parametric survival models that estimate the cumulative incidence of revision in the presence of the Analysis was based on 434,560 THRs between April competing risk of death were used. Standard survival 2003 and September 2011 where patient-identifiers analysis treats death simply as censored information are present that allow revisions to be linked to primary but this has been shown to overestimate revision operations (81.8% of 531,247 operations). These rates. In all models, head size, age and ASA grade operations were performed in 447 units under the care were selected as predictors of revision and age and of 2,584 consultant surgeons. ASA grade were predictors of the competing risk of The analysis estimated all-cause revision rates. The death. The effect of age was allowed to differ for the unit of analysis was implant (rather than patient) so main and competing risks. These models produce bilateral procedures were included (2,645). We have hazard ratios, which are a measure of relative risk compared revision rates for three types of bearing (averaged over time). To illustrate the absolute effect of surface: resurfacing, ceramic-on-ceramic, and metal- these factors, the models have been used to predict on-polyethylene. A range of commonly used head revision rates for a typical patient by estimating the sizes were compared for the resurfacing and ceramic- covariate-adjusted cumulative incidence function in the on-ceramic groups whereas the most commonly used presence of competing risks. head size of 28mm (Table 1) was been chosen for the One resurfacing brand, the Birmingham Hip metal-on-polyethylene group. A confounding variable Resurfacing system (BHR), was used in more than was the fixation of the implant. We have addressed half of cases giving large enough numbers to repeat this by reporting ceramic-on-ceramic results for the multivariable analysis approach for the BHR cases uncemented fixation as this was most commonly alone (15,386/26,199, 58.9%). This enabled the used (45,099/57,748, 78.1%) while for metal-on- potentially confounding influence of different brand polyethylene, results were shown separately for effects to be removed. We have compared the BHR cemented, uncemented and hybrid fixations. with the most commonly used alternatives in each group: the Exeter V40 Contemporary metal-on-

150 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

polyethylene 28mm cemented THR (23,906/69,569, of the competing risk of death but not of revision for 34.4%) and the Corail Pinnacle 36mm ceramic-on- either men or women. ASA grade was a significant ceramic uncemented THR (9,604/18,370, 52.3%). predictor of revision for men only (hazard ratio 1.267, 95% CI 1.042-1.540, p=0.018) indicating that patients The reasons for revision across all groups were also with ASA grade 2 were more likely to be revised than examined by dividing the number of revisions for each those with ASA grade 1. reason by the total person time at risk of revision (per 1,000 patient-years). This was equivalent to a person- Predicted revision rates from these models illustrate time incidence rate and 95% confidence intervals have the absolute effect of head size. For a 55-year-old been estimated assuming a Poisson distribution. man, the five-year revision rate with a 46mm head was 4.1% compared with 2.6% for a 54mm head (Table Results L2). Resurfacing revision rates for women were much Overall, 7.4% (31,932/434,560) of THRs were higher than for men (five-year revision rate of 8.3% resurfacing procedures. Annual numbers in England with a 42mm head size) (Table L3). This was true even and Wales reached a peak of 6,650 operations in with the same head size: a 55-year-old woman had 2007 but thereafter fell to around 2,000 in 2011. a five-year revision rate of 6.1% with a 46mm head Resurfacing patients were most commonly aged 59 size compared with 4.1% for a 55-year-old man. Only years while the median age was 55. Overall, 78.4% 0.4% (35/9,856) of women had a head size of 54mm or of patients were aged 45-65 years and 69.1% of above associated with the lowest revision rates in men. resurfacings were undertaken on men (Table L1). Resurfacing procedures with a large head size Unadjusted analysis suggested that revision rates for (54mm or greater) resulted in revision rates which resurfacing procedures were higher than for stemmed were not significantly worse than those of other THR (five-year revision rate of 5.2% (95% CI 4.9-5.5%) common surgical options (Table L2). In men, a compared with 2.8% (95% CI 2.7-2.9%). However, 54mm resurfacing head had a five-year revision rate this overall rate disguised differences according to of 2.6% compared with 2.1% for an uncemented gender (Figure L1). In women, 8.5% of resurfacings 40mm ceramic-on-ceramic prosthesis and 1.9% had been revised by five years (95% CI 7.8-9.2%) for a cemented 28mm metal-on-polyethylene compared with 3.6% in men (95% CI 3.3-3.9%). A articulation. However, the smaller resurfacing head head size of between 46mm and 54mm was used for sizes compared less well. Key results are illustrated 92.1% (20342/22076) of male resurfacing patients in Figure L2. Overall, just 23% (5,085/22,076) of male while 76.5% (7541/9856) of females had a head size resurfacings used the 54mm or larger head size that of 42-46mm. Just 1% (98/9,856) of women received had the lowest revision rates. the larger head sizes above 50mm. In total, 15 In women, resurfacing, even with larger heads, offered resurfacing brands were used. significantly poorer implant survivorship than all other The multivariable models for resurfacing procedures common surgical options (Table L3). In 55-year-old confirmed that head size was an independent women, a 46mm resurfacing procedure had a fiveyear predictor of revision for both men (hazard ratio 0.951, revision rate of 6.1% compared with 2.5% for a 95% CI 0.945-0.978, p<0.0005 per unit increase 36mm ceramic-on-ceramic prosthesis and 1.5% for in head size) and women (hazard ratio 0.921, 95% a cemented 28mm metal-on-polyethylene articulation CI 0.892-0.951, p<0.0005) indicating that smaller (Figure L3). head sizes were more likely to be revised. These Generally, the BHR had better implant survivorship hazard ratios can be broadly interpreted as each than other resurfacing brands. Unadjusted analysis 1mm increase in head size being associated with a for all male patients shows a five-year revision rate of 5-8% reduction in the hazard (the risk of revision at a 2.4% (95% CI 2.1-2.7%) for the BHR compared with particular point in time). Age was a significant predictor 8.1% (95% CI 6.7-9.8%) for the now-withdrawn ASR

www.njrcentre.org.uk 151 and 4.2% (95% CI 3.7-4.9%) for other resurfacing Resurfacings were most commonly revised for pain, brands. The equivalent five-year revision rates in periprosthetic fracture, and aseptic loosening (Table women were more than twice as high for all brands L5). Revision for periprosthetic fracture was up to (5.8% (95% CI 5.1-6.6%) for the BHR, 21.0% (95% eight times more common in resurfacing than in other CI 17.7-24.9%) for the ASR and 9.2% (95% CI 8.1- surgical options. Revisions for pain in women were 10.5%) for other brands). However, the multivariable up to 10 times higher with resurfacing than with other models indicated that in men aged 55, the BHR with surgical options and revisions for aseptic loosening the 54mm head size did not demonstrate better were up to four times higher. Revisions for dislocation implant survivorship than the Exeter/Contemporary in resurfacing were lower than other options in men. or the Corail/Pinnacle and had considerably worse implant survivorship in women (Table L4).

Figure L1 Cumulative hazard of revision after resurfacing by gender (with 95% CI).

14%

12%

10%

6% © National Joint Registry 2013 4%

2% Female Male 0 1 2 3 4 5 6 7 Years since primary surgery

152 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure L2 Estimated cumulative incidence of revision for 55-year-old male by prosthesis type.

5% Resurfacing 46mm Resurfacing 48mm 4% Resurfacing 50mm Resurfacing 52mm 3% Resurfacing 54mm 2% Uncemented ceramic-on-ceramic

40mm © National Joint Registry 2013 1% Cemented Cumulative incidence of revision metal-on-polyethylene 0 28mm 0 1 2 3 4 5 6 7 Years since primary surgery

Figure L3 Estimated cumulative incidence of revision for 55-year-old female by prosthesis type.

12% 11% 10% 9% 8% Resurfacing 42mm 7% Resurfacing 44mm 6% Resurfacing 46mm 5% 4% Uncemented ceramic-on-ceramic 3%

36mm © National Joint Registry 2013 2% Cemented Cumulative incidence of revision 1% metal-on-polyethylene 0 28mm 0 1 2 3 4 5 6 7 Years since primary surgery

www.njrcentre.org.uk 153 Table L1 Description of comparison groups: number of cases (percentage of total).

Bearing surface Ceramic-on-ceramic Metal-on-polyethylene Resurfacing (N=31,932) (N=57,748) (N=226,165) Head size Less than 28mm 0 (0.0%) 0 (0.0%) 22,140 (9.9%) 28mm 0 (0.0%) 12,274 (21.3%) 154,305 (68.2%) 30-34mm 0 (0.0%) 18,680 (32.3% 30,059 (13.3%) 36-40mm 577 (1.8%) 26,339 (46.0%) 16,113 (7.2%) 42mm 2,107 (6.6%) 0 (0.0%) 0 (0.0%) 44mm 2,032 (6.4%) 0 (0.0%) 349 (0.2%) 46mm 5,626 (17.6%) 0 (0.0%) 0 (0.0%) 48mm 3,735 (11.7%) 0 (0.0%) 0 (0.0%) 50mm 8,905 (27.9%) 0 (0.0%) 0 (0.0%) 52mm 3,397 (10.6%) 0 (0.0%) 0 (0.0%) 54mm 4,054 (12.7%) 0 (0.0%) 0 (0.0%) 56-62mm 1,066 (3.3%) 0 (0.0%) 0 (0.0%) Unknown size 379 (1.2%) 455 (0.8%) 3,199 (1.4%) Age (mean age in years) 54.1 59.6 72.7 Aged less than 40 years 1,934 (6.1%) 2,874 (5.0%) 558 (0.3%) Aged 40-49 6,877 (21.5%) 6,065 (10.5%) 1,929 (0.9%) Aged 50-59 14,067 (44.1%) 16,877 (29.2%) 12,451 (5.5%) Aged 60-69 8,229 (25.8%) 22,918 (39.7%) 58,865 (26.0%) Aged 70 or over 825 (2.6%) 9,014 (15.6%) 152,362 (67.4%) Gender Male 22,076 (69.1%) 25,290 (43.8%) 78,891 (34.9%) Female 9,856 (30.9%) 32,458 (56.2%) 147,274 (65.1%) ASA grade 1 - fit and healthy 15,613 (48.9%) 16,048 (27.8%) 29,320 (13.0%)

© National Joint Registry 2013 2 - mild systemic disease that does not 15,336 (48.0%) 37,196 (64.4%) 157,492 (69.6%) limit activity 3 - systemic disease that limits activity but 947 (3.0%) 4,385 (7.6%) 37,764 (16.7%) is not incapacitating 4/5 - incapacitating, life-threatening systemic disease/not expected to survive 36 (0.1%) 119 (0.2%) 1589 (0.7%) 24 hours without an operation Diagnosis Osteoarthritis only 29,550 (92.5%) 50,626 (87.7%) 208,173 (92.0%) Other 2,382 (7.5%) 7,122 (12.3%) 17,992 (8.0%) Total (% of all THR) 31,932 (7.4%) 57,748 (13.3%) 226,165 (52.0%) Multivariable model samples Total in model 26,119 (81.8%) 36,009 (62.4%) 115,574 (51.1%) Number of surgeons 698 (27.0%) 1,202 (46.5%) 2,144 (83.0%) Number of units 372 (83.2%) 397 (88.8%) 434 (97.1%) Numbers followed up by year Less than 1 year 1,681 (6.4%) 9,034 (25.1%) 14,871 (12.9%) 1 - < 2 years 2,692 (10.3%) 8,868 (24.6%) 16,967 (14.7%) 2 - < 3 years 3,708 (14.2%) 6,419 (17.8%) 17,257 (14.9%) 3 - < 4 years 4,457 (17.1%) 4,577 (12.7%) 17,852 (15.4%) 4 - < 5 years 4,335 (16.6%) 2,940 (8.2%) 16,906 (14.6%) 5 - < 6 years 3,653 (14.0%) 1,980 (5.5%) 13,218 (11.4%) 6 - < 7 years 2,835 (10.9%) 1,182 (3.3%) 10,265 (8.9%) 7 or more years 2,756 (10.6%) 1,008 (2.8%) 8,237 (7.1%)

Note: among the full sample of hip replacements (N=434,560), there are other bearing surfaces which are not shown in this table: 46,290 ceramic-on-polyethylene (10.7% of the total), 31,171 stemmed metal-on-metal (7.2% of total), 2,430 other combinations (0.6%), and 38,824 unknown bearings (8.9%).

154 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table L2 Predicted revision rates for 55-year-old males by prosthesis and head size (95% CI).

Type/head size Year 1 Year 3 Year 5 Year 7 Resurfacing 46mm 1.57 (1.26-1.95) 2.81 (2.31-3.42) 4.05 (3.34-4.90) 5.23 (4.29-6.38) 48mm 1.40 (1.17-1.68) 2.52 (2.15-2.94) 3.63 (3.12-4.21) 4.69 (3.99-5.51) 50mm 1.26 (1.06-1.48) 2.25 (1.96-2.58) 3.25 (2.85-3.70) 4.20 (3.65-4.84) 52mm 1.12 (0.94-1.34) 2.02 (1.74-2.34) 2.91 (2.52-3.35) 3.77 (3.23-4.38) 54mm 1.00 (0.82-1.23) 1.80 (1.50-2.17) 2.60 (2.18-3.11) 3.37 (2.80-4.06) Ceramic-on-ceramic Uncemented 28mm 1.24 (0.92-1.68) 2.46 (1.87-3.22) 3.41 (2.64-4.40) 4.08 (3.14-5.28) Uncemented 32mm 1.05 (0.84-1.30) 2.07 (1.72-2.50) 2.88 (2.39-3.47) 3.45 (2.81-4.23) Uncemented 36mm 0.88 (0.71-1.10) 1.75 (1.43-2.14) 2.43 (1.96-3.02) 2.91 (2.28-3.71)

Uncemented 40mm 0.74 (0.55-1.01) 1.48 (1.09-2.00) 2.05 (1.49-2.83) 2.46 (1.73-3.47) © National Joint Registry 2013 Metal-on-polyethylene Cemented 28mm 0.56 (0.43-0.73) 1.28 (1.00-1.64) 1.91 (1.50-2.44) 2.40 (1.87-3.06) Hybrid 28mm 1.26 (0.83-1.92) 2.19 (1.48-3.24) 3.21 (2.20-4.67) 4.10 (2.80-5.97) Uncemented 28mm 1.41 (1.03-1.93) 2.57 (1.92-3.42) 3.48 (2.63-4.59) 4.18 (3.14-5.56)

Note: results are estimated from multivariable competing risks flexible parametric survival models based on 18,375 resurfacing cases, 16,136 uncemented ceramic- on-ceramic cases, and for 28mm metal-on-polyethylene: 22,407 cemented, 6,634 hybrid and 9,352 uncemented cases. Results are based on an ASA grade of 2.

Table L3 Predicted revision rates for 55-year-old females by prosthesis and head size (95% CI).

Type/head size Year 1 Year 3 Year 5 Year 7 Resurfacing 42mm 1.81 (1.47-2.24) 5.09 (4.33-5.97) 8.33 (7.18-9.65) 11.67 (10.00-13.58) 44mm 1.54 (1.26-1.88) 4.33 (3.75-4.99) 7.11 (6.24-8.09) 9.99 (8.71-11.43) 46mm 1.31 (1.07-1.60) 3.69 (3.17-4.28) 6.07 (5.28-6.95) 8.54 (7.39-9.85) Ceramic-on-ceramic Uncemented 28mm 0.87 (0.68-1.12) 1.87 (1.50-2.33) 2.63 (2.12-3.25) 3.31 (2.63-4.17) Uncemented 32mm 0.84 (0.69-1.01) 1.81 (1.53-2.13) 2.54 (2.15-3.01) 3.20 (2.62-3.92) Uncemented 36mm 0.81 (0.65-1.02) 1.75 (1.41-2.17) 2.46 (1.95-3.09) 3.10 (2.38-4.03) Metal-on-polyethylene

Cemented 28mm 0.60 (0.32-1.10) 1.12 (0.62-2.01) 1.47 (0.82-2.60) 1.81 (0.99-3.28) © National Joint Registry 2013 Hybrid 28mm 0.70 (0.49-1.02) 1.23 (0.87-1.73) 1.82 (1.30-2.53) 2.48 (1.77-3.47) Uncemented 28mm 0.81 (0.60-1.07) 1.46 (1.12-1.91) 1.94 (1.49-2.53) 2.25 (1.71-2.95)

Note: results are estimated from multivariable competing risks flexible parametric survival models based on 7,744 resurfacing cases, 19,873 uncemented ceramic-on- ceramic cases, and for 28mm metal-on-polyethylene: 47,162 cemented, 13,383 hybrid and 16,636 uncemented cases. Results are based on an ASA grade of 2.

www.njrcentre.org.uk 155 Table L4 Comparison of most commonly used brands; predicted revision rates for 55-year old patients (95% CI).

Type/head size Year 1 Year 3 Year 5 Year 7 Males aged 55 BHR 46mm 1.52 (1.12-2.05) 2.48 (1.87-3.28) 3.43 (2.61-4.50) 4.32 (3.26-5.70) BHR 50mm 1.15 (0.92-1.43) 1.87 (1.54-2.28) 2.60 (2.15-3.12) 3.27 (2.69-3.97) BHR 54mm 0.86 (0.65-1.15) 1.41 (1.09-1.84) 1.96 (1.52-2.52) 2.47 (1.90-3.20) Uncemented Corail /Pinnacle 0.93 (0.65-1.35) 1.90 (1.36-2.64) 2.52 (1.72-3.67) - ceramic-on-ceramic 36mm Cemented Exeter V40/ Contemporary metal-on- 0.53 (0.28-1.02) 1.00 (0.54-1.86) 1.31 (0.71-2.41) 1.61 (0.86-3.02) polyethylene 28mm Females aged 55 BHR 42mm 2.01 (1.52-2.66) 5.02 (4.03-6.25) 8.23 (6.73-10.04) 11.76 (9.59-14.36)

© National Joint Registry 2013 BHR 46mm 1.20 (0.91-1.57) 3.00 (2.43-3.71) 4.96 (4.08-6.00) 7.14 (5.86-8.67) Uncemented Corail/Pinnacle 0.63 (0.42-0.94) 1.53 (1.07-2.20) 2.00 (1.35-2.95) - ceramic-on-ceramic 36mm Cemented Exeter V40/ Contemporary metal-on- 0.42 (0.25-0.71) 0.87 (0.54-1.41) 1.38 (0.86-2.20) 1.94 (1.18-3.17) polyethylene 28mm

Note: results are estimated from multivariable competing risks flexible parametric survival models based on 10,889 male and 4,497 female BHR resurfacing cases, 5,389 male and 4,213 female Corail Pinnacle 36mm ceramic-on-ceramic cases and 7,857 male and 16,049 female cases for the Exeter/Contemporary. Results are based on an ASA grade of 2. Seven-year revision rates cannot be estimated for the Corail/Pinnacle ceramic-on-ceramic 36mm as this has only been in use in England and Wales since 2005.

156 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table L5 Reasons for revision (95% CI), expressed as incidence per 1,000 patient-years, by articulation and fixation.

Uncemented Uncemented metal-on- Hybrid11 metal- Cemented metal- ceramic-on- polyethylene on-polyethylene on-polyethylene Resurfacing ceramic 28mm 28mm 28mm Males Aseptic loosening 1.42 (1.17-1.72) 1.80 (1.42-2.27) 1.88 (1.48-2.41) 0.73 (0.47-1.14) 0.75 (0.58-0.97) Dislocation/subluxation 0.18 (0.10-0.31) 1.13 (0.84-1.52) 1.56 (1.19-2.04) 1.77 (1.32-2.36) 0.75 (0.58-0.97) Implant failure 0.07 (0.03-0.16) 0.46 (0.29-0.73) 0.35 (0.20-0.62) 0.12 (0.04-0.37) 0.10 (0.05-0.20) Implant fracture 0.29 (0.19-0.44) 0.62 (0.41-0.92) 0.03 (0.00-0.21) 0.15 (0.06-0.41) 0.05 (0.02-0.13) Incorrect sizing 0.03 (0.01-0.11) 0.05 (0.01-0.21) 0.03 (0.00-0.21) 0.00 - 0.00 - Infection 0.55 (0.40-0.75) 1.13 (0.84-1.52) 1.12 (0.81-1.54) 1.15 (0.80-1.65) 1.09 (0.88-1.34) Lysis 0.25 (0.16-0.39) 0.08 (0.02-0.24) 0.29 (0.16-0.55) 0.19 (0.08-0.46) 0.24 (0.15-0.37) Malalignment 0.64 (0.48-0.86) 0.59 (0.39-0.89) 0.68 (0.45-1.02) 0.27 (0.13-0.56) 0.21 (0.13-0.34) Other 1.12 (0.90-1.39) 0.69 (0.48-1.01) 0.18 (0.08-0.39) 0.31 (0.15-0.61) 0.15 (0.09-0.26) Pain 1.70 (1.42-2.02) 1.13 (0.84-1.52) 1.00 (0.72-1.40) 0.77 (0.50-1.19) 0.46 (0.34-0.64) Periprosthetic fracture 1.54 (1.28-1.86) 0.64 (0.43-0.95) 0.38 (0.22-0.66) 0.69 (0.44-1.10) 0.51 (0.38-0.70) Person time (years) 73,143.3 38,917.4 33,959.7 26,061.9 79,971.8 Number revised 457 255 185 128 267 Number of operations 18,375 16,136 9,352 6,634 22,407 Females Aseptic loosening 3.73 (3.14-4.43) 1.25 (0.97-1.61) 1.17 (0.92-1.49) 0.83 (0.61-1.12) 0.75 (0.63-0.90) Dislocation/subluxation 0.77 (0.53-1.13) 1.09 (0.83-1.42) 1.72 (1.41-2.10) 1.38 (1.09-1.75) 0.81 (0.69-0.96) Implant failure 0.37 (0.22-0.64) 0.41 (0.26-0.64) 0.21 (0.12-0.38) 0.28 (0.16-0.47) 0.05 (0.03-0.10) © National Joint Registry 2013 Implant fracture 0.32 (0.17-0.57) 0.66 (0.46-0.93) 0.05 (0.02-0.17) 0.06 (0.02-0.18) 0.03 (0.01-0.07) Incorrect sizing 0.09 (0.03-0.27) 0.12 (0.06-0.27) 0.05 (0.02-0.17) 0.02 (0.00-0.14) 0.01 (0.00-0.05) Infection 0.63 (0.42-0.96) 0.41 (0.26-0.64) 0.39 (0.26-0.59) 0.32 (0.19-0.52) 0.56 (0.46-0.69) Lysis 0.86 (0.60-1.23) 0.12 (0.06-0.27) 0.07 (0.03-0.19) 0.18 (0.09-0.34) 0.14 (0.10-0.21) Malalignment 1.15 (0.84-1.56) 0.78 (0.57-1.07) 0.82 (0.61-1.09) 0.47 (0.32-0.71) 0.30 (0.23-0.39) Other 2.78 (2.28-3.40) 0.74 (0.53-1.02) 0.41 (0.27-0.62) 0.18 (0.09-0.34) 0.17 (0.12-0.25) Pain 5.28 (4.57-6.10) 1.00 (0.76-1.33) 0.68 (0.49-0.93) 0.51 (0.35-0.75) 0.41 (0.33-0.52) Periprosthetic fracture 1.61 (1.24-2.09) 0.94 (0.71-1.26) 0.55 (0.39-0.78) 0.32 (0.19-0.52) 0.19 (0.14-0.27) Person time (years) 34,859.4 48,806.1 56,259.4 50,637.5 167,174.5 Number revised 477 274 248 167 411 Number of operations 7,744 19,873 16,636 13,383 47,162

11 Not including reverse hybrid

www.njrcentre.org.uk 157 3.2.8 Conclusions As we now have more data, we have been able to extend brand analysis to more stem/ This year’s report confirms many of the trends shown cup combinations. Brand analysis of stem/cup in previous years. Once again we stress that implant combinations shows higher unadjusted failure rates survivorship alone is an incomplete measure of with uncemented stem/cup combinations than outcome. Implant survivorship gives little indication of with hybrid or all cemented combinations. The satisfaction, relief of pain, improvement in function and best survivorship at nine years is achieved with all- greater participation in society. Readers should also cemented components. understand that each year more data are available and thus analysis can give a more accurate picture. Each The in-depth analysis of hip resurfacing highlights year’s analysis thus supersedes the previous analyses the high failure rate, particularly amongst women. It and may, in places, appear contradictory with previous also demonstrates that failure rate is related to both reports from a smaller dataset. Furthermore, the data gender and head size, with smaller femoral heads and are imperfect and rely on accurate and comprehensive female gender independently associated with higher reporting by surgeons. Some fields, such as body rates of implant revision. In men with the largest head mass index (BMI), were not initially reported but we sizes the failure rates are not significantly worse than are encouraged that in 2012 over 60% of entries alternatives. Only 23% of men who have undergone included BMI data. We would urge surgeons to resurfacing in England and Wales have these larger complete all fields in the NJR forms in order to allow femoral heads. meaningful analysis. This year we have analysed Mortality in the first 30 or 90 days after surgery survivorship for metal-on-polyethylene bearings by remains very low, which is reassuring. However, head size for the first time. We have not attempted mortality at nine years after hip replacement is high in to sub-classify polyethylene as what constitutes so those over the age of 75 suggesting than long-term called third generation polyethylene is not clearly implant survivorship is unlikely to be important in this defined. This analysis shows that there appears to be age group. an association between increased implant failure and larger head sizes.

We note that metal-on-metal stemmed hip replacement and hip resurfacing have virtually ceased with fewer than one in one thousand hip replacements performed in 2012 belonging to each of these classes of implant. The failure rates in these classes continue to be markedly higher than the alternatives. Other bearing surfaces continue to have very low failure rates regardless of fixation, especially for ceramic-on- polyethylene bearings.

158 www.njrcentre.org.uk Part 3 3.3 Outcomes after primary knee replacement In the period 2003 to 2012 589,028 knees were implant to deal with additional ligament deficiency or replaced for the first time. The knee is made bone loss, where a constrained condylar (CCK) or up of three compartments: medial, lateral and hinged knee would be used, even in a primary situation. patello femoral compartments; when a ‘total’ knee The tibial construct may be modular with a metallic replacement (TKR) is implanted two out of the three tibial tray and a polyethylene insert or non-modular compartments are always replaced (medial and being constructed of polyethylene alone. In recent years lateral) and the patella is resurfaced if the surgeon all-polyethylene tibial components have increased in considers this to be of benefit to the patient. If popularity so as to contain costs of the implant. a single compartment is replaced then the term ‘unicompartmental’ is applied to the implant (UKR). The tibial insert may be mobile or remain in a fixed The medial, lateral or patello-femoral compartments position on the tibial tray. This also applies to medial can be replaced independently, if clinically appropriate. and lateral unicompartmental knees. Many brands of total knee implant exist in fixed and mobile forms with There is variation in the constraint of the tibial insert CR or PS constraint. depending on whether the posterior cruciate ligament is preserved (cruciate retaining CR) or sacrificed The method of fixation used to secure the vast (posterior-stabilised, PS) at the time of surgery. majority of knee replacements in place is cement Additional constraint may be necessary to allow the (83.4% from table 3.14), unlike hip replacements.

Terminology note replacements, but these are not reported on here as numbers are too small. Both total and partial knee replacement procedures are discussed in this section. At With regard to the use of the phrase ‘constraint’ present, the NJR does not distinguish between here, for brevity, total knee replacements are medial and lateral unicondylar knee replacements termed unconstrained (instead of posterior cruciate- during the data collection process. In addition, retaining) or posterior-stabilised (instead of posterior there are other possible knee designs, such as cruciate-stabilised). combinations of unicondylar and patello-femoral

3.3.1 Overview of primary knee surgery posterior-stabilised and fixed. A number of primary operations could not be classified according to their Table 3.14 shows the proportion of all kinds of bearing/constraint (approximately 2.5%). primary knee operations performed, broken down by fixation method and bearing type. The vast majority Table 3.15 shows the annual change in usage of of replacements were of the total knee joint with primary knee replacement. Overall, over 80% of all an all cemented implant being the most common primaries utilised an all cemented fixation method fixation method used (83.4% of all primary knee and, since 2003, the share of all implant replacements operations were total replacements using all cemented of this type has increased by some 5%. The main components). Nearly 9% of all primary operations were decline in type of primary surgery has been in the use unicondylar and, of the remaining total knee replaced of all uncemented and hybrid total knee replacements joints, 6.5% were either all uncemented or a hybrid type over time. Each approximately has halved in terms (made up of both cemented and uncemented parts). of proportion of implants of this type now used compared to figures for 2003. There has also been a Over half of all operations (54%) were total knee steady increase in use of all cemented all polyethylene replacements which were all cemented, unconstrained tibial implants (2.1% of primary surgeries in 2012 and fixed, followed by 21% which were all cemented, compared to less than half a percent of all primary surgeries between 2003 and 2005).

160 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.14 Numbers and percentages of primary knee replacements by fixation method and bearing type.

Type of primary knee operation Number of primary knee Fixation method Bearing Type operations Percentage of total Total knee replacement All cemented 490,939 83.4 Cemented and unconstrained, fixed 317,914 54.0 unconstrained, mobile 25,931 4.4 posterior-stabilised, fixed 123,162 20.9 posterior-stabilised, mobile 8,276 1.4 constrained condylar 2,173 0.4 all polyethylene tibia 6,124 1.0 bearing type unknown 7,359 1.3 All uncemented 31,382 5.3 All hybrid 7,307 1.2 Uncemented/hybrid and unconstrained, fixed 18,865 3.2 unconstrained, mobile 16,502 2.8 posterior stabilised, fixed 2,479 0.4

other constraint 350 0.06 © National Joint Registry 2013 bearing type unknown 493 0.08 Unicompartmental knee replacement All unicondylar 51,474 8.7 Unicondylar and fixed 13,763 2.3 mobile 36,982 6.3 bearing type unknown 729 0.1 All patello-femoral 7,881 1.3 Knee type unknown 45 0.01 All types 589,028 100.0

www.njrcentre.org.uk 161 Table 3.15 Percentage of primary knee replacements performed each year by method of fixation and, within each fixation group, by bearing type.

Percentage of knee replacements each year by fixation and bearing type 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Total knee replacement All cemented 81.5 80.8 81.7 81.4 81.9 81.9 82.7 84.0 85.6 86.9 Cemented and unconstrained, fixed 53.3 53.0 53.0 50.7 50.5 51.4 53.1 54.5 56.6 59.8 unconstrained, 4.0 4.2 5.2 6.3 6.3 5.6 4.6 3.9 2.8 2.3 mobile posterior-stabilised, 20.8 20.6 19.8 20.1 20.4 21.0 21.4 21.7 21.6 20.6 fixed posterior-stabilised, 0.9 1.1 1.6 1.9 1.7 1.4 1.4 1.4 1.2 1.1 mobile constrained condylar 0.4 0.5 0.4 0.3 0.3 0.3 0.3 0.4 0.4 0.5 all polyethylene tibia 0.3 0.2 0.4 0.6 0.9 0.9 0.7 1.0 1.7 2.1 bearing type 1.8 1.3 1.4 1.6 1.8 1.4 1.2 1.1 1.2 0.6 unknown All uncemented 6.7 6.6 6.2 6.5 6.5 6.2 5.7 4.7 4.1 3.3 All hybrid 2.8 2.8 2.4 1.7 1.4 1.4 1.2 0.9 0.5 0.4 Uncemented/hybrid and unconstrained, fixed 4.9 4.7 4.3 4.0 4.1 4.0 3.7 2.7 1.9 1.4 unconstrained, mobile 3.5 3.7 3.3 3.2 3.1 3.1 2.7 2.6 2.4 2.1 © National Joint Registry 2013 posterior stabilised, 0.9 0.7 0.6 0.7 0.5 0.4 0.4 0.3 0.2 0.2 fixed other constraint 0.0 0.0 0.2 0.2 0.1 0.1 0.0 0.0 0.0 0.0 bearing type 0.3 0.3 0.2 0.2 0.1 0.1 0.1 0.0 0.0 0.0 unknown Unicompartmental knee replacement All unicondylar 8.0 8.7 8.6 9.2 8.8 9.1 9.0 9.0 8.4 8.1 Unicondylar and fixed 1.4 1.8 2.1 2.3 2.0 2.1 2.2 2.6 2.6 2.9 mobile 6.5 6.7 6.5 6.8 6.7 6.8 6.6 6.2 5.8 5.2 bearing type 0.2 0.2 0.1 0.2 0.1 0.2 0.2 0.1 0.1 0.0 unknown All patello-femoral 1.0 1.0 1.0 1.1 1.4 1.5 1.5 1.4 1.5 1.3 Knee type unknown 0.2 0.01 All types 13,512 27,697 41,846 49,459 66,587 73,806 75,494 78,061 81,032 81,534

162 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

2,722 consultant surgeons carried out at least one type operations over the whole period were excluded (i.e. of primary knee arthroplasty between 2003 and 2012. 475 surgeons and 15 units, respectively). The median number of primary operations performed by a surgeon was 107 (IQR 21 to 315) over the whole Surgeons performing cemented TKR carried out a period. The total number of surgical units in which at median of 137 operations of this type over the whole least one primary knee operation was carried out in the period they were observed with an IQR of 47 to 316 time period was 448. The median number of operations procedures. This means that 25% of surgeons had performed in a unit was 954 (IQR 460 to 1825). a caseload of fewer than 47 cemented total knee replacements over the time and 25% of surgeons Table 3.16 summarises the key features of the carried out more than 316 procedures. The 10% distribution of primary operations carried out by of surgeons with the highest caseload completed surgeons and units in terms of the proportions who between 517 and 2,599 all cemented primary TKRs performed each type of procedure and the median (not shown in table 3.16). The majority of surgeons and IQR of the number of procedures they carried and units carried out very few, if any, uncemented and out. Surgeons and units who performed fewer than 10 hybrid TKRs.

Table 3.16 Distribution of consultant surgeon and unit caseload12 for each fixation type.

Distribution of primary knee operations by fixation type performed by (i) Consultant Surgeons (n=2,247) (ii) Surgical Units (n=433) % performing % performing this fixation this fixation type Median IQR type Median IQR Total knee replacement All cemented 99.7 137 (47-316) 100.0 838 (390-1591) All uncemented 26.6 6 (1-41) 64.0 16 (2-77) All hybrid 29.5 2 (1-5) 71.4 4 (2-14)

Unicompartmental knee replacement © National Joint Registry 2013 All unicondylar 57.2 13 (3-43) 97.7 69 (25-148) All patello-femoral 30.7 5 (2-13) 82.7 11 (4-27)

Table 3.17 shows the age distribution of patients (patello-femoral). The 99th percentile of patient age undergoing a first replacement of their knee joint. for all types of surgery ranged between 85 and 88 The median age of a person receiving a cemented years, indicating that surgery was rarely undertaken in total knee replacement was 70 years (with an inter- a person aged 90 or older, although the maximum age quartile range of 64 to 77 years). However, for of a patient who underwent primary surgery over the unicompartmental primary knee surgery, patients nine year period was aged 102 years. were typically 6 (unicondylar) and 11 years younger

12 Only surgeons or units with at least 10 primary operations recorded in the NJR are presented in the tables. The total count of surgeons who had performed any knee operation between 2003 and 2012 is 2,722. Of these, 475 have performed fewer than 10 operations over this period. Excluding these from reported results leaves 2,247 surgeons. The total count of units who had performed any knee operation between 2003 and 2012 is 448. Of these, 15 have performed fewer than 10 operations over this period. Excluding these from reported results leaves 433 units.

www.njrcentre.org.uk 163 Table 3.17 Age (in years) at primary operation13 for different types of knee replacement; by fixation and bearing type.

Median (IQR)14 Fixation Bearing Type of ages Minimum age15 Maximum age Total knee replacement All cemented 70 (64-77) 14 102 Cemented and unconstrained, fixed 70 (64-77) 16 101 unconstrained, mobile 69 (62-75) 23 97 posterior-stabilised, fixed 71 (64-77) 15 102 posterior-stabilised, mobile 65 (59-73) 22 95 constrained, condylar 72 (63-79) 20 96 all polyethylene tibia 75 (70-79) 25 96 bearing type unknown 70 (63-77) 14 99 All uncemented 69 (62-75) 20 101 All hybrid 69 (62-76) 23 96 Hybrid and unconstrained, fixed 69 (62-76) 24 99 unconstrained, mobile 69 (62-76) 25 101 posterior-stabilised, fixed 66 (59-74) 20 93 © National Joint Registry 2013 other bearing type 65 (57-73) 33 95 bearing type unknown 69 (61-76) 23 91 Unicompartmental knee replacement All unicondylar 64 (57-70) 18 100 Unicondylar and fixed 63 (56-70) 18 100 mobile 64 (58-71) 23 100 bearing type unknown 63 (56-70) 31 90 All patello-femoral 59 (51-68) 1 93 Fixation unknown 69 (59-77) 43 85 All types 70 (63-76) 1 102

13 Ages at primary operation excluding those 19 cases where age was recorded as zero years. Based on a total of 589,009 joints. 14 The interquartile range (IQR) shows the age range of the middle 50% of patients arranged in order of their age at time of primary knee operation. 15 The lowest age excluding 19 cases where age was recorded as zero years. The patello-femoral joint replacement in a one year old in the table above certainly represents a data input error.

164 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.3.2 Revisions after primary but are generally used in younger patients. This may be a function of milder disease in these patients, or knee surgery the desire to delay a total knee replacement for as Table 3.18 shows Kaplan-Meier estimates of the long as possible. Younger patients too may be more cumulative percentage probability of first revision, active which puts more strain on their implants. for any cause, for all cases combined and then Figure 3.6 compares the cumulative hazard of a subdivided by knee types. Estimates are shown, first revision for the main types of primary knee together with 95% Confidence Intervals (95% CIs), replacement surgery over time. Patello-femoral and at 30 and 90 days after the primary operation and at unicompartmental knee replacements were revised each anniversary up to the 9th. much earlier than total knee replacements. All curves These estimates are not adjusted for other factors increase in a reasonably linear fashion as time such as age and gender and these will be looked increases, indicating that the hazard rate was, by and at in the future. The unicondylar and patello-femoral large, constant over time. replacements seem to have done particularly badly

www.njrcentre.org.uk 165 © National Joint Registry 2013 Registry Joint National © 3.60 3.12 4.48 3.75 2.90 2.64 6.70 3.78 4.36 4.28 3.32 4.03 4.47 5.01 3.44 9 years (3.04-4.27) (2.87-6.96) (3.24-4.32) (2.77-3.04) (1.96-3.55) (5.15-8.70) (3.24-4.40) (3.34-5.60) (2.42-7.53) (3.08-3.58) (2.74-7.28) (3.55-7.06) (3.01-3.24) (3.61-4.50) (2.70-4.39) Continued > 3.12 2.88 4.48 3.50 2.71 2.64 6.70 3.31 3.92 4.28 3.06 3.69 4.47 3.84 3.11 8 years (2.79-3.48) (2.87-6.96) (3.10-3.94) (2.60-2.81) (1.96-3.55) (5.15-8.70) (2.99-3.66) (3.31-4.64) (2.42-7.53) (2.89-3.24) (2.74-7.28) (3.28-4.49) (2.80-2.97) (3.37-4.02) (2.62-3.68) 3.02 2.63 3.65 3.38 2.44 2.64 5.25 3.14 3.80 4.28 2.79 3.47 4.47 3.84 2.93 7 years (2.71-3.36) (2.64-5.04) (3.02-3.79) (2.36-2.53) (1.96-3.55) (4.18-6.57) (2.86-3.44) (3.23-4.47) (2.42-7.53) (2.65-2.94) (2.74-7.28) (3.28-4.49) (2.56-2.70) (3.20-3.77) (2.48-3.46) 2.36 3.35 3.13 2.19 2.41 4.65 2.89 3.40 4.28 2.52 3.16 3.95 3.53 2.55 2.66 6 years (2.47-4.54) (2.80-3.49) (2.12-2.26) (1.84-3.15) (3.72-5.80) (2.65-3.16) (2.92-3.96) (2.42-7.53) (2.40-2.64) (2.41-6.45) (3.04-4.09) (2.40-2.95) (2.31-2.42) (2.92-3.41) (2.16-3.00) 3.35 2.11 2.79 1.92 2.29 4.11 2.65 3.27 3.22 2.28 2.86 3.20 3.08 2.32 2.51 5 years (2.47-4.54) (2.50-3.11) (1.86-1.99) (1.76-2.98) (3.28-5.15) (2.43-2.89) (2.81-3.80) (1.74-5.90) (2.18-2.40) (1.90-5.36) (2.66-3.56) (2.27-2.77) (2.06-2.16) (2.64-3.08) (1.96-2.73) 3.04 1.81 2.42 1.66 1.74 3.31 2.35 2.80 3.22 1.89 2.50 2.91 2.73 2.00 2.22 4 years (2.24-4.13) (2.17-2.71) (1.61-1.71) (1.33-2.27) (2.61-4.19) (2.15-2.57) (2.40-3.27) (1.74-5.90) (1.80-1.99) (1.70-4.97) (2.35-3.17) (2.00-2.46) (1.76-1.85) (2.31-2.70) (1.69-2.38) 2.56 1.45 2.01 1.33 1.48 2.63 1.84 2.27 2.88 1.53 2.11 2.41 2.22 1.75 1.96 3 years (1.87-3.51) (1.79-2.26) (1.29-1.38) (1.12-1.95) (2.03-3.40) (1.67-2.03) (1.92-2.68) (1.51-5.47) (1.45-1.61) (1.34-4.32) (1.89-2.60) (1.75-2.18) (1.41-1.49) (1.94-2.29) (1.46-2.10)

1.85 0.96 1.38 0.88 0.93 1.80 1.22 1.56 2.23 1.03 1.50 1.24 1.56 1.31 1.47 2 years (1.30-2.61) (1.21-1.59) (0.84-0.92) (0.68-1.27) (1.32-2.44) (1.09-1.37) (1.29-1.89) (1.07-4.62) (0.97-1.09) (0.56-2.74) (1.29-1.87) (1.30-1.66) (0.93-0.99) (1.36-1.65) (1.06-1.61)

1.02 0.38 0.63 0.33 0.39 0.42 0.51 0.63 0.31 0.43 0.61 0.82 0.68 0.54 0.59 1 year (0.66-1.58) (0.51-0.76) (0.31-0.35) (0.25-0.61) (0.23-0.78) (0.43-0.60) (0.47-0.85) (0.04-2.18) (0.39-0.47) (0.31-2.17) (0.52-0.89) (0.49-0.71) (0.36-0.40) (0.53-0.70) (0.39-0.74) 0 0.47 0.07 0.15 0.06 0.10 0.08 0.13 0.19 0.08 0.12 0.20 0.13 0.07 0.07 90 days (0.25-0.87) (0.10-0.22) (0.05-0.07) (0.05-0.23) (0.02-0.32) (0.09-0.18) (0.11-0.32) (0.07-0.10) (0.03-1.43) (0.07-0.24) (0.04-0.13) (0.07-0.08) (0.08-0.16) (0.03-0.17) Cumulative percentage probability of a first revision (95% CI) at time shown if elapsed since primary operation is: of a first revision probability Cumulative percentage 0 0 0.03 0.14 0.05 0.02 0.04 0.09 0.01 0.04 0.04 0.20 0.07 0.03 0.02 30 days (0.04-0.43) (0.03-0.11) (0.02-0.03) (0.01-0.29) (0.06-0.13) (0.05-0.20) (0.03-0.05) (0.03-1.43) (0.03-0.16) (0.01-0.05) (0.03-0.04) (0.02-0.07) (0.01-0.11) Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) at specified times after primary knee of first revision probability Kaplan-Meier estimates of the cumulative percentage

constrained condylar unconstrained, mobile all polyethylene tibia unconstrained, fixed posterior stabilised, fixed unconstrained, mobile bearing type unknown other constraint posterior-stabilised, posterior-stabilised, fixed bearing type unknown posterior-stabilised, posterior-stabilised, mobile unconstrained, fixed Fixation/bearing type Total knee replacement Total All cemented Cemented and All uncemented All hybrid Uncemented/hybrid and Table 3.18 Table replacement, by fixation and bearing type. replacement,

166 www.njrcentre.org.uk

National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report © National Joint Registry 2013 Registry Joint National © 0 4.12 9.11 11.57 16.11 11.86 10.72 9 years (4.00-4.24) (10.92-12.26) (14.09-18.39) (11.11-12.65) (9.34-12.29) (5.27-15.50) 2.63 3.79 9.74 7.22 10.63 14.75 10.94 8 years (3.70-3.88) (10.13-11.15) (13.15-16.52) (10.35-11.56) (0.37-17.25) (8.89-10.67) (4.52-11.42) 2.63 9.40 3.43 9.28 9.50 6.07 12.88 7 years (9.04-9.99) (4.00-9.17) (9.00-9.81) (3.36-3.50) (11.65-14.23) (0.37-17.25) (8.52-10.11) 0 8.22 3.05 8.37 8.24 5.41 10.81 6 years (7.71-9.07) (7.86-8.64) (3.61-8.09) (7.89-8.56) (2.99-3.12) (9.85-11.86) 0 9.69 7.06 2.70 7.41 7.02 4.14 5 years (6.83-8.02) (6.70-7.35) (2.76-6.18) (6.78-7.34) (2.65-2.76) (8.83-10.62) 0 5.87 2.29 6.22 5.82 3.81 7.43 4 years (5.73-6.76) (5.55-6.11) (2.53-5.71) (5.64-6.12) (6.74-8.19) (2.24-2.34) 0 4.59 1.83 4.70 4.59 3.17 5.50 3 years (4.29-5.14) (4.36-4.84) (2.05-4.89) (4.39-4.80) (1.79-1.87) (4.94-6.12) 0 3.12 1.22 3.01 3.20 1.43 3.22 2 years (2.70-3.34) (3.01-3.40) (0.77-2.65) (2.96-3.29) (1.19-1.25) (2.81-3.68) 0 1.22 0.48 0.89 1.35 0.84 1.09 1 year (0.74-1.07) (1.23-1.47) (0.38-1.86) (1.13-1.32) (0.46-0.49) (0.87-1.36) 0 0.15 0.08 0.10 0.17 0.14 0.05 90 days (0.06-0.17) (0.13-0.22) (0.02-0.97) (0.12-0.19) (0.07-0.09) (0.02-0.14) Cumulative percentage probability of a first revision (95% CI) at time shown if elapsed since primary operation is: of a first revision probability Cumulative percentage 0 0 0 0.05 0.03 0.05 0.05 30 days (0.02-0.11) (0.03-0.08) (0.03-0.07) (0.03-0.04) fixed mobile bearing type unknown Fixation/bearing type Unicompartmental knee replacement Unicompartmental knee replacement All unicondylar All types Unicondylar and All patello-femoral Others/unknown Table 3.18 (continued)

www.njrcentre.org.uk 167 Figure 3.6 Cumulative hazard (x100) of a first revision for different types of primary knee replacement at increasing years after the primary surgery (with 95% CI).

0 1 2 3 4 5 6 7 8 9 Years since primary surgery

Numbers at risk

Cemented 490,939 413,315 337,600 267,439 202,689 142,760 91,079 54,676 25,627 7,913 Uncemented 31,382 28,199 24,441 20,415 15,975 11,425 7,320 4,369 2,142 670 Hybrid 7,307 6,845 6,295 5,471 4,508 3,504 2,595 1,778 896 280 Patello-femoral 7,881 6,679 5,361 4,181 3,071 2,028 1,202 712 335 105 Unicondylar 51,474 44,107 36,376 28,915 21,999 15,428 9,918 5,825 2,762 828

Figure 3.7 compares the cumulative hazard of a first hazard of the need of revision surgery. However, the revision of a prosthesis for different bearing types of curvature of the hazard curves of each type over time cemented TKR. There is no marked separation of indicates that the hazard rate changes over time with the cumulative hazard curves for any of the types of the risk of needing a revision of the implant in earlier bearing shown and the confidence intervals for each years (up to five years after surgery) being higher than type overlap. This strongly suggests that the different later times after surgery. bearing types by and large have similar cumulative

168 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.7 Comparison of the cumulative hazard (x100) of a knee prosthesis first revision for different bearing types at increasing years after the primary surgery when the primary arthroplasty method of fixation was cemented only (with 95% CI).

0 1 2 3 4 5 6 7 8 9 Years since primary surgery Numbers at risk

PS mobile 8,276 7,257 6,147 4,958 3,854 2,812 1,782 942 352 104 C cond 2,173 1,661 1,294 956 730 537 335 214 106 25 PT 6,124 4,356 2,973 2,147 1,588 976 465 211 84 30 UC mobile 25,931 23,743 21,054 17,676 14,007 9,881 5,945 3,151 1,279 378 PS xed 123,162 104,490 85,328 67,223 50,491 35,159 22,327 13,414 6,469 1,986 UC xed 317,914 265,085 215,218 169,856 128,326 90,705 58,653 35,853 16,905 5,226 Unknown 7,359 6,723 5,586 4,623 3,693 2,690 1,572 891 432 164

Key: UC fixed=unconstrained and fixed, UC mobile= unconstrained and mobile, PS fixed=posterior-stabilised and fixed, PS mobile=posterior-stabilised and mobile, C cond=constrained and condylar, PT=polyethylene tibia.

3.3.3 Revisions for different causes after primary knee surgery

Methodological note using patient–time incidence rates; the total number of revisions for that reason divided by the total The previous section looked at revisions for any number of individual patient-years at risk. The figures reason. For any revision, a number of reasons may are given as the numbers of revisions for that reason be related to the first revision of the implant. The per 1,000 years at risk. This method is appropriate if reasons are not mutually exclusive of each other. the hazard rate remains constant. Incidence rates for each reason have been calculated

Table 3.19 shows the revision rates for each reason subdivided by fixation type and whether the primary recorded on the clinical assessment forms for joint procedure was a TKR or an UKR. Table 3.20 shows replacement/revision surgery for all cases and then these first knee revision rates for each reason

www.njrcentre.org.uk 169 by fixation/bearing type. In earlier versions of the The main reasons for revision were pain, aseptic clinical assessment forms recording the type of joint loosening and infection in a primary TKR. Pain and replacement or revision operation to be undertaken aseptic loosening were also primary reasons given for and reasons for this, both ‘stiffness’ and ‘arthritis’ revision of an UKR (and amongst the highest incidence were not originally included as options for indicating rates for revision) alongside implant instability and failure the need for revising the implant. Therefore, for these and dislocation. Progression of arthritis was a key reason reasons for revision, there are fewer years at which for revision of both unicondylar and patello-femoral UKRs. patients are at risk specifically for these categories.

Table 3.19 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first knee revision. Rates shown are for all revised cases and by fixation type.

Patient- Number of revisions per 1,000 patient-years for: years at risk Dislocation/ Aseptic Periprosthetic Implant (x1,000) Pain subluxation Infection loosening Lysis fracture fracture 1.18 0.23 1.10 1.41 0.25 0.14 0.02 All cases* 2173.5 (1.14-1.23) (0.21-0.25) (1.06-1.15) (1.36-1.46) (0.23-0.28) (0.12-0.15) (0.02-0.03) By fixation: 0.78 0.14 1.16 1.05 0.23 0.11 0.02 Cemented 1,800.0 (0.74-0.82) (0.12-0.16) (1.11-1.21) (1.01-1.10) (0.21-0.26) (0.10-0.13) (0.01-0.02) 1.30 0.30 0.90 2.01 0.28 0.15 0.07 Uncemented 130.8 (1.12-1.51) (0.22-0.41) (0.75-1.08) (1.78-2.27) (0.20-0.38) (0.10-0.24) (0.04-0.13) 0.95 0.20 1.23 1.23 0.25 0.08 0.06 © National Joint Registry 2013 Hybrid 35.9 (0.68-1.33) (0.09-0.41) (0.91-1.65) (0.91-1.65) (0.13-0.48) (0.03-0.26) (0.01-0.22) 4.16 0.93 0.76 4.29 0.48 0.33 0.05 Unicondylar 191.8 (3.88-4.46) (0.81-1.08) (0.64-0.89) (4.01-4.59) (0.39-0.59) (0.26-0.43) (0.02-0.09) 6.37 1.01 0.36 2.24 0.07 0.25 0.11 Patello-femoral 27.6 (5.50-7.38) (0.70-1.47) (0.19-0.67) (1.75-2.88) (0.02-0.29) (0.12-0.53) (0.04-0.34)

Table 3.19 (continued)

Patient- Number of revisions per 1,000 patient-years for: Patient- Patient- years Component years years at risk Implant size Other at risk at risk (x1,000) failure mismatch Malalignment indication (x1,000) Stiffness** (x1,000) Arthritis*** 0.27 0.73 0.44 0.64 0.39 0.42 All cases* 2173.5 2000.0 893.1 (0.25-0.29) (0.70-0.77) (0.41-0.47) (0.61-0.68) (0.36-0.42) (0.38-0.46) By fixation: 0.17 0.67 0.37 0.42 0.39 0.16 Cemented 1,800.0 1700.0 744.1 (0.15-0.19) (0.63-0.70) (0.34-0.40) (0.39-0.45) (0.36-0.42) (0.14-0.20) 0.23 0.93 0.58 0.50 0.53 0.22 Uncemented 130.8 119.2 46.3 (0.16-0.33) (0.78-1.11) (0.46-0.73) (0.39-0.63) (0.41-0.68) (0.12-0.40) 0.33 0.84 0.39 0.25 0.36 Hybrid 35.9 30.9 9.7 0 (0.19-0.59) (0.58-1.20) (0.23-0.66) (0.13-0.48) (0.20-0.64) © National Joint Registry 2013 1.06 1.14 0.80 2.37 0.29 2.27 Unicondylar 191.8 178.1 12.7 (0.92-1.21) (1.00-1.30) (0.69-0.94) (2.16-2.60) (0.22-0.38) (1.96-2.63) Patello- 1.73 1.09 1.85 4.05 0.66 4.81 27.6 25.9 80.2 femoral (1.31-2.31) (0.76-1.55) (1.40-2.43) (3.37-4.88) (0.41-1.06) (3.74-6.18)

*Including the one case with bearing type unknown. **This reason was not asked in the earliest version of the clinical assessment form MDSv1 for joint replacement or revision surgery. Therefore, there are fewer patient-years at risk. ***This reason was not asked in the early versions of the clinical assessment forms MDSv1 and MDSv2 for joint replacement/revision surgery and hence, for these reasons, there are fewer patient-years at risk.

170 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.20 Revision rates (95% CI), expressed as number of revisions per 1,000 patient-years, for each recorded reason for first knee revision. Rates shown are for each fixation/bearing surface sub-group.

Patient- Number of revisions per 1,000 patient-years for: years By fixation/ at risk Dislocation/ Aseptic Periprosthetic Implant bearing (x1,000) Pain subluxation Infection loosening Lysis fracture fracture Total knee replacement Cemented unconstrained, 0.76 0.13 1.05 0.93 0.21 0.08 0.02 1,100.0 fixed (0.71-0.81) (0.11-0.15) (0.99-1.11) (0.87-0.99) (0.18-0.24) (0.07-0.10) (0.01-0.03) unconstrained, 1.09 0.24 1.23 1.35 0.35 0.15 0.01 110.2 mobile (0.91-1.30) (0.16-0.35) (1.04-1.46) (0.58-1.43) (0.26-0.48) (0.10-0.25) (0.00-0.06) posterior- 0.68 0.13 1.38 1.24 0.26 0.16 0.01 448.5 stabilised, fixed (0.61-0.76) (0.10-0.17) (1.27-1.49) (1.14-1.34) (0.21-0.31) (0.13-0.20) (0.00-0.02) posterior- 1.30 0.28 1.14 1.18 0.31 0.28 0.09 32.3 stabilised, mobile (0.96-1.76) (0.14-0.54) (0.83-1.58) (0.86-1.62) (0.17-0.57) (0.14-0.54) (0.03-0.29) constrained, 0.58 0.58 3.61 1.45 0.43 0.43 6.9 0 condylar (0.22-1.54) (0.22-1.50) (2.44-5.35) (0.78-2.69) (0.14-1.34) (0.14-1.34) all polyethylene 0.63 0.32 1.39 0.89 0.25 0.13 15.8 0 tibia (0.34-1.18) (0.13-0.76) (0.92-2.11) (0.52-1.49) (0.09-0.67) (0.03-0.51) Bearing type 1.30 0.17 1.46 1.70 0.23 0.13 0.07 30.1 unknown (0.95-1.77) (0.07-0.40) (1.09-1.97) (1.29-2.23) (0.11-0.49) (0.05-0.35) (0.02-0.27) Uncemented/Hybrid unconstrained, 0.98 0.17 1.00 1.79 0.20 0.10 0.05 83.9 fixed (0.79-1.21) (0.10-0.40) (0.81-1.24) (1.52-2.10) (0.13-0.33) (0.05-0.19) (0.02-0.13) unconstrained, 1.37 0.37 0.92 1.85 0.33 0.19 0.09 67.0 mobile (1.12-1.68) (0.25-0.55) (0.72-1.19) (1.55-2.21) (0.22-0.50) (0.11-0.33) (0.04-0.20) posterior- 1.96 0.62 1.16 2.13 0.36 0.18 11.3 0 stabilised, fixed (1.29-2.97) (0.30-1.30) (0.67-1.99) (1.43-3.18) (0.13-0.95) (0.04-0.71) © National Joint Registry 2013 3.26 0.54 other constraint 1.8 0 0 0 0 0 (1.46-7.25) (0.08-3.85) Bearing type 0.74 0.74 3.33 0.74 0.37 2.7 0 0 unknown (0.19-2.96) (0.19-2.96) (1.73-6.40) (0.19-2.96) (0.05-2.6) Unicompartmental knee replacement Unicondylar 4.96 0.13 0.90 4.70 0.43 0.32 0.06 fixed 46.4 (4.35-5.64) (0.06-0.29) (0.67-1.22) (4.11-5.36) (0.28-0.67) (0.19-0.54) (0.02-0.20) 3.91 1.21 0.71 4.20 0.51 0.34 0.04 mobile 142.2 (3.60-4.25) (1.04-1.40) (0.58-0.86) (3.88-4.56) (0.40-0.64) (0.26-0.46) (0.01-0.08) Bearing type 3.77 0.31 0.63 2.20 0.31 3.2 0 0 unknown (2.14-6.63) (0.04-2.23) (0.16-2.51) (1.05-4.61) (0.04-2.20) Patello-femoral 6.37 1.00 0.36 2.24 0.07 0.25 0.11 27.6 (5.50-7.38) (0.70-1.50) (0.19-0.67) (1.75-2.88) (0.02-0.29) (0.12-0.53) (0.04-0.34) Others/unknown 2.92 0.3 0 0 0 0 0 0 (0.41-20.76)

Blue italic estimates based on a small group size (n<30), therefore estimates are unreliable. *This reason was not asked in the earliest version of the clinical assessment form MDSv1 for joint replacement or revision surgery. Therefore, there are fewer patient-years at risk. **This reason was not asked in the early versions of the clinical assessment forms MDSv1 and MDSv2 for joint replacement/revision surgery and hence, for these reasons, there are fewer patient-years at risk.

Continued >

www.njrcentre.org.uk 171 Table 3.20 (continued)

Number of revisions per 1,000 Patient- patient-years for: Patient- Patient- years Component years years By fixation/ at risk Implant size Other at risk at risk bearing (x1,000) failure mismatch Malalignment indication (x1,000) Stiffness* (x1,000) Arthritis** Cemented unconstrained, 0.14 0.62 0.38 0.42 0.38 0.14 1,100.0 1,100.0 477.2 fixed (0.12-0.16) (0.58-0.67) (0.34-0.42) (0.38-0.46) (0.35-0.42) (0.12-0.19) unconstrained, 0.28 1.06 0.48 0.44 0.56 0.10 110.2 103.9 40.8 mobile (0.20-0.40) (0.89-1.27) (0.37-0.63) (0.34-0.59) (0.43-0.72) (0.04-0.26) posterior- 0.19 0.64 0.34 0.35 0.31 0.21 448.5 414.0 191.0 stabilised, fixed (0.16-0.24) (0.57-0.72) (0.29-0.39) (0.30-0.41) (0.26-0.37) (0.15-0.29) posterior- 0.28 1.21 0.25 1.08 0.97 0.16 stabilised, 32.3 30.9 12.4 (0.14-0.54) (0.88-1.65) (0.12-0.49) (0.78-1.51) (0.68-1.39) (0.04-0.64) mobile constrained, 0.43 1.01 0.14 0.29 0.31 6.9 6.4 2.7 0 condylar (0.14-1.34) (0.48-2.12) (0.02-1.03) (0.07-1.16) (0.08-1.25) all polyethylene 0.25 0.57 0.25 0.70 0.39 0.21 15.8 15.4 9.5 tibia (0.09-0.67) (0.30-1.09) (0.09-0.67) (0.39-1.26) (0.17-0.87) (0.05-0.84) bearing type 0.27 0.70 0.37 0.76 0.33 0.38 30.1 27.5 10.6 unknown (0.13-0.53) (0.45-1.07) (0.20-0.66) (0.51-1.15) (0.17-0.63) (0.14-1.00) Uncemented/Hybrid unconstrained, 0.24 0.85 0.44 0.41 0.44 0.07 83.9 75.2 29.6 fixed (0.15-0.37) (0.67-1.07) (0.32-0.61) (0.29-0.57) (0.31-0.62) (0.02-0.27) unconstrained, 0.27 0.85 0.58 0.34 0.43 0.36 67.0 60.9 22.2 mobile (0.17-0.43) (0.66-1.10) (0.43-0.79) (0.23-0.52) (0.29-0.63) (0.18-0.72) posterior- 0.18 1.33 1.16 1.16 0.99 11.3 10.1 3.4 0 © National Joint Registry 2013 stabilised, fixed (0.04-0.71) (0.80-2.21) (0.67-1.99) (0.67-1.99) (0.53-1.84) 0.54 2.17 0.54 1.10 other constraint 3.3 0 1.8 0.3 0 (0.08-3.85) (0.81-5.78) (0.08-3.85) (0.28-4.40) Bearing type 0.37 1.85 0.37 1.11 1.40 2.7 2.1 0.5 0 unknown (0.05-2.63) (0.77-4.45) (0.05-2.63) (0.36-3.44) (0.45-4.33) Unicompartmental knee replacement Unicondylar 0.82 0.93 0.80 1.94 0.41 2.84 fixed 46.4 43.9 21.1 (0.60-1.13) (0.69-1.25) (0.58-1.10) (1.58-2.38) (0.26-0.65) (2.21-3.66) 1.14 1.20 0.82 2.53 0.25 2.08 mobile 142.2 131.3 57.8 (0.98-1.33) (1.03-1.40) (0.68-0.98) (2.28-2.81) (0.18-0.35) (1.74-2.48) Bearing type 0.94 1.26 0.31 1.26 1.56 3.2 3.0 0 1.3 unknown (0.30-2.92) (0.47-3.35) (0.04-2.23) (1.05-4.61) (0.39-6.22) Patello-femoral 1.74 1.09 1.85 4.05 0.66 4.81 27.6 25.9 12.7 (1.31-2.31) (0.76-1.55) (1.40-2.43) (3.37-4.88) (0.41-1.06) (3.74-6.18) Other/unknown 2.92 0.3 0 0 0 0.3 0 - (0.41-21.00) Blue italic estimates based on a small group size (n<30), therefore estimates are unreliable. *This reason was not asked in the earliest version of the clinical assessment form MDSv1 for joint replacement or revision surgery. Therefore, there are fewer patient-years at risk. **This reason was not asked in the early versions of the clinical assessment forms MDSv1 and MDSv2 for joint replacement/revision surgery and hence, for these reasons, there are fewer patient-years at risk.

172 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.3.4 Revisions after primary knee where fewer than 100 primary knee joint replacements remain at risk. No attempt has been made to adjust surgery by main brands for TKR for other factors that may influence the chance of and UKR revision so the figures are unadjusted probabilities.

Tables 3.21 and 3.22 show the Kaplan-Meier Table 3.23 shows Kaplan-Meier estimates of the estimates of the cumulative percentage probability of cumulative percentage probability of first revision of first revision (for any reason) of a primary TKR (table a primary TKR or primary UKR by implant brand and 3.21) and primary UKR (table 3.22) by implant brand. bearing type for those brands/bearing types which We have only included those brands that have been were implanted on at least 1,000 occasions. used in a primary knee procedure in 1,000 or more operations. Figures in blue indicate those time points

www.njrcentre.org.uk 173 Table 3.21 Kaplan-Meier estimated cumulative percentage probability of first revision (95% CI) of a primary total knee replacement by main type of implant brand at the indicated number of years after primary operation.16

Cumulative percentage probability of a first revision (95% CI) at time shown if time Number elapsed since primary operation is of knee Brand17 joints 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Advance 0.29 1.20 1.62 1.98 2.22 2.94 3.17 3.42 3.98 5,258 MP (0.17-0.49) (0.91-1.58) (1.27-2.06) (1.58-2.48) (1.78-2.77) (2.33-3.71) (2.50-4.01) (2.63-4.45) (2.79-5.66) 0.28 0.92 1.38 1.74 1.97 2.30 2.54 2.85 3.12 AGC 53,440 (0.24-0.33) (0.83-1.01) (1.28-1.50) (1.62-1.88) (1.83-2.11) (2.14-2.47) (2.35-2.74) (2.62-3.11) (2.83-3.45) 0.45 1.54 2.07 2.52 2.88 3.29 4.37 4.37 Columbus 4,319 - (0.28-0.72) (1.16-2.04) (1.60-2.66) (1.97-3.23) (2.23-3.73) (2.48-4.37) (2.58-7.35) (2.58-7.35) E-Motion 0.69 1.55 2.54 2.64 2.79 3.55 3.55 3.55 3.55 Bicondylar 2,001 (0.40-1.19) (1.05-2.27) (1.84-3.50) (1.92-3.62) (2.03-3.83) (2.41-5.22) (2.41-5.22) (2.41-5.22) (2.41-5.22) Knee Endoplus 0.70 1.34 1.81 2.18 2.44 2.68 2.85 3.00 3.00 Bicondylar 13,833 (0.57-0.86) (1.16-1.55) (1.59-2.05) (1.94-2.45) (2.17-2.74) (2.38-3.02) (2.53-3.23) (2.58-3.48) (2.58-3.48) Knee 0.42 0.96 1.39 1.84 2.07 2.27 2.34 2.43 2.43 Genesis 2 30,010 (0.35-0.51) (0.84-1.10) (1.23-1.56) (1.64-2.06) (1.85-2.32) (2.01-2.56) (2.07-2.66) (2.11-2.79) (2.11-2.79) Genesis 2 0.55 1.35 2.25 2.69 3.37 3.75 4.50 4.82 4.82 4,832 Oxinium (0.37-0.82) (1.03-1.77) (1.80-2.82) (2.17-3.34) (2.71-4.19) (2.99-4.71) (3.51-5.77) (3.69-6.29) (3.69-6.29) Insall- 0.30 0.90 1.58 2.17 2.69 3.08 3.45 3.79 3.79 2,058 Burstein 2 (0.13-0.66) (0.57-1.43) (1.11-2.24) (1.61-2.93) (2.04-3.53) (2.36-4.00) (2.67-4.46) (2.91-4.93) (2.91-4.93) 0.24 1.03 1.76 2.21 2.67 3.00 3.49 3.91 4.30 Kinemax 10,862 (0.16-0.35) (0.86-1.25) (1.52-2.03) (1.95-2.51) (2.37-3.00) (2.68-3.36) (3.13-3.89) (3.50-4.37) (3.81-4.86) 0.64 1.09 1.75 2.06 2.33 2.39 2.61 2.74 2.94 LCS 2,038 (0.37-1.10) (0.72-1.65) (1.26-2.43) (1.52-2.79) (1.75-3.10) (1.80-3.16) (1.99-3.43) (2.10-3.58) (2.25-3.83) LCS 0.47 1.13 1.70 2.25 2.64 2.88 3.12 3.12 18,235 - Complete (0.38-0.56) (0.98-1.31) (1.50-1.93) (2.01-2.53) (2.35-2.96) (2.56-3.24) (2.73-3.56) (2.73-3.56) 0.30 0.93 1.63 2.02 2.33 2.33 3.02 3.33 5.05 Maxim 2,124 (0.13-0.66) (0.59-1.47) (1.14-2.32) (1.46-2.80) (1.71-3.18) (1.71-3.18) (2.18-4.19) (2.35-4.72) (3.00-8.46)

© National Joint Registry 2013 0.27 0.76 1.32 1.63 1.73 1.83 1.83 3.33 3.33 MRK 5,534 (0.16-0.45) (0.55-1.06) (1.01-1.71) (1.28-2.08) (1.36-2.21) (1.42-2.36) (1.42-2.36) (1.74-6.31) (1.74-6.31) Natural 0.33 0.88 1.32 1.76 2.21 2.66 2.80 2.80 2.80 2,538 Knee II (0.17-0.67) (0.57-1.37) (0.91-1.91) (1.26-2.45) (1.62-3.01) (1.96-3.61) (2.06-3.79) (2.06-3.79) (2.06-3.79) 0.37 0.88 1.38 1.74 2.16 2.47 2.75 3.06 3.26 Nexgen 73,966 (0.33-0.42) (0.81-0.96) (1.28-1.48) (1.63-1.87) (2.03-2.31) (2.31-2.64) (2.56-2.95) (2.82-3.32) (2.96-3.60) 0.73 2.14 2.56 3.26 3.74 4.09 4.62 5.43 5.43 Opetrak 2,232 (0.44-1.20) (1.57-2.91) (1.92-3.41) (2.49-4.27) (2.87-4.87) (3.12-5.35) (3.33-6.39) (3.63-8.10) (3.63-8.10) PFC Sigma 0.36 0.87 1.28 1.55 1.81 1.96 2.14 2.30 2.45 Bicondylar 187,728 (0.34-0.39) (0.82-0.91) (1.22-1.34) (1.49-1.62) (1.73-1.89) (1.88-2.05) (2.04-2.24) (2.19-2.42) (2.30-2.61) Knee 0.36 0.98 1.22 1.52 1.81 2.28 2.36 2.36 2.94 Profix 3,941 (0.22-0.61) (0.71-1.36) (0.91-1.63) (1.17-1.98) (1.41-2.33) (1.78-2.91) (1.84-3.01) (1.84-3.01) (1.90-4.53) 0.21 0.58 1.62 2.01 2.59 3.08 3.80 3.80 3.80 Rotaglide 1,067 (0.05-0.84) (0.24-1.40) (0.92-2.85) (1.18-3.41) (1.55-4.30) (1.81-5.21) (2.15-6.66) (2.15-6.66) (2.15-6.66) 0.57 1.85 2.79 3.40 3.66 4.12 4.62 4.74 5.45 Rotaglide + 2,108 (0.33-1.01) (1.35-2.53) (2.15-3.61) (2.69-4.31) (2.91-4.60) (3.30-5.14) (3.71-5.74) (3.81-5.90) (4.19-7.07) 0.40 1.17 1.71 2.08 2.43 2.70 3.13 3.26 3.53 Scorpio 34,185 (0.34-0.47) (1.06-1.30) (1.57-1.86) (1.92-2.26) (2.25-2.63) (2.49-2.92) (2.87-3.40) (2.98-3.56) (3.13-3.97) 0.44 1.07 1.54 1.84 1.93 2.11 2.98 2.98 Triathlon 29,766 - (0.36-0.53) (0.94-1.22) (1.36-1.74) (1.62-2.10) (1.69-2.21) (1.78-2.50) (1.89-4.69) (1.89-4.69) 0.35 0.95 1.57 2.04 2.36 2.56 3.41 3.41 Vanguard 17,098 - (0.26-0.46) (0.77-1.16) (1.30-1.89) (1.68-2.48) (1.92-2.91) (2.00-3.26) (2.02-5.70) (2.02-5.70)

16 Estimates in blue indicate that fewer than 100 cases remain at risk at the time shown. 17 Brands shown have been used in at least 1,000 primary total knee replacement operations.

174 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.22 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) of a primary unicompartmental knee replacement by main type of implant brand at the indicated number of years after primary operation.18

Cumulative percentage probability of a first revision (95% CI) at time shown if time Number elapsed since primary operation is of knee Brand19 joints 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Patello-femoral 0.76 2.35 4.01 5.58 7.68 8.51 10.94 12.53 13.88 Avon 3,794 (0.52-1.11) (1.89-2.93) (3.37-4.76) (4.79-6.50) (6.69-8.82) (7.42-9.75) (9.49-12.60) (10.72-14.61) (11.61-16.56) 0.89 4.16 6.35 8.74 9.16 9.16 9.16 FPV 1,227 - - (0.48-1.65) (3.06-5.64) (4.88-8.23) (6.80-11.21) (7.09-11.78) (7.09-11.78) (7.09-11.78) Journey PFJ 1.98 4.60 8.20 9.69 13.89 17.52 17.52 1,064 - - Oxinium (1.26-3.08) (3.39-6.23) (6.43-10.43) (7.68-12.19) (11.08-17.34) (12.98-23.41) (12.98-23.41) Unicondylar 2.58 5.18 6.98 8.36 9.07 10.57 11.78 11.78 11.78 AMC/Uniglide 2,279 (1.99-3.35) (4.28-6.26) (5.89-8.26) (7.12-9.81) (7.73-10.63) (8.94-12.48) (9.75-14.19) (9.75-14.19) (9.75-14.19) 0.89 2.53 3.83 4.72 5.71 6.64 7.03 7.23 7.23 MG Unicondylar 2,361 (0.58-1.37) (1.97-3.25) (3.12-4.70) (3.92-5.69) (4.80-6.79) (5.62-7.84) (5.93-8.32) (6.08-8.59) (6.08-8.59) Oxford Partial 1.18 2.91 4.25 5.43 6.61 7.82 9.10 10.52 11.49

35,947 © National Joint Registry 2013 Knee (1.07-1.30) (2.73-3.10) (4.03-4.50) (5.16-5.71) (6.30-6.95) (7.44-8.21) (8.63-9.58) (9.92-11.15) (10.72-12.31) 2.26 4.86 7.53 9.57 11.13 12.55 13.85 15.08 17.31 Preservation 1,509 (1.62-3.15) (3.89-6.08) (6.29-8.99) (8.17-11.20) (9.60-12.89) (10.88-14.46) (12.00-15.97) (13.00-17.46) (14.43-20.68) 1.04 3.17 4.96 6.51 6.51 Sigma HP 3,142 - - - - (0.72-1.50) (2.47-4.06) (3.84-6.41) (4.66-9.05) (4.66-9.05) Zimmer 0.52 2.11 3.36 4.69 5.27 5.27 5.27 5.27 5.27 3,734 Unicompartment (0.32-0.85) (1.62-2.76) (2.67-4.24) (3.74-5.87) (4.15-6.69) (4.15-6.69) (4.15-6.69) (4.15-6.69) (4.15-6.69)

18 Estimates in blue indicate that fewer than 100 cases remain at risk at the time shown. 19 Brands shown have been used in at least 1,000 primary unicompartmental knee replacement operations.

www.njrcentre.org.uk 175 Table 3.23 Kaplan-Meier estimates of the cumulative percentage probability of first revision (95% CI) of a total knee replacement at the indicated number of years after primary operation, by main implant brands and type of fixation and constraint20.

Cumulative percentage probability of a first revision (95% CI) at time shown if time elapsed since primary operation is No. of Brand 21 joints 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Advance MP cemented 0.30 1.22 1.61 1.98 2.23 2.96 3.18 3.45 4.03 unconstrained, 5,189 (0.18-0.50) (0.93-1.60) (1.27-2.06) (1.58-2.49) (1.78-2.78) (2.34-3.73) (2.51-4.04) (2.64-4.49) (2.80-5.76) fixed AGC cemented 0.25 0.85 1.32 1.68 1.90 2.24 2.45 2.78 3.07 unconstrained, 50,160 (0.21-0.30) (0.76-0.94) (1.21-1.43) (1.55-1.82) (1.76-2.05) (2.07-2.41) (2.26-2.66) (2.53-3.04) (2.76-3.41) fixed Uncemented/ hybrid 1.14 2.51 2.91 3.42 3.54 3.99 4.26 4.26 4.26 2,051 unconstrained, (0.76-1.72) (1.90-3.31) (2.25-3.77) (2.66-4.39) (2.76-4.54) (3.09-5.14) (3.25-5.57) (3.25-5.57) (3.25-5.57) fixed Columbus cemented 0.39 1.39 1.85 2.29 2.68 3.14 4.38 4.38 unconstrained, 4,051 - (0.23-0.67) (1.02-1.89) (1.40-2.45) (1.74-3.00) (2.02-3.57) (2.29-4.30) (2.41-7.91) (2.41-7.91) fixed E-Motion Bicondylar Knee Uncemented/ hybrid 0.89 1.50 2.08 2.20 2.37 3.26 3.26 1,416 - - unconstrained, (0.51-1.56) (0.95-2.34) (1.41-3.08) (1.50-3.23) (1.61-3.47) (2.04-5.20) (2.04-5.20) mobile Endoplus Bicondylar Knee cemented 0.75 1.43 1.84 2.21 2.48 2.72 3.05 3.05 3.05 unconstrained, 7,440 (0.58-0.97) (1.18-1.72) (1.56-2.18) (1.89-2.59) (2.12-2.91) (2.31-3.20) (2.55-3.65) (2.55-3.65) (2.55-3.65) fixed cemented 0.55 1.11 1.56 1.91 2.19 2.41 2.47 2.47 2.47 unconstrained, 4,235 (0.37-0.83) (0.83-1.48) (1.22-1.99) (1.53-2.40) (1.76-2.71) (1.95-2.97) (2.00-3.05) (2.00-3.05) (2.00-3.05) mobile Uncemented/ hybrid 0.66 1.14 1.90 2.51 2.65 2.86 2.86 4.23 4.23 © National Joint Registry 2013 1,446 unconstrained, (0.34-1.26) (0.69-1.89) (1.27-2.86) (1.73-3.64) (1.84-3.83) (1.98-4.14) (1.98-4.14) (2.15-8.26) (2.15-8.26) mobile Genesis 2 cemented 0.32 0.81 1.18 1.62 1.85 2.11 2.16 2.27 2.27 uncons-trained, 21,935 (0.25-0.41) (0.69-0.96) (1.02-1.38) (1.40-1.87) (1.60-2.14) (1.81-2.45) (1.85-2.53) (1.90-2.71) (1.90-2.71) fixed cemented 0.79 1.48 2.02 2.54 2.80 2.80 2.80 2.80 2.80 posterior 6,736 (0.59-1.05) (1.18-1.85) (1.64-2.49) (2.08-3.10) (2.27-3.46) (2.27-3.46) (2.27-3.46) (2.27-3.46) (2.27-3.46) stabilised, fixed Genesis 2 Oxinium cemented 0.47 1.03 1.93 2.25 3.04 3.35 3.73 4.08 4.08 uncons- 3,259 (0.28-0.79) (0.71-1.50) (1.44-2.58) (1.70-2.97) (2.31-3.99) (2.53-4.42) (2.80-4.96) (2.99-5.57) (2.99-5.57) trained, fixed Insall-Burstein 2 cemented 0.33 0.66 1.35 2.01 2.43 2.68 3.10 3.48 3.48 posterior 1,869 (0.15-0.72) (0.38-1.16) (0.91-2.01) (1.44-2.78) (1.79-3.29) (2.00-3.60) (2.33-4.13) (2.59-4.67) (2.59-4.67) stabilised, fixed Kinemax Cemented 0.25 1.05 1.78 2.22 2.69 3.03 3.51 3.94 4.28 uncons- 10,656 (0.17-0.36) (0.88-1.27) (1.54-2.05) (1.95-2.53) (2.39-3.03) (2.71-3.40) (3.14-3.92) (3.52-4.40) (3.79-4.83) trained, fixed LCS Uncemented/ 0.74 1.12 1.80 2.11 2.35 2.35 2.43 2.43 2.43 hybrid uncons- 1,356 (0.40-1.38) (0.67-1.85) (1.21-2.67) (1.46-3.04) (1.66-3.32) (1.66-3.32) (1.72-3.42) (1.72-3.42) (1.72-3.42) trained, mobile Continued > 176 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.23 (continued)

Opetrak © National Joint Registry 2013 cemented posterior 0.69 2.27 2.66 3.40 4.07 4.58 4.58 4.58 4.58 1,494 stabilised, (0.37-1.28) (1.59-3.23) (1.90-3.71) (2.48-4.64) (2.99-5.52) (3.35-6.25) (3.35-6.25) (3.35-6.25) (3.35-6.25) fixed PFC Sigma Bicondylar Knee Cemented 0.33 0.79 1.14 1.39 1.61 1.77 1.93 2.12 2.18 unconstrained, 115,473 (0.30-0.37) (0.74-0.85) (1.07-1.21) (1.31-1.48) (1.52-1.70) (1.67-1.87) (1.81-2.05) (1.98-2.27) (2.03-2.35) fixed unconstrained, 0.61 1.26 1.89 2.18 2.47 2.59 2.79 2.79 2.79 6,504 mobile (0.45-0.84) (1.01-1.58) (1.56-2.29) (1.82-2.62) (2.06-2.96) (2.15-3.11) (2.29-3.40) (2.29-3.40) (2.29-3.40) posterior 0.37 0.91 1.39 1.67 1.96 2.10 2.31 2.44 2.79 stabilised, 51,706 (0.32-0.43) (0.83-1.01) (1.28-1.51) (1.54-1.81) (1.81-2.12) (1.94-2.27) (2.12-2.51) (2.22-2.67) (2.43-3.21) fixed posterior 0.70 1.62 2.31 2.90 3.27 3.67 4.00 4.00 7.10 stabilised, 5,346 (0.50-0.98) (1.29-2.04) (1.89-2.81) (2.41-3.50) (2.72-3.94) (3.03-4.44) (3.25-4.92) (3.25-4.92) (3.00-16.28) mobile all 0.33 0.59 1.31 1.43 2.59 2.59 2.59 2.59 2.59 polyethylene 4,148 (0.18-0.60) (0.35-0.98) (0.85-2.00) (0.94-2.19) (1.64-4.08) (1.64-4.08) (1.64-4.08) (1.64-4.08) (1.64-4.08) tibia Constraint 0.26 0.88 1.61 1.89 2.31 2.31 2.57 2.57 2.57 1,618 unknown (0.10-0.68) (0.51-1.51) (1.06-2.44) (1.28-2.80) (1.58-3.37) (1.58-3.37) (1.73-3.79) (1.73-3.79) (1.73-3.79) Uncemented/ hybrid 0.31 0.57 1.07 1.30 1.58 1.58 1.58 1.58 1.58 1,618 uncons- (0.13-0.75) (0.30-1.10) (0.65-1.74) (0.83-2.04) (1.04-2.40) (1.04-2.40) (1.04-2.40) (1.04-2.40) (1.04-2.40) trained, fixed Continued >

www.njrcentre.org.uk 177 Table 3.23 (continued)

© National Joint Registry 2013 stabilised, 1,372 (0.15-0.89) (0.57-1.70) (0.95-2.31) (1.28-2.84) (1.44-3.10) (1.54-3.28) (1.67-3.56) (1.67-3.56) (1.80-12.02) mobile Uncemented/ hybrid 0.59 1.34 1.64 1.78 2.13 2.26 3.29 3.45 3.97 4,712 uncons- (0.40-0.86) (1.04-1.73) (1.30-2.07) (1.42-2.23) (1.71-2.66) (1.81-2.83) (2.52-4.29) (2.64-4.52) (2.81-5.60) trained, fixed Triathlon cemented 0.38 0.94 1.48 1.76 1.83 2.03 2.91 2.91 unconstrained, 22,353 - (0.30-0.48) (0.79-1.11) (1.27-1.72) (1.50-2.05) (1.56-2.15) (1.67-2.47) (1.81-4.66) (1.81-4.66) fixed posterior 0.51 1.24 1.47 1.95 2.10 2.10 stabilised, 6,483 - - - (0.35-0.73) (0.96-1.60) (1.15-1.89) (1.45-2.61) (1.54-2.87) (1.54-2.87) fixed Vanguard cemented 0.33 0.91 1.52 2.04 2.48 2.81 3.82 3.82 unconstrained, 14,185 - (0.24-0.46) (0.72-1.14) (1.23-1.88) (1.63-2.55) (1.94-3.18) (2.05-3.84) (2.17-6.69) (2.17-6.69) fixed posterior 0.53 1.08 1.89 2.49 2.49 2.49 stabilised, 1,932 - - - (0.26-1.06) (0.63-1.84) (1.17-3.06) (1.52-4.07) (1.52-4.07) (1.52-4.07) fixed

20 Estimates in blue indicate that fewer than 100 cases remain at risk at the time shown. 21 Brands shown have been used in at least 1,000 primary knee replacement operations for that type of fixation and bearing type.

178 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.3.5 Mortality after primary both knees replaced on the same day. Patients identified as having a bilateral operation have had the knee surgery second recorded joint excluded from the sample used This section gives the cumulative likelihood of a for mortality analysis. patient dying at different lengths of time after their This identified a sample of 581,253 distinct patients primary operation date by age and gender. Cumulative who had had a primary operation to replace one or probabilities of a knee replacement patient dying in the both knees within the NJR data collection and follow short term (by 30 days or 90 days after the primary up period. Analysis of the cumulative likelihood of a operation) and in the longer term, up to nine years patient dying in the short or medium term after the after their primary operation, are shown. For simplicity primary knee surgery are based on this sample of here, we do not take into account whether the patient patients. In total, 43,155 deaths were linked to this set had further operations (i.e. revision of any primary of patients over the time period. joint) after the primary operation in calculating the cumulative probability of death. Table 3.24 shows, by gender, the age distribution of patients who underwent primary TKR or UKR surgery Of the 589,028 records of a primary operation to over the data collection period. Fewer men than replace a knee joint over the period 1 April 2003 to 31 women, overall, have had a primary knee replacement December 2012, 181 not did have an NHS number and, proportionally, more women than men undergo therefore the death details could not be traced. A surgery above the age of 75. further 22 had missing age (19) or gender (3). Amongst the remainder, 7,572 were bilateral operations, having

Table 3.24 Age and gender distribution of patients undergoing all types of primary knee replacement operations for the period 2003 to 2012.

Gender Number of patients Percentage Female Under 55 23,576 4.1 55-59 29,001 5.0 60-64 47,426 8.2 65-69 57,169 9.8 70-74 63,898 11.0 75-79 59,293 10.2 80 and above 50,660 8.7 Male Under 55 16,887 2.9 55-59 22,126 3.8

60-64 41,468 7.1 © National Joint Registry 2013 65-69 48,537 8.4 70-74 49,842 8.6 75-79 41,114 7.1 80 and above 30,256 5.2 Total 581,253 100.0

Table 3.25 shows the Kaplan-Meier estimated older age groups, had a higher cumulative percentage cumulative percentage probability of a patient dying at probability of dying in the short or longer term after the indicated number of years after surgery stratified their primary knee replacement operation. by age group and gender. Males, particularly in the

www.njrcentre.org.uk 179 4.63 6.68 3.55 9.78 5.29 7.50 24.97 38.26 17.15 61.10 29.08 51.00 22.27 14.58 10.31 9 years (3.80-5.63) (5.90-7.56) (2.88-4.37) (4.60-6.08) (6.73-8.35) (9.61-11.06) (8.99-10.63) (23.87-26.11) (36.90-39.66) (16.30-18.03) (59.24-62.96) (28.03-30.16) (49.55-52.46) (13.74-15.46) (21.95-22.59) 8.56 4.03 5.95 2.79 8.51 4.53 5.92 32.02 13.82 53.30 23.98 41.82 18.51 12.55 20.82 8 years (8.07-9.06) (3.39-4.79) (5.37-6.60) (2.36-3.31) (7.93-9.13) (4.02-5.10) (5.45-6.44) (31.07-32.99) (13.27-14.40) (52.01-54.61) (20.08-21.58 (23.24-24.73) (40.87-42.78) (11.94-13.19) (18.29-18.73) 6.69 3.08 5.20 2.22 6.69 3.64 4.66 25.97 11.18 45.78 18.60 34.40 15.03 17.26 10.33 7 years (6.35-7.05) (2.63-3.60) (4.72-5.73) (1.90-2.59) (6.28-7.13) (3.27-4.05) (4.32-5.02) (9.87-10.82) (25.25-26.71) (10.76-11.62) (44.75-46.83) (18.07-19.15) (33.67-35.15) (16.69-17.84) (14.87-15.19) 5.46 8.67 2.35 3.99 1.87 5.48 2.73 8.31 3.61 21.00 38.16 14.50 27.55 12.00 13.74 6 years (5.18-5.75) (8.35-9.01) (2.03-2.73) (3.63-4.40) (1.61-2.18) (5.16-5.83) (2.46-3.03) (7.94-8.69) (3.35-3.88) (20.42-21.60) (37.29-39.03) (14.09-14.93) (26.96-28.15) (13.29-14.21) (11.87-12.13) 4.18 6.49 1.89 3.05 1.41 4.23 2.00 9.06 6.26 2.71 16.05 29.35 10.86 20.86 10.22 5 years (3.96-4.41) (6.24-6.75) (1.62-2.19) (2.77-3.37) (1.22-1.64) (3.97-4.50) (1.80-2.23) (5.97-6.56) (2.52-2.92) (8.96-9.16) (9.87-10.59) (15.58-16.54) (28.65-30.06) (10.53-11.20) (20.39-21.35) 3.02 4.48 1.36 7.69 2.27 1.02 3.05 1.44 6.54 4.63 2.00 7.48 11.59 21.84 15.10 4 years (2.85-3.20) (4.29-4.68) (1.16-1.59) (7.44-7.96) (2.04-2.53) (0.87-1.19) (2.86-3.26) (1.28-1.61) (4.40-4.87) (1.85-2.16) (7.20-7.77) (6.46-6.62) (11.22-11.98) (21.27-22.43) (14.71-15.49) 2.01 7.75 2.98 1.03 5.08 1.56 0.71 9.99 2.12 0.96 4.37 3.03 1.40 4.96 14.99 3 years (1.88-2.15) (7.46-8.05) (2.84-3.14) (0.87-1.22) (4.88-5.29) (1.39-1.76) (0.60-0.85) (1.97-2.29) (0.84-1.09) (2.86-3.21) (1.28-1.53) (4.75-5.19) (4.31-4.44) (9.69-10.29) (14.53-15.46) 4.63 1.76 0.70 9.30 2.99 0.96 0.39 5.74 1.33 0.60 2.58 1.70 0.76 2.84 1.14 2 years (1.05-1.24) (4.42-4.86) (1.65-1.87) (0.58-0.86) (8.95-9.66) (2.85-3.14) (0.83-1.11) (0.31-0.49) (5.52-5.96) (1.22-1.46) (0.51-0.71) (1.58-1.83) (0.68-0.85) (2.69-3.01) (2.53-2.62) 2.12 0.74 0.28 4.40 1.31 0.41 0.15 2.64 0.53 0.25 1.15 0.75 0.35 1.21 0.46 1 year (1.98-2.27) (0.67-0.81) (0.21-0.38) (4.17-4.65) (1.22-1.41) (0.34-0.51) (0.11-0.21) (2.50-2.79) (0.46-0.61) (0.19-0.31) (0.67-0.83) (0.30-0.41) (1.11-1.31) (0.41-0.52) (1.12-1.17) 0.63 0.20 0.07 1.60 0.37 0.13 0.05 0.92 0.14 0.06 0.36 0.21 0.10 0.34 0.14 90 days (0.56-0.71) (0.17-0.24) (0.04-0.13) (1.47-1.75) (0.33-0.43) (0.09-0.19) (0.03-0.08) (0.84-1.00) (0.11-0.18) (0.03-0.09) (0.17-0.25) (0.08-0.13) (0.29-0.39) (0.11-0.18) (0.34-0.37) Cumulative percentage probability of patient death (95% CI) at time shown if time elapsed since primary knee arthroplasty is: of patient death (95% CI) at time shown if elapsed since primary knee arthroplasty probability Cumulative percentage 0.38 0.11 0.04 0.95 0.20 0.07 0.03 0.48 0.08 0.04 0.20 0.12 0.06 0.18 0.08 30 days (0.32-0.44) (0.08-0.14) (0.02-0.08) (0.85-1.07) (0.17-0.24) (0.04-0.11) (0.01-0.06) (0.42-0.54) (0.06-0.11) (0.02-0.07) (0.09-0.16) (0.04-0.09) (0.14-0.22) (0.06-0.11) (0.19-0.21) No. of 41,114 63,898 16,887 30,256 59,293 22,126 23,576 50,660 41,468 29,001 48,537 47,426 49,842 57,169 581,253 patients Kaplan-Meier estimated cumulative percentage probability (95% CI) of a patient dying at the indicated number probability Kaplan-Meier estimated cumulative percentage

By gender and age group 75-79 years 70-74 years Males <55 years 80+ years 75-79 years 55-59 years Females <55 years 80+ years 60-64 years 55-59 years All cases 65-69 years 60-64 years 70-74 years 65-69 years

Table 3.25 Table of years after a primary knee joint replacement operation (i) by age group and gender (ii) for all patients. operation (i) by age group of years after a primary knee joint replacement © National Joint Registry 2013 Registry Joint National ©

180 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.3.6 Conclusions Patello-femoral joint replacements have a very high failure rate and were typically revised for pain. The data shows some strong trends that merit further However, it should be remembered that patello- discussion. We hope that this will provoke debate and femoral joint replacements are undertaken for different encourage surgeons and manufacturers to re-evaluate reasons than total knee replacements and so a direct their practice in light of the evidence provided. We comparison of revision rates would be erroneous. accept that the data are open to other interpretations Patello-femoral joint replacements may be revised to and we welcome this. Once again, we must stress a total knee replacement because of problems with very strongly that the NJR provides only part of the a different part of the knee and so the reason for picture, that of survivorship, and only survivorship of revision may be unrelated to the original procedure. a short- to medium-term duration; and now for the In addition, there may be reasons related to the first time some Patient Related Outcome Measures aetiology of patello-femoral arthritis that could explain (PROMs). We do not know whether these trends will why replacing the joint, without significantly correcting continue in the longer term. Indeed, one of the lessons the underlying biomechanical cause, may not always that we have learnt is that survivorship is not linear. be a successful strategy. Patello-femoral joints lead Moreover, the data are imperfect and we are reliant to a significant improvement in PROMs, but the on surgeons completing the data accurately and improvement is not as marked as with unicondylar and recording every procedure without exception. Some total knee replacements. data fields continue to be poorly completed making meaningful analysis difficult. Unicondylar knee replacements also have a higher failure rate than total knee replacements. They were Unlike hip arthroplasty, the practice of knee commonly revised for pain and loosening. Again, arthroplasty has not changed significantly over unicondylar knee replacements may be undertaken the past eight years. This year’s analysis shows a for different reasons than total knee replacements continuation of the trends shown previously with total and they may be revised to a total knee replacement knee replacements surviving markedly better than because of disease progression in a non-operated partial knee replacements. compartment which is unrelated to the original procedure. Therefore, comparing revision rates Overall, the data show that short to medium-term with total knee replacements is not straightforward. survivorship is excellent after almost all common The improvement in PROMs after unicondylar knee types of total knee replacement regardless of fixation, replacement is very similar to that achieved by total constraint and bearing type. For the first time we knee replacement. report on PROMs data. This shows that all types of total knee replacement give significant improvement in We report mortality after knee replacement using the majority of patients (see section 3.5). slightly different methodology to last year. Once again we show that knee replacement is associated For bicondylar knee replacements, unconstrained with a slightly lower risk of death compared to hip implants tended to have slightly lower revision rates replacement. We are currently conducting an in-depth than posterior cruciate-stabilised implants while mobile study of post-operative mortality. bearing prostheses tended to have a slightly higher failure rate than fixed bearing prostheses. Thus, the lowest revision rates were associated with a cemented, unconstrained, fixed bearing prosthesis. However, these differences are small and the results in all these groups are acceptable. This also holds true for analysis within brands. This year the brand analysis has been extended to more brands as more have reached the threshold number of cases needed to do this.

www.njrcentre.org.uk 181 Part 3 3.4 Outcomes after primary ankle replacement National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.4.1 Overview of primary There were 1,417 primary operations (see Tables 3.1 and 3.2), including two bilateral operations performed ankle surgery at the same time. This section looks at revision and mortality for Table 3.26 summarizes the procedures by year of all primary ankle operations performed up to 31 primary operation; December 2012.

Table 3.26 Number of primary ankle operations by year.

The median age at primary surgery was 67 years (IQR The maximum number of procedures carried out by 61 to 74 years) and, overall, the ages of those having any consultant was 102. a first replacement of an ankle joint ranged between 24 and 91 years. Men made up 57.2% of all who Similarly the total number of units involved was 175; underwent primary ankle replacement surgery. 37 (21%) of them carried out 10 or more procedures. The maximum number of procedures carried out by Nearly all procedures were of a cemented type any unit was 94. (1,411, 99.6%), only three recorded procedures used a cementless implant and three primary ankle Table 3.27 below shows the distribution of ankle procedures were recorded in the NJR as ‘unsure’ in brands. Mobility was the main brand used (56.5% of terms of the type of implant used. procedures) followed by Zenith (19.6%).

The procedures were carried out by 165 consultants; 44 (27%) of them performed 10 or more procedures.

Table 3.27 Number of primary ankles by ankle brand.

Brand Number (%) Mobility 800 (56.5) Zenith 280 (19.8) Box 91 (6.4) Salto 86 (6.1) Star 69 (4.9) Hintegra 56 (4.0) Rebalance 17 (1.2) Inbone 2 (0.1) Taric 1 (0.1) © National Joint Registry 2013 Not known 15 (1.1) Total 1,417 (100)

www.njrcentre.org.uk 183 3.4.2 Revisions after primary probabilities of revision (based on Kaplan-Meier estimates) at 30 days, 90 days, 1 year and 2 years, ankle surgery respectively, were: 0.07% (95% CI: 0.01%-0.50%), The maximum follow up time available was 4.63 0.14% (95% CI: 0.04%-0.58%), 0.24% (95% CI: years and, up to the end of December 2012, only 0.08%-0.74%) and 1.40% (95% CI: 0.69%-2.83%). nine revisions of the primary ankle implant had been Table 3.28 below lists the stated reasons for the recorded in the NJR. Estimated cumulative percentage nine revisions.

Table 3.28 Reasons for ankle revision (not mutually exclusive).

Reason Number Infection High suspicion (e.g. pus or confined micro) 0 Low suspicion (awaiting micro/histology) 4 Aseptic loosening Tibial component 2 Talar component 0 Lysis Tibia 1* Talus 1* Malalignment 2 Implant fracture Tibial component 0 Talar component 1

© National Joint Registry 2013 Meniscal component 0 Wear of polyethylene component 0 Meniscal insert dislocation 0 Component migration/Dissociation 1 Pain (undiagnosed) 1 Stiffness 0 Soft tissue impingement 1 Other indication for revision 2

*Same patient had both of these reasons.

3.4.3 Mortality after primary Kaplan-Meier) at 30 days, 90 days, 1 year and 2 years were, respectively: 0.0%, 0.15% (95% CI: 0.04%- ankle surgery 0.59%), 1.05% (95% CI: 0.59%-1.85%) and 1.49% There were 17 deaths in total. The estimated (95% CI: 0.88%-2.51%) There were too few deaths for cumulative percentage probability of death (using further breakdown by age and gender.

184 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Annexe to survival analysis failure rates are less than 10%, the numerical value of e (-H(t)) is numerically close to the value of 1-H(t) and In a departure from our previous reports, in the tables therefore F(t) is close to H(t). The values of F(t) and H(t) above, we have used Kaplan-Meier estimates for do not remain numerically similar once failure rates the cumulative probability of failure (i.e. where failure exceed 10% and then, in general, H(t) will be greater is either a revision or death) at given times t, F(t). than F(t). This has become particularly noticeable in Previously we had reported Nelson-Aalen estimates of the NJR data set for mortality where, after nine years, the ‘cumulative hazard’ (also called ‘integrated hazard’). the estimate of H(t) for 80+ year old men now has exceeded 100 (expressed as number of deaths per This change in reported values brings us more in 100), whereas, because F(t) is a probability, its value line with the form of reporting undertaken by other cannot exceed 100%. National Registries and, therefore, makes for easier comparison with published work from these and other It is useful to use cumulative hazards for plotting, e.g. research bodies. to compare revision rates for various sub-groups. As with the cumulative probability of failure, groups The cumulative hazard at time t after primary with greater cumulative hazards will have the highest operation, say H(t), is not the same as the cumulative revision rates. The shape of each cumulative hazard probability of failure, F(t). To understand H(t), first start curve, however, carries useful information about how with the concept of the ‘hazard rate’ (analogous to the revisions occur because the gradient of the curve the ‘force of mortality’ used in actuarial science). In at any time t estimates the hazard rate. If the plot of the case of revision, the hazard rate at time t is the the cumulative hazard against time is a straight line, (instantaneous) rate at which revisions are performed then revisions will have occurred at a constant rate amongst those cases that have not been revised throughout, i.e. the revision rate is independent from previously. For discrete observations over time, the time of operation. If the graph increases exhibiting a cumulative hazard at time t is the cumulative sum concave shape, then the risk of revision will be lower of the hazard rates; in continuous time it is the area initially, increasing with time after operation. The under (i.e. the integral of) the hazard rate curve. The converse is true; if the shape is convex, revision rate cumulative hazard here is estimated from the data would be higher initially. using Nelson-Aalen estimates to account for the discrete time nature of observations. For the above reasons we give Kaplan-Meier estimates in the tables but have retained estimates of Although the ‘cumulative hazard’ (H(t)) and ‘cumulative ‘cumulative hazards’ for plotting. probability of failure’ (F(t)) are estimated using different approaches, they are mathematically related through the expression F(t)=(1 - e (-H(t))) and so it is possible to estimate one using an estimate of the other. Where

www.njrcentre.org.uk 185 Part 3 3.5 PROMs outcomes National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.5.1 Background to Patient Reported complete but it was not possible to choose the ‘higher quality’ record from the remaining 90 multiple entries Outcome Measures (PROMs) and so these were omitted. This left 324,101 PROMs PROMs assess the patient’s experience and entries, some of which related to primary and some to perspective of the quality of care and treatment they revision procedures. receive when they undergo a NHS-funded hip and/or knee replacement. PROMs are collected via pre- and HES to NJR matching: post-operative questionnaires; the items allow the Within the HES data set, seven rankings were available patient to self report on their state of health and general resulting from linkage methods designed to determine lifestyle before and after their joint replacement surgery. the most likely NJR operation identifier for a given These outcomes provide a means of determining the patient/episode in HES. The linkage method utilises impact the procedure has subsequently had on their and ranks the availability of various combinations of quality of life. PROMs have been routinely collected patient information associated with a particular HES by all providers of NHS-funded care in England episode. Relevant patient related information included since April 2009. The initial questionnaire, Q1, is patient NHS number, date of episode/entry, provider normally administered just prior to the operation and code, local patient ID, date or year of birth and patient questionnaire Q2 approximately six months following gender. In addition, an episode in HES must have the operation. Additional information is available from: an ‘orthopaedically relevant’ OPCS procedure code http://www.hscic.gov.uk/PROMs associated with the particular episode i.e. OPCS codes must identify an orthopaedic type of event. In this section we have concentrated on the variables The linkage rankings were used to assign the most that described health benefits following primary hip or probable NJR operation number to the HES person ID knee replacement. (HESID) and episode (EPIKEY).

3.5.2 Data linkage from PROMs to Not all of such ‘orthopaedically relevant’ HES episodes HES to NJR had entries in the NJR. However, of the 517, 220 HES entries between 2003 and 2012 we were able to Our base PROMs data file had 445,134 entries assign 391,240 (75.6%) to an NJR procedure. (207,436 hips and 230,429 knees). PROMs/HES to NJR/HES matching: Q1 and Q2 had already been linked together when Finally, we were able to link 274,729 of the above received by the NJR. 324,101 PROMs entries to HES patient episodes Linkage between PROMs and the NJR data set could with an assigned NJR operation identifier. There were only be made via Hospital Episode Statistics (HES) data some duplications i.e. instances where more than one spanning the same time period as the PROMs data set. PROMs entry could be ascribed to the same NJR operation. At this juncture we linked the PROMs data PROMs to HES matching: separately to our linked data files for primary hips and knees, after first removing bilateral operations from A total of 3% of the PROMs entries had no HES patient the latter. Any remaining duplications were removed identifier (HESID) and a further 24% of the remainder pragmatically, by checking gender match, NJR patient could not be matched to a particular episode in HES identifiers and finally by choosing the PROMs entry (via EPIKEY). Of the remaining 326,876 PROMs entries, where the date of completion of Q1 was closest to the there were 2,653 duplicate matches and 16 triplicate date of the operation. matches to HES, meaning that multiple PROMs entries existed for the same HES patient episode. For these The remainder of this section provides some simple multiple matches, we chose the ‘best match’, i.e. the summary statistics for the PROMs entries that could PROMs entry that had the best episode matched rank. be matched to our unilateral linked primary hip and In 234 cases there were two equal best matches; we knee files. gave priority then to the 144 PROMs entries that were

www.njrcentre.org.uk 187 Methodological note The EQ-5D Health Scale is a visual analogue scale (VAS), asking about how the patients rate themselves The three health measures in PROMs are the EQ-5D ‘on the day’ (0=worst, 100=best). Index, the EQ-5D Health Scale and the Oxford Hip/ Knee Scores. The Oxford Hip/Knee Scores are based on responses to 12 hip or knee specific questions The EQ-5D Index22 is derived from a profile of about health ‘in the last four weeks’. Each question responses to five questions about health ‘today’, is scored from 0 to 4 and the scores are summed to covering activity, anxiety/depression, discomfort, give a score from 0 (worst) to 48 (best). If fewer than mobility and self care. Weights had been applied to three responses to individual questions are missing, the responses to these questions to calculate the they can be replaced by the mean of the remaining ‘index’. All five questions had to be answered in questions. order to do this. The higher the index the better the patient, with one being the best possible score. The The VAS was not normally distributed, neither distribution of the EQ-5D Index is such that parametric were the Oxford Hip and Knee scores at Q2; non- statistical methods are not suitable for statistical parametric methods of comparison, therefore, were comparison. Non-parametric unpaired and paired used for both these variables. comparisons (two-tailed Mann-Whitney and Wilcoxon matched-pairs signed-ranks) have been used here Summary statistics used throughout this section are but, because of the ‘ceiling effect’ (scores cannot the median and Inter-Quartile Ranges (IQRs). The latter improve beyond 1), paired comparisons were double- is less sensitive to outliers than a simple range. (Note checked with simple sign tests. however that, for paired data, the median difference may not be equal to the difference in medians.)

3.5.3 PROMs outcomes for primary hip replacements

A total of 124,136 of the linked primary hip operations in NJR had an associated PROMs entry; Q1 was complete in 99.7% of these and both Q1 and Q2 were complete in 75.6%. The median (IQR) interval of time from completion of Q1 to the primary operation was 0.59 months (IQR 0.23-1.22; n=84,078) and from primary operation to completion of Q2 was 6.47 months (IQR 6.31 to 6.87; n=83,808).

Table 3.29 opposite shows the overall outcomes after primary hip surgery.

22 As the index is bounded between 0 and 1, we present results in this section with respect to the EQ-5D Index to three decimal places for greater precision.

188 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.29 Overall outcomes after primary hip surgery.

Variable Frequency (%) Q2 Satisfaction: Satisfaction after the operation (Describe the results of the operation) 1 Excellent 35,313 (38.5%) 2 Very good 32,147 (35.0%) 3 Good 16,826 (18.3%) 4 Fair 5,739 (6.3%) 5 Poor 1,735 (1.9%) Q2 Success: Problems now compared to before operation 1 Much better 78,617 (85.6%) 2 A little better 8,715 (9.5%) 3 About the same 2,298 (2.5%) 4 A little worse 1,332 (1.5%) 5 Much worse 908 (1.0%) Q2 Wound: Wound problems since operation 1 Yes 8,498 (10.0%) 2 No 76,070 (90.0%) Q2 Bleeding: Bleeding since operation 1 Yes 4,797 (5.8%) 2 No 78,165 (94.2%) Q2 Urine: Urinary problems after the operation © National Joint Registry 2013 1 Yes 11,260 (13.3%) 2 No 73,302 (86.7%) Q2 Allergy: Experienced allergy with respect to drugs after the operation 1 Yes 10,022 (11.7%) 2 No 75,813 (88.3%) Q2 Re-admitted: Re-admitted to hospital since operation 1 Yes 6,726 (7.3%) 2 No 85,120 (92.7%) Q2 Further Surgery: Has had another operation on affected site 1 Yes 1,924 (2.1%) 2 No 89,998 (97.9%)

Bias in completion of PROMs is difficult to assess complete measures at Q2 tended to have higher or because we do not have details of how many lower scores at Q1; we found that the Q2 completers PROMs questionnaires were sent out and how many tended to have had better scores at Q1, as seen in were returned. Table 3.30. It may be that some of the initially less healthy group had died in the interim; this will be Amongst those NJR cases for whom we had PROMs explored later. entries, we checked whether those who had available/

www.njrcentre.org.uk 189 Table 3.30 Bias in Q2 completion of EQ-5D Index, EQ-5D Health Scale (VAS), Oxford Hip Score.

Median (IQR) Measure Was measure available and complete at Q2? of the score at Q1 EQ-5D Index Yes 83,202 0.516 (0.055 to 0.656) No 33,607 0.159 (-0.016 to 0.620) P<0.001* EQ-5D Health Scale (VAS) Yes 80,394 70 (50-80) No 31,513 65 (50-80) P<0.001* Oxford Hip Score © National Joint Registry 2013 Yes 92,133 18 (12-24) No 30,872 15 (10-22) P<0.001*

*Mann-Whitney U-test

Completion rates for both EQ-5D Index and EQ-5D Histograms showing comparative distributions of Health Scale (VAS) tended to be slightly higher EQ-5D Scale (VAS) and Oxford Hip Scores at Q1 in men than in women at both Q1 and Q2. The and Q2 for complete pairs are shown in Figures 3.8 differences, although ‘statistically significant’, were and 3.9 respectively. in fact very small (e.g. for VAS at Q1: completion rates were 91.5% versus 89.3%, respectively, for All hip sub-groups showed significant improvement in men and women). EQ-5D Index (Table 3.31).

Significant improvements from Q1 to Q2 were seen All the main sub-groups showed improvements in the in all three health measures overall (see Tables 3.31 VAS (Table 3.32) and in the Oxford Hip Score to 3.33). (Table 3.33).

Table 3.31 Changes in EQ-5D Index for hip primaries with index scores at both time points.

Q1 Before Q2 After Change (Q2 operation operation minus Q1) EQ-5D Index Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 0.516 0.796 0.380 All cases 83,202 P<0.001 (0.055-0.656) (0.691-1) (0.175-0.694) 0.364 0.796 0.413 All cemented 27,329 P<0.001 (0.030-0.620) (0.639-1) (0.162-0.694) Cemented by bearing surface: 0.293 0.779 0.413 MoP 23,482 P<0.001 (0.030-0.620) (0.623-1) (0.159-0.694) 0.516 0.805 0.309 MoM 74 P<0.001 (0.055-0.620) (0.639-1) (0.159-0.633) 0.516 0.796 0.380 © National Joint Registry 2013 CoP 2,884 P<0.001 (0.055-0.656) (0.691-1) (0.177-0.697) 0.516 0.848 0.380 Others/unsure 889 P<0.001 (0.055-0.691) (0.691-1) (0.192-0.684)

Continued >

*Wilcoxon matched-pairs signed-ranks test and checked with sign test. **Sign test.

190 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.31 (continued)

Q1 Before Q2 After Change (Q2 operation operation minus Q1) EQ-5D Index Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 0.516 0.814 0.380 All uncemented 38,693 P<0.001 (0.055-0.691) (0.691-1) (0.176-0.691) Uncemented by bearing surface: 0.516 0.814 0.380 MoP 15,054 P<0.001 (0.055-0.691) (0.691-1) (0.176-0.694) 0.516 0.815 0.344 MoM 2,049 P<0.001 (0.055-0.691) (0.691-1) (0.165-0.661) 0.516 0.796 0.380 CoP 5,328 P<0.001 (0.055-0.689) (0.689-1) (0.181-0.694) 0.516 0.848 0.380 CoC 15,175 P<0.001 (0.055-0.691) (0.691-1) (0.192-0.672) 0.587 0.796 0.309 CoM 514 P<0.001 (0.088-0.691) (0.691-1) (0.123-0.535) 0.516 0.796 0.377 Others/unsure 573 P<0.001 (0.055-0.620) (0.620-1) (0.105-0.694) 0.516 0.814 0.413 All hybrid 13,014 P<0.001 (0.055-0.656) (0.691-1) (0.192-0.697) Hybrids by bearing surface: 0.516 0.812 0.413 MoP 8,829 P<0.001 (0.055-0.620) (0.691-1) (0.192-0.705) 0.516 0.850 0.413 MoM 88 P<0.001 (0.025-0.620) (0.701-1) (0.235-0.649) © National Joint Registry 2013 0.516 0.848 0.396 CoP 1,602 P<0.001 (0.055-0.656) (0.691-1) (0.204-0.695) 0.516 0.850 0.380 CoC 2,291 P<0.001 (0.055-0.689) (0.691-1) (0.192-0.697) 0.516 0.796 0.388 Others/unsure 204 P<0.001 (0.055-0.673) (0.691-1) (0.198-0.667) 0.260 0.779 0.413 All reverse hybrid 2,479 P<0.001 (-0.003 to 0.620) (0.620-1) (0.159-0.707) Reverse hybrids by bearing surface: 0.260 0.779 0.413 MoP 1,636 P<0.001 (-0.003 to 0.620) (0.620-1) (0.140-0.708) 0.260 0.796 0.413 CoP 831 P<0.001 (-0.003 to 0.620) (0.620-1) (0.173-0.703) 0.209 0.760 0.404 P=0.003 Others/unsure 12 (0.026-0.568) (0.587-0.908) (0.202-0.613) (P=0.002**) All resurfacing 0.587 0.309 MoM 1,687 1 (0.743-1) P<0.001 (0.189-0.691) (0.165-0.484)

*Wilcoxon matched-pairs signed-ranks test and checked with sign test. **Sign test.

www.njrcentre.org.uk 191 Table 3.32 Changes in EQ-5D Health Scale Score (VAS) for hip primaries with scores at both time points.

Q1 Before Q2 After Change (Q2 EQ-5D Health Scale operation operation minus Q1) Score (VAS) Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 70 80 9 All cases* 80,394 (50-80) (69-90) (-2 to 20) P<0.001 70 80 6 All cemented 26,339 (50-80) (65-90) (-5 to 20) P<0.001 Cemented by bearing surface: 70 80 6 MoP 22,556 (50-80) (65-90) (-5 to 20) P<0.001 72.5 85 0.5 MoM 74 (50-80) (60-90) (-4 to 19) P=0.009 70 80 10 CoP 2,823 (50-80) (70-90) (0 to 25) P<0.001 72.5 80 5 Others/unsure 886 (59-85) (70-90) (-3 to 20) P<0.001 70 80 10 All uncemented 37,451 (50-80) (70-90) (-1 to 22) P<0.001 Uncemented by bearing surface: 70 80 6 MoP 14,385 (50-80) (70-90) (-4 to 20) P<0.001 70 80 9 MoM 1,958 (51-80) (70-90) (-2 to 20) P<0.001 70 80 10 CoP 5,170 (50-80) (70-90) (-1 to 25) P<0.001 70 80 10 CoC 14,879 (50-80) (70-90) (0 to 25) P<0.001 70 80 10 CoM 505 (50-83) (70-90) (-2 to 20) P<0.001 70 80 10 P<0.001 Others/unsure 554 (50-80) (65-90) (-1 to 21) © National Joint Registry 2013 70 80 10 All hybrid 12,564 (50-80) (70-90) (-1 to 22) P<0.001 Hybrids by bearing surface: 70 80 8 MoP 8,450 (50-80) (68-90) (-3 to 20) P<0.001 70 80 10 MoM 83 (50-80) (70-90) (0 to 25) P<0.001 70 80 10 CoP 1,567 (50-80) (70-90) (0 to 25) P<0.001 70 80 10 CoC 2,254 (50-80) (70-90) (0 to 25) P<0.001 70 80 5 Others/unsure 210 (50-80) (65-90) (0 to 20) P<0.001 70 80 9 All reverse hybrid 2,378 (50-80) (64-90) (-3 to 21) P<0.001 Reverse hybrids by bearing surface: 70 80 8 MoP 1,561 (50-80) (65-90) (-5 to 20) P<0.001 70 80 10 CoP 806 (50-80) (60-90) (0 to 25) P<0.001 60 55 0 Others/unsure 11 (33-70) (40-90) (-18 to 20) P=0.65 All resurfacing 70 85 10 MoM 1,662 (57-85) (75-90) (0-20) P<0.001

*Wilcoxon matched-pairs signed-ranks test.

192 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Table 3.33 Changes in Oxford Hip Score for hip primaries with scores at both time points.

Q1 Before Q2 After Change (Q2 operation operation minus Q1) Oxford Hip Score Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 18 41 21 All cases 92,133 (12-24) (34-46) (14-28) P<0.001 17 40 21 All cemented 30,551 (11-23) (32-45) (14-27) P<0.001 Cemented by bearing surface: 17 39 21 MoP 26,294 (11-23) (32-44) (14-27) P<0.001 18 41 21 MoM 83 (11-22) (30-46) (12-28) P<0.001 17 42 21 CoP 3,189 (12-24) (34-46) (14-28) P<0.001 19 42 21 Others/unsure 985 (13-24) (36-46) (15-27) P<0.001 18 42 21 All uncemented 42,497 (12-24) (35-46) (14-28) P<0.001 Uncemented by bearing surface: 18 42 21 MoP 16,704 (12-24) (34-46) (14-28) P<0.001 19 43 21 MoM 2,240 (13-25) (35-46) (14-27) P<0.001 18 42 21 CoP 5,805 (12-23) (34-46) (14-28) P<0.001 18 43 22 CoC 16,547 (13-24) (36-47) (15-28) P<0.001 20 43 20 CoM 565 (14-26) (35-46) (13-27) P<0.001 18 42 21 Others/unsure 636 (13-23) (33-46) (14-28) P<0.001 18 42 22 © National Joint Registry 2013 All hybrid 14,493 (12-24) (35-46) (15-28) P<0.001 Hybrids by bearing surface: 17 41 21 MoP 9,875 (12-24) (34-46) (14-28) P<0.001 18 43 21 MoM 97 (13-23) (37-46) (16-27) P<0.001 18 43 22 CoP 1,770 (12-24) (36-46) (15-28) P<0.001 18 43 23 CoC 2,516 (12-24) (37-47) (15-29) P<0.001 18 41 21 Others/unsure 235 (13-23) (34-46) (14-28) P<0.001 16 40 20 All reverse hybrid 2,755 (11-23) (32-45) (13-28) P<0.001 Reverse hybrids by bearing surface: 16 40 21 MoP 1,825 (11-23) (32-45) (13-27) P<0.001 16 40 20 CoP 918 (11-23) (32-45) (14-28) P<0.001 11.5 38 16.5 Others/unsure 12 (10-16.5) (21.5-44) (10-28.5) P=0.002 All resurfacing 22 44 19 MoM 1,837 (16-28) (38-47) (12-25) P<0.001

*Wilcoxon matched-pairs signed-ranks test.

www.njrcentre.org.uk 193 Figure 3.8 Histogram to compare the distributions of the EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=80,394). (i) At Q1 15,000

0 20 40 60 80 100

Q1 EQ-5D HEALTH SCALE

Figure 3.8 Histogram to compare the distributions of the EQ-5D Health Scale Score (VAS) between Q1 and Q2 in

cases with scores at both time points (n=80,394). (ii) At Q2 15,000

0 0 20 40 60 80 100

Q2 EQ-5D HEALTH SCALE

194 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.9 Histogram to compare the distributions of the Oxford Hip Score between Q1 and Q2 in cases with scores at both time points (n=92,133). (i) At Q1 20,000

15,000

5,000

0 0 10 20 30 40 50

Q1 Oxford Hip Score

Figure 3.9 Histogram to compare the distributions of the Oxford Hip Score between Q1 and Q2 in cases with scores at both time points (n=92,133). (ii) At Q2 20,000

15,000

5,000

0 0 10 20 30 40 50

Q2 Oxford Hip Score

www.njrcentre.org.uk 195 3.5.4 PROMs outcomes for primary completion of Q1 to the primary operation was 0.56 months (IQR 0.20-1.22; n=86,908) and from primary knee replacements operation to completion of Q2 was 6.47 months (IQR 132,019 of the linked primary knee operations in NJR 6.31 to 6.90; n=87,171). had an associated PROMs entry; Q1 was complete in Table 3.34 below shows the overall outcomes after 99.7% of these and both Q1 and Q2 were complete primary hip surgery. in 74.6%. The median (IQR) interval of time from

Table 3.34 Overall outcomes after primary knee surgery.

Variable Frequency (%) Q2 Satisfaction: Satisfaction after the operation (Describe the results of the operation) 1 Excellent 21,962 (22.7%) 2 Very good 33,626 (34.8%) 3 Good 25,185 (26.1%) 4 Fair 12,186 (12.6%) 5 Poor 3,668 (3.8%) Q2 Success: Problems now compared to before operation 1 Much better 68,540 (70.8%) 2 A little better 17,127 (17.7%) 3 About the same 4,901 (5.1%) 4 A little worse 3,727 (3.8%) 5 Much worse 2,577 (2.7%) Q2 Wound: Wound problems since operation 1 Yes 11,825 (13.4%) 2 No 76,265 (86.6%) Q2 Bleeding: Bleeding since operation

© National Joint Registry 2013 1 Yes 6,768 (7.9%) 2 No 78,915 (92.1%) Q2 Urine: Urinary problems after the operation 1 Yes 10,635 (12.2%) 2 No 76,231 (87.8%) Q2 Allergy: Experienced allergy with respect to drugs after the operation 1 Yes 12,244 (13.7%) 2 No 76,938 (86.3%) Q2 Re-admitted: Re-admitted to hospital since operation 1 Yes 9,295 (9.6%) 2 No 87,467 (90.4%) Q2 Further Surgery: Has had another operation on affected site 1 Yes 3,236 (3.3%) 2 No 93,646 (96.7%)

196 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Amongst these NJR cases for whom we have PROMs Table 3.35 below. As for hips, it may be that some of entries, those who had available completed measures those with poorer scores at Q1 had died before the at Q2 tended to have had better scores at Q1, see Q2 assessment.

Table 3.35 Bias in Q2 completion of EQ-5D Index, EQ-5D Health Scale (VAS), Oxford Knee Score.

Median (IQR) Measure Was measure available and complete at Q2? of the score at Q1 EQ-5D Index Yes 87,306 0.587 (0.088-0.691) No 36,451 0.264 (0.055-0.691) P<0.001* EQ-5D Health Scale (VAS) Yes 84,031 70 (55-80) No 33,999 70 (50-80) P<0.001* Oxford Knee Score

*Mann-Whitney U-test

Completion rates for all three health measures, at All knee sub-groups showed significant improvement both Q1 and Q2, tended to be slightly higher in men in EQ-5D Index (Table 3.36). than women; the differences, although ‘statistically significant’ ,were very small (e.g. 99.0% vs 98.9% All sub-groups except patello-femoral showed small respectively at Q1 and 73.5% vs 72.2% at Q2 for the improvements in VAS (Table 3.37). Oxford Knee Score). All sub-groups showed improvements in the Oxford Significant improvements from Q1 to Q2 were seen in Knee score (Table 3.38). all three health measures (see Tables 3.36 to 3.38).

The distributions of EQ-5D Scale (VAS) and Oxford Knee Scores at Q1 and Q2 for complete pairs are shown in Figures 3.10 and 3.11.

www.njrcentre.org.uk 197 Table 3.36 Changes in EQ-5D Index for knee primaries with index scores at both time points.

Q1 Before Q2 After Change (Q2 operation operation minus Q1) EQ-5D Index Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 0.587 0.727 0.275 All cases 87,306 (0.088-0.691) (0.620-0.883) (0.069-0.568) P<0.001 0.587 0.727 0.280 All cemented 81,358 (0.088-0.691) (0.620-0.883) (0.069-0.568) P<0.001 0.587 0.727 0.240 All uncemented 3,782 (0.101-0.691) (0.620-0.848) (0.036-0.532) P<0.001 0.587 0.725 0.240 All hybrid 636 (0.101-0.691) (0.587-0.849) (0.036-0.532) P<0.001 All patello-femoral 257 0.620 0.691 0.105 P<0.001 © National Joint Registry 2013 (0.124-0.691) (0.516-0.796) (0.000-0.311) 0.620 0.760 0.240 All unicondylar 1,273 (0.159-0.691) (0.620-1.000) (0.036-0.532) P<0.001

*Wilcoxon matched-pairs signed-ranks test and checked with sign test.

Table 3.37 Changes in EQ-5D Health Scale score (VAS) for knee primaries with scores at both time points.

Q1 Before Q2 After Change (Q2 EQ-5D Health Scale operation operation minus Q1) Score (VAS) Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 70 75 3 All cases 84,031 (55-80) (60-85) (-7 to 15) P<0.001 70 75 3 All cemented 78,239 (55-80) (60-85) (-7 to 15) P<0.001 70 76 4 All uncemented 3,681 (55-82) (60-86) (-5 to 15) P<0.001 70 75 2 All hybrid 609 (60-85) (60-85) (-10 to 15) P<0.001 70 70 0 © National Joint Registry 2013 All patello-femoral 259 (50-80) (50-85) (-10 to 15) P=0.16 75 80 3 All unicondylar 1,243 (60-85) (65-89) (-9 to 15) P<0.001

*Wilcoxon matched-pairs signed-ranks test and checked with sign test.

Table 3.38 Changes in Oxford Knee Score for knee primaries with scores at both time points.

Q1 Before Q2 After Change (Q2 operation operation minus Q1) Oxford Knee Score Number of pairs Median (IQR) Median (IQR) Median (IQR) P-value* 19 36 16 All cases 93,353 (13-24) (28-42) (9-22) P<0.001 19 36 16 All cemented 88,921 (13-24) (28-42) (9-22) P<0.001 19 35 14 All uncemented 4,067 (14-25) (27-42) (7-21) P<0.001 19 35 15 All hybrid 678 (13-24) (26-42) (7-21) P<0.001 All patello-femoral 282 20 32 10 P<0.001

*Wilcoxon matched-pairs signed-ranks test and checked with sign test.

198 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Figure 3.10 Histogram to compare the distributions of EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=84,031). (i) At Q1 15,000

0 0 20 40 60 80 100

Q1 EQ-5D HEALTH SCALE

Figure 3.10 Histogram to compare the distributions of EQ-5D Health Scale Score (VAS) between Q1 and Q2 in cases with scores at both time points (n=84,031). (ii) At Q2 15,000

10,000 Frequency

0 0 20 40 60 80 100

Q2 EQ-5D HEALTH SCALE

www.njrcentre.org.uk 199 Figure 3.11 Histogram to compare the distributions of the Oxford Knee Score between Q1 and Q2 in cases with scores

at both time points (n=93,353). (i) At Q1

6,000

5,000

4,000

2,000

1,000

0 0 10 20 30 40 50

Q1 Oxford Knee Score

Figure 3.11 Histogram to compare the distributions of the Oxford Knee Score between Q1 and Q2 in cases with scores

at both time points (n=93,353). (ii) At Q2

6,000

5,000

4,000

2,000

1,000

0 0 10 20 30 40 50

Q2 Oxford Knee Score

200 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

3.5.5 Conclusions quality of life post surgery. The median improvements in score on the three health scales were 0.38 on the Data linkage via HES allowed a match of PROMs data EQ-5D Index, 10 points on the VAS score and 21 to the subset of NJR records associated with English points on the Oxford Hip Score respectively. patients who had had NHS funded primary hip or knee replacement procedures (from the start of the collection In general, of the knee surgery patient group who of PROMs in April 2009 until the end of the current completed both questionnaires, most reported NJR annual report review period of 31 December enhanced lifestyle and health benefits post operation. 2012). Over a quarter of the identified orthopaedic Operation outcomes were rated as a success overall, HES episodes were lost initially due to the absence of although the proportion of those who were at least a traceable NHS number. Of the remaining PROMs satisfied with the outcomes were lower than the records, between a quarter and 29% of this group of ratings for hip surgery on the same items. That is, English NHS funded primary hip/knee replacement more than 96% of knee patients with responses to patients had only completed the pre-operation survey both Q1 and Q2 said they were at least fairly satisfied and so we are unable to report on how this patient with the results of their surgery. In addition, just over group evaluated their quality of life post-operation. In 70% of these patients intimated that the problems addition, based on the median responses (and IQR) they were faced with at circa six months post surgery to Q1 for this sub-group compared to those who did as a result of their knee condition were ‘much better’ complete both surveys, these patients tended to be in having had knee surgery. a poorer state of health prior to surgery. Therefore, in As noted already for the hip patient group, the knee interpreting the PROMs results for those patients who patient surgical group who had completed both completed both Q1 and Q2, it should be remembered surveys tended to have higher scores at Q1 and thus that these are patients who generally were healthier were in better health overall compared to the group of prior to operation and would have a better prospect of Q1 only respondees. surviving to six months after the primary surgery. Thus, pre- to post-operative changes in PROMs health scores Across all knee surgery cases, for the three knee are likely to be less improved with the inclusion of the PROMs measures of interest here, there was marked less healthy pre-operative subset of patients lost to improvement in patient self-reported well-being and follow up. health post operation. This was seen as a general increase in median scores and IQR values after The vast majority of the 93,846 hip replacement surgery i.e. responses to Q2 only and in the median patients who had responded to both Q1 and Q2 change in Q1 and Q2 knee health scores of 0.275 indicated at least a ‘fair’ amount of satisfaction post on the EQ-5D Index, 3 points on the VAS score and surgery (98.1%) and over 85% of these reported 16 points on the Oxford Knee Score respectively. that the problems they had had living with their hip However, the VAS score increase overall was more condition before surgery were ‘much better’ after moderate than for the other two measures. receiving a hip implant. The median score change was fairly consistent across It is clear that the hip patient group who completed the main fixation and knee replacement type sub- both Q2 and Q1 tended to be in a better state of groups apart from for those undergoing a patello- health before surgery than those who had only femoral type knee surgery. Here, the change in health completed Q1 (as indicated by higher scores on all score ratings compared to the other types of main three health measures prior to operation for these; knee replacement types suggested that post-surgery most notably on the EQ-5D Index). Therefore, these benefits to health and lifestyle were slightly less patients are more likely to have greater resilience improved than for other types of knee replacements. to undergoing surgery and are less likely to be susceptible to complications post surgery. This group reported a significant overall improvement in their

www.njrcentre.org.uk 201 Part 4 Trust-, Local Health Board- and unit-level activity and outcomes 2012

4.1 Introduction National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Part Four shows indicators for hip and knee joint replacement procedures by unit and Trust. This section was first included last year and has been updated as part of the Government’s transparency agenda. It is based on procedures carried out during the 2012 calendar year and submitted to the NJR by 28 February 2013.

Part Four information is based on the actual operation date (1 January to 31 December 2012) whereas data in Part One is based on the date the procedure was submitted (1 April 2012 to 31 March 2013). It is therefore possible for a hospital to have zero submissions in Part Four but not be listed as a nil returner in Part One.

www.njrcentre.org.uk 203 Part 4 4.2 Unit outlier analysis methodology National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Unit outlier analysis covers the period from much more extreme, i.e. above the upper limit of the 1 April 2003 to 31 December 2012. 99.99% control limits.

The outlier analyses made use of funnel plots from Previous annual reports have discussed in detail how statistical process control methodology. These aim to metal-on-metal bearing surfaces are associated with distinguish normal variation between units (which is to higher revision rates than other bearing surfaces. Many be expected) from unusual differences (termed ‘special of the outlying hip units identified from this analysis cause’ variation), which may indicate the need for further have more commonly used stemmed metal-on-metal investigation. Funnel plots enable units of different sizes and resurfacing procedures, and although use of these to be compared; performance indicators based on implants has now almost ceased in the UK, the higher smaller numbers of patients will have greater variability revision rates for these implants are likely to be reflected and this in turn is reflected by wider control limits. in the returns made by these units for some years to come. This may explain their higher revision rates. Summary of the methodology for unit revision rates: Summary of the methodology for unit 90-day • The standardised revision ratio (SRR) is plotted against mortality rates: the number of expected revisions. The SRR is the total number of revisions divided by the number of expected • The standardised mortality ratio (SMR) is plotted revisions for that unit’s caseload in respect of the age against number of expected deaths. The SMR is the group, gender, and the diagnosis of the patients. If the number of actual deaths within 90 days of primary SRR is 1, the number of revisions is compatible with joint replacement divided by the number of expected the number expected for that unit. deaths in this period. The number of expected deaths has been calculated after adjustment for patient • Control limits for funnel plots are normally set at 95% characteristics (age, gender, and ASA grade of and 99.8% (roughly equivalent to 2 or 3 standard patients). In addition, for hip units, adjustment is made deviations). Our normal six-monthly review process for whether the operation was undertaken for trauma. focuses on those above the upper limit of the latter; one might expect, however, that one in 500 would • Control limits are as for revisions above. be outside the 99.8% limits (above or below) just by Any units identified as potential outliers in Part Four chance. In the table below we have highlighted in light have been notified. All units are provided with an Annual red (1) units above the upper limit of the 99.8% control Clinical Report and additionally have access to an online limits but also indicate in dark red (2) those that are NJR Management Feedback system (See section 1.4.6).

Please note for the data in following table: Compliance, Consent and Linkability are: Red if lower than 80% Amber if equal to or greater than 80% and lower than 95% Green if 95% or more • Compliance figures may be low due to delayed data entry • Linkability for some hospitals will be lower than expected if they have private patients from outside England and Wales • Part Four data covers procedures carried out between 1 January 2012 and 31 December 2012

Outlier analyses are: 1 Light red if units are outside 99.8% control limits (approx 3 standard deviations (SDs)) 2 Dark red if units are outside 99.99% control limits

The table uses the following footnotes: 23. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. 24. Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed. 25. Linkability - the proportion of records which include a valid patient’s NHS number compared with the number of procedures recorded on the NJR. 26. Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). 27. Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2).

www.njrcentre.org.uk 205 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

1 2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 6% 0% 15% 20% 43% 91% 75% 90% 99% 99% 82% 97% 41% 32% 35% 51% 2012 Rated of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 6% 1% 3% 0% 2% 4% 4% 0% 0% 0% 0% 0% 13% 18% 14% 74% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 71.1 67.6 71.9 68.8 73.2 69.3 71.0 66.5 67.4 70.0 69.1 70.7 68.1 66.2 69.4 71.3 2012 Age At Average Average Operation 40% 44% 44% 43% 35% 43% 40% 43% 45% 41% 42% 43% 44% 75% 38% 37% 2012 Male Patients Percentage Percentage 2.0 2.4 2.0 2.1 2.1 2.2 2.1 2.0 2.2 2.2 2.3 2.0 1.5 1.9 2.2 2.2 2012 Average Average ASA Grade

25 98% 97% 98% 94% 86% 94% 99% 95% 99% 98% 97% 96% 100% 100% 100% 100% Linkability (%) 2012

24 5% Rate 99% 96% 94% 61% 88% 90% 86% 98% 91% 99% 75% 91% 86% 87% 100% Consent (%) 2012 6 9 9 7 3 11 10 14 16 11 14 16 17 19 13 16 2012 No. of Consultants 23 4 37 701 505 170 286 348 576 494 500 766 358 607 791 159 288 Only) 95% 83% 75% 86% 93% 97% 79% 80% 2012 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Abertawe Bro Morgannwg University Abertawe Bro Local Health Board Morriston Hospital Queens Hospital Neath Port Talbot Hospital Neath Port Talbot Barnet Farm Hospitals and Chase NHS Trust Barnet Hospital Princess of Wales Hospital Princess of Wales Chase Farm Hospital Aintree University Hospitals NHS University Aintree Foundation Trust University Hospital Aintree Barnsley NHS Foundation Hospital Trust Barnsley District General Hospital Airedale NHS Foundation Trust NHS Airedale General Hospital Airedale Aneurin Bevan Local Health Board Nevill Hall Hospital Royal Gwent Hospital St Woolos Hospital St Woolos Ysbyty Ystrad Fawr Ashford and St Peter's Hospitals Ashford NHS Foundation Trust Hospital Ashford St Peter's Hospital Barking Havering and Redbridge University Hospitals NHS Trust King George Hospital

206 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

2 2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 94% 53% 83% 96% 82% 67% 72% 89% 35% 61% 60% 33% 63% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 1% 3% 0% 82% 26% 39% 36% 36% 38% 60% 11% 45% 31% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 71.4 69.6 65.4 70.2 69.6 70.9 69.9 74.9 70.7 70.4 71.1 67.5 69.6 74.0 2012 Age At Average Average Operation 23% 38% 40% 40% 42% 43% 49% 47% 42% 36% 41% 43% 42% 29% 2012 Male Patients Percentage Percentage 2.4 2.0 2.0 2.2 2.2 2.2 1.9 2.9 2.2 2.2 2.0 2.1 2.6 2.5 2012 Average Average ASA Grade

25 96% 97% 98% 97% 99% 99% 97% 96% 99% 99% 99% 96% 100% 100% Linkability (%) 2012

24 96% Rate 92% 96% 99% 93% 97% 93% 97% 94% 87% 98% 84% 65% 100% Consent (%) 2012 6 9 9 8 8 7 7 20 11 11 12 14 11 16 2012 No. of Consultants 23 23 86 69 608 347 760 486 436 717 882 568 639 346 992 77% 88% 89% 88% 79% 85% 2012 Only) No. of 210% 118% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Buckinghamshire Healthcare NHS Trust Healthcare Buckinghamshire Wycombe Hospital Stoke Mandeville Hospital Barts and The London NHS Trust Barts and The London The Royal London Hospital Basildon and Thurrock University Basildon and Thurrock Trust Hospitals NHS Foundation Basildon University Hospital Bedford Hospital NHS Trust Hospital Bedford South Wing Hospital Bedford Betsi Cadwaladr University Local Betsi Cadwaladr University Health Board Abergele Hospital Glan Clwyd General Hospital Wrexham Maelor Hospital Wrexham Ysbyty Gwynedd Blackpool Teaching Hospitals NHS Blackpool Teaching Foundation Trust Blackpool Victoria Hospital Bradford Teaching Hospitals NHS Hospitals NHS Teaching Bradford Foundation Trust Royal Infirmary Bradford Brighton and Sussex University Hospitals NHS Trust Princess Royal Hospital Royal Sussex County Hospital Sussex Orthopaedic NHS Treatment Sussex Orthopaedic NHS Treatment Centre 26 27 24 25 23

www.njrcentre.org.uk 207 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 99% 94% 56% 99% 51% 70% 27% 98% 89% 2012 Rated 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 9% 0% 0% 0% 50% 47% 15% 47% 36% 26% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 71.3 70.4 69.3 70.1 67.2 67.8 69.2 69.1 68.9 67.6 70.9 2012 Age At Average Average Operation 40% 45% 43% 44% 40% 43% 41% 41% 43% 44% 39% 2012 Male Patients Percentage Percentage 2.2 2.0 2.1 2.1 2.0 2.5 2.1 2.1 2.2 2.2 2.2 2012 Average Average ASA Grade

25 98% 99% 98% 98% 93% 93% 83% 95% 98% 100% 100% Linkability (%) 2012

24 Rate 97% 96% 86% 98% 86% 88% 86% 47% 95% 86% 100% Consent (%) 2012 8 9 9 14 15 16 12 11 10 11 10 2012 No. of Consultants 23 758 602 276 337 374 762 844 284 Only) 98% 93% 62% 89% 90% 87% 84% 97% 80% 2012 1,113 1,529 1,117 No. of 103% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Burton Hospitals NHS Foundation Burton Hospitals NHS Foundation Trust Queens Hospital Burton Upon Trent Queens Hospital Burton Calderdale and Huddersfield NHS Calderdale Foundation Trust Royal Hospital Calderdale Cambridge University Hospitals NHS Cambridge University Hospitals Foundation Trust Hospital Addenbrooke's Cardiff and Vale University Local and Vale Cardiff Health Board Llandough Hospital Central Manchester University Hospitals NHS Foundation Trust Manchester Royal Infirmary Trafford General Hospital Trafford Chelsea and Westminster Hospital Chelsea and Westminster NHS Foundation Trust Hospital Chelsea & Westminster Chesterfield Royal Hospital NHS Foundation Trust Chesterfield Royal Hospital City Hospitals Sunderland NHS Foundation Trust Sunderland Royal Hospital Colchester Hospital University NHS Foundation Trust Colchester General Hospital Countess of Chester Hospital NHS Foundation Trust Countess of Chester Hospital

208 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 0% 90% 49% 75% 97% 86% 92% 95% 80% 70% 84% 52% 97% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 8% 2% 6% 0% 11% 28% 23% 11% 23% 21% 71% 13% 21% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 67.3 68.6 69.0 57.9 70.3 72.9 68.1 67.6 68.3 70.9 69.3 70.2 66.8 71.0 2012 Age At Average Average Operation 0% 48% 40% 39% 44% 38% 25% 45% 45% 40% 40% 43% 43% 42% 2012 Male Patients Percentage Percentage 1.9 2.1 2.4 3.0 2.2 2.3 2.4 2.2 2.0 2.0 2.1 2.1 2.1 2.2 2012 Average Average ASA Grade

25 99% 99% 98% 97% 98% 91% 99% 99% 97% 98% 98% 100% 100% 100% Linkability (%) 2012

24 Rate 87% 97% 98% 96% 72% 77% 96% 97% 98% 82% 100% 100% 100% 100% Consent (%) 2012 2 4 7 6 7 9 4 19 13 12 12 11 16 15 2012 No. of Consultants 23 2 22 408 256 536 227 341 583 575 173 441 821 180 87% 84% 67% 84% 86% 54% 99% 79% 2012 Only) 1,673 No. of 118% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local County Durham and Darlington NHS County Durham and Darlington Foundation Trust Bishop Auckland Hospital Darlington Memorial Hospital East and North Hertfordshire NHS East and North Hertfordshire Trust Lister Hospital University Hospital of North Durham University Hospital of North Queen Elizabeth II Hospital Croydon Health Services NHS Trust Health Croydon Hospital University Croydon Cwm Taf Local Health Board Cwm Taf Prince Charles Hospital The Royal Glamorgan Hospital Dartford and Gravesham NHS Trust Dartford Hospital Valley Darent Derby Hospitals NHS Foundation Trust Royal Derby Hospital Doncaster and Bassetlaw Hospitals NHS Foundation Trust Bassetlaw Hospital Doncaster Royal Infirmary Dorset County Hospital NHS Foundation Trust Dorset County Hospital Ealing Hospital NHS Trust Ealing Hospital 26 27 24 25 23

www.njrcentre.org.uk 209 www.njrcentre.org.uk 2 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 87% 99% 96% 88% 76% 65% 41% 99% 99% 65% 92% 91% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 2% 0% 0% 0% 1% 0% 8% 4% 38% 34% 15% 13% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 71.1 71.4 69.6 69.1 69.9 69.7 74.6 68.9 72.6 70.4 69.5 68.6 69.2 2012 Age At Average Average Operation 43% 39% 39% 44% 40% 40% 33% 43% 35% 38% 42% 44% 41% 2012 Male Patients Percentage Percentage 2.1 2.1 2.0 2.0 2.1 2.1 2.6 2.1 2.4 2.0 2.0 2.3 2.1 2012 Average Average ASA Grade

25 99% 98% 96% 98% 74% 96% 99% 90% 98% 99% 98% 98% 100% Linkability (%) 2012

24 Rate 97% 93% 96% 83% 73% 54% 99% 64% 98% 96% 92% 90% 100% Consent (%) 2012 3 9 9 7 12 11 13 14 18 15 25 19 18 2012 No. of Consultants 23 70 318 766 865 298 706 481 420 927 455 Only) 68% 62% 85% 70% 95% 96% 94% 2012 1,070 2,710 1,211 No. of 101% 111% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local East Cheshire NHS Trust East Cheshire Macclesfield District General Macclesfield District General Hospital Gloucestershire Hospitals NHS Gloucestershire Foundation Trust Cheltenham General Hospital East Kent Hospitals University NHS East Kent Hospitals University Foundation Trust Mother Queen Elizabeth The Queen Hospital Gloucestershire Royal Hospital Gloucestershire William Hospital (Ashford) Harvey East Lancashire Hospitals NHS Trust East Lancashire Burnley Hospital General Royal Blackburn Infirmary East Sussex Hospitals NHS Trust Conquest Hospital Eastbourne General Hospital District Epsom and St Helier University Hospitals NHS Trust The Elective Orthopaedic Centre Frimley Park Hospital NHS Foundation Trust Frimley Park Hospital Gateshead Health NHS Foundation Trust The North East NHS Surgery Centre George Eliot Hospital NHS Trust George Eliot Hospital - Acute Services

210 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 0% 99% 92% 88% 29% 92% 95% 89% 97% 2012 Rated 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 0% 0% 0% 6% 0% 0% 1% 6% 1% 66% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 69.1 69.5 63.4 66.9 70.1 70.1 70.3 70.2 70.5 68.8 69.4 71.6 2012 Age At Average Average Operation 41% 42% 42% 50% 21% 39% 41% 45% 43% 41% 39% 41% 2012 Male Patients Percentage Percentage 2.1 2.2 2.0 2.0 2.0 2.0 2.2 2.0 2.1 2.1 2.0 2.4 2012 Average Average ASA Grade

25 97% 93% 96% 64% 98% 97% 93% 95% 98% 99% 100% 100% Linkability (%) 2012

24 89% Rate 97% 38% 50% 27% 94% 96% 94% 54% 60% 96% 94% Consent (%) 2012 7 2 4 7 9 14 14 11 13 23 11 11 2012 No. of Consultants 23 4 22 624 783 947 649 765 323 785 636 201 89% 93% 91% 87% 70% 98% 89% 2012 Only) 1,113 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Guy's Hospital Great Western Hospitals NHS Western Great Foundation Trust Western Hospital The Great Guy's and St Thomas' NHS Foundation Guy's and St Thomas' NHS Trust Guy's Nuffield House Guy's Nuffield St Thomas' Hospital Hampshire Hospitals NHS Hampshire Foundation Trust Basingstoke and North Hampshire Hospital Royal Hampshire County Hospital Royal Hampshire Harrogate and District NHS Harrogate Foundation Trust District Hospital Harrogate Heart of England NHS Foundation Trust Good Hope Hospital Solihull Hospital Heatherwood and Wexham Park Heatherwood and Wexham Hospitals NHS Foundation Trust Heatherwood Hospital Wexham Park Hospital Wexham Hinchingbrooke Health Care NHS Health Care Hinchingbrooke Trust Hospital Hinchingbrooke 26 27 24 25 23

www.njrcentre.org.uk 211 www.njrcentre.org.uk 2 2 27 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 80% 91% 89% 93% 95% 71% 85% 18% 99% 2012 Rated 100% 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 0% 0% 0% 34% 10% 14% 50% 38% 36% 77% 27% 21% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 72.2 67.9 66.7 67.5 69.7 70.2 68.5 68.6 67.8 70.9 70.2 71.6 68.8 2012 Age At Average Average Operation 38% 41% 35% 29% 42% 48% 17% 44% 37% 40% 39% 38% 43% 2012 Male Patients Percentage Percentage 2.5 2.1 2.1 1.8 2.2 2.0 2.3 2.1 2.1 2.2 2.1 2.3 2.0 2012 Average Average ASA Grade

25 94% 88% 92% 97% 99% 99% 99% 98% 100% 100% 100% 100% 100% Linkability (%) 2012

24 Rate 84% 99% 73% 98% 77% 93% 96% 99% 98% 78% 100% 100% 100% Consent (%) 2012 4 9 5 6 7 5 5 7 8 19 10 16 10 2012 No. of Consultants 23 46 13 741 176 372 606 176 432 629 798 535 628 760 94% Only) 77% 96% 75% 17% 87% 90% 94% 98% 2012 No. of 452% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local King's College Hospital NHS Foundation Trust King's College Hospital (Denmark Hill) Prince Philip Hospital Castle Hill Hospital Hywel Dda Local Health Board Homerton University Hospital NHS Homerton University Hospital Foundation Trust Homerton University Hospital Hospital General Bronglais Hull and East Yorkshire Hospitals Hospitals Hull and East Yorkshire NHS Trust General Hospital Wales West Kingston Hospital NHS Trust Kingston Hospital Withybush General Hospital Imperial College Healthcare NHS Imperial College Healthcare Trust Hospital Charing Cross Ipswich Hospital NHS Trust Ipswich Hospital Isle of Wight NHS PCT St Mary's Hospital James Paget University Hospitals NHS Foundation Trust James Paget University Hospital Kettering General Hospital NHS Foundation Trust Kettering General Hospital

212 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 97% 73% 99% 99% 93% 37% 84% 85% 96% 76% 83% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 6% 0% 7% 0% 4% 88% 39% 16% 17% 18% 11% 16% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 68.0 68.3 67.5 68.9 67.8 70.6 70.3 68.2 70.6 70.7 68.4 75.6 2012 Age At Average Average Operation 40% 45% 42% 36% 39% 42% 41% 44% 43% 42% 40% 45% 2012 Male Patients Percentage Percentage 2.2 2.2 2.1 2.2 2.2 2.0 2.2 2.2 2.1 2.1 2.8 2.1 2012 Average Average ASA Grade

25 89% 99% 96% 99% 99% 98% 98% 98% 96% 99% 100% 100% Linkability (%) 2012

24 Rate 98% 85% 89% 88% 98% 99% 87% 97% 95% 96% 62% 98% Consent (%) 2012 6 9 9 6 8 12 12 20 12 13 12 11 2012 No. of Consultants 23 47 760 642 269 628 687 299 833 591 354 94% 88% 90% 88% 99% 89% 54% 86% 83% 2012 Only) 1,133 1,029 No. of 105% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Lancashire Teaching Hospitals NHS Teaching Lancashire Foundation Trust Chorley and South Ribble Hospital Chorley and South Ribble Pinderfields General Hospital Leeds Teaching Hospitals NHS Trust Hospitals Leeds Teaching Chapel Allerton Hospital Lewisham Healthcare NHS Trust Lewisham Healthcare Centre Riverside Treatment Luton and Dunstable Hospital NHS Luton and Dunstable Hospital Foundation Trust Luton & Dunstable Hospital Maidstone and Tunbridge Wells NHS Wells Maidstone and Tunbridge Trust Hospital Wells The Tunbridge Medway NHS Foundation Trust Medway Maritime Hospital Mid Cheshire Hospitals NHS Mid Cheshire Foundation Trust Leighton Hospital Mid Essex Hospital Services NHS Trust Hospital Broomfield Mid Staffordshire NHS Foundation Mid Staffordshire Trust Cannock Chase Hospital Stafford Hospital Stafford Mid Yorkshire Hospitals NHS Trust Mid Yorkshire Dewsbury and District Hospital 26 27 24 25 23

www.njrcentre.org.uk 213 www.njrcentre.org.uk 2 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 88% 60% 97% 58% 91% 67% 96% 75% 98% 88% 2012 Rated 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 3% 0% 1% 54% 14% 24% 64% 26% 50% 67% 54% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 69.2 69.1 67.3 68.1 70.9 72.2 69.9 73.4 68.1 70.7 69.5 70.0 69.1 2012 Age At Average Average Operation 41% 46% 44% 47% 40% 43% 36% 43% 43% 44% 45% 42% 45% 2012 Male Patients Percentage Percentage 2.1 1.9 2.4 2.0 2.3 2.2 2.2 2.4 2.2 2.2 2.5 2.1 2.1 2012 Average Average ASA Grade

25 98% 98% 99% 97% 93% 95% 97% 98% 98% 100% 100% 100% 100% Linkability (%) 2012

24 Rate 98% 97% 98% 99% 85% 69% 88% 93% 69% 91% 98% 100% 100% Consent (%) 2012 7 4 6 5 3 7 6 20 12 12 29 11 10 2012 No. of Consultants 23 18 10 72 389 322 179 613 401 296 687 387 Only) 86% 87% 99% 94% 88% 77% 2012 1,369 1,671 No. of 107% 100% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Milton Keynes Hospital NHS Milton Keynes Hospital Foundation Trust Milton Keynes Hospital Goole Treatment Centre Goole Treatment Newham University Hospital NHS Newham University Hospital Trust Gateway Surgical Centre Scunthorpe General Hospital Newham General Hospital Norfolk and Norwich University Norfolk and Norwich University Trust Hospitals NHS Foundation Norfolk & Norwich Hospital Northampton General Hospital NHS Trust Northampton General Hospital (Acute) North Bristol NHS Trust Hospital Frenchay Southmead Hospital North Cumbria University Hospitals NHS Trust Cumberland Infirmary West Cumberland Hospital West Northern NHS Healthcare Devon Trust North Devon District Hospital Northern and Goole Lincolnshire Hospitals NHS Foundation Trust Hospital Diana Princess of Wales

214 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 2 2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 82% 99% 66% 80% 26% 81% 56% 64% 77% 2012 Rated 100% 100% 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 0% 0% 0% 7% 9% 0% 19% 19% 16% 41% 16% 29% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 68.6 68.6 69.5 67.5 68.4 75.4 68.5 69.4 69.2 69.0 73.1 69.2 69.4 74.4 2012 Age At Average Average Operation 43% 27% 39% 44% 42% 38% 41% 44% 45% 38% 48% 45% 41% 24% 2012 Male Patients Percentage Percentage 2.1 2.3 2.2 2.1 2.2 2.1 1.8 2.1 2.1 2.0 2.2 2.2 2.1 1.9 2012 Average Average ASA Grade

25 96% 97% 91% 89% 99% 99% 99% 96% 96% 100% 100% 100% 100% 100% Linkability (%) 2012

24 Rate 94% 77% 92% 84% 78% 98% 89% 72% 98% 99% 100% 100% 100% 100% Consent (%) 2012 6 8 4 6 2 18 17 10 13 15 13 15 17 14 2012 No. of Consultants 23 9 25 221 616 844 366 556 741 785 357 101 313 75% 73% 93% 94% 77% 99% 2012 Only) 1,172 1,791 No. of 102% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local North Middlesex University Hospital North Middlesex University NHS Trust North Middlesex Hospital Pennine Acute Hospitals NHS Trust Fairfield General Hospital North Tees and Hartlepool NHS and Hartlepool North Tees Foundation Trust University Hospital of Hartlepool North Manchester General Hospital University Hospital of North Tees University Hospital of North Northumbria Healthcare NHS Northumbria Healthcare Foundation Trust Hexham General Hospital North Tyneside General Hospital North Tyneside Wansbeck Hospital Wansbeck North West London Hospitals NHS North West Trust Central Middlesex Hospital Northwick Park Hospital Nottingham University Hospitals NHS Trust Nottingham City Hospital Queens Medical Centre Nottingham Queens Medical Centre University Hospital Oxford University Hospitals NHS Oxford Trust Hospital John Radcliffe Nuffield Orthopaedic Centre Nuffield 26 27 24 25 23

www.njrcentre.org.uk 215 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 6% 99% 54% 82% 99% 70% 99% 92% 97% 98% 56% 42% 2012 Rated of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 1% 0% 4% 0% 38% 79% 68% 15% 24% 13% 14% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 70.5 68.1 69.3 74.9 71.1 69.8 69.3 68.8 69.1 67.8 70.4 68.0 2012 Age At Average Average Operation 44% 38% 34% 34% 40% 40% 43% 40% 43% 48% 40% 43% 2012 Male Patients Percentage Percentage 2.2 2.2 2.5 2.2 2.2 2.0 2.0 2.1 2.4 2.2 1.8 2.1 2012 Average Average ASA Grade

25 99% 95% 87% 96% 99% 99% 86% 94% 92% 97% 100% 100% Linkability (%) 2012

24 Rate 97% 64% 83% 77% 86% 98% 99% 86% 84% 84% 92% 100% Consent (%) 2012 8 15 12 10 16 22 35 17 11 19 17 12 2012 No. of Consultants 23 860 106 202 281 816 925 424 565 390 47% Only) 76% 87% 84% 92% 89% 96% 89% 93% 2012 1,162 2,944 1,447 No. of 101% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Poole Hospital NHS Foundation Trust Poole Hospital NHS Foundation Poole Hospital Derriford Hospital Derriford Royal Oldham Hospital Royal Free London NHS Foundation Royal Free Trust Hospital The Royal Free Peterborough and Stamford and Stamford Peterborough Trust Hospitals NHS Foundation City Hospital Peterborough Portsmouth Hospitals NHS Trust Queen Alexandra Hospital Plymouth Hospitals NHS Trust Plymouth Hospitals NHS Robert Jones and Agnes Hunt Orthopaedic And District Hospital NHS Trust Robert Jones & Agnes Hunt Orthopaedic Hospital Royal Berkshire NHS Foundation Royal Berkshire Trust Hospital Royal Berkshire Royal Bolton Hospital NHS Foundation Trust Royal Bolton Hospital Royal Cornwall Hospitals NHS Trust Royal Cornwall Hospital (Treliske) St Michael's Hospital Royal Devon and Exeter NHS Foundation Trust Royal Devon & Exeter Hospital (Wonford)

216 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 96% 89% 51% 96% 73% 99% 17% 70% 66% 49% 35% 46% 2012 Rated of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 3% 0% 7% 4% 0% 0% 0% 0% 65% 72% 25% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 70.3 69.8 67.2 69.4 72.1 62.9 70.2 69.5 67.2 71.0 68.2 68.6 2012 Age At Average Average Operation 49% 45% 43% 44% 41% 38% 40% 39% 47% 41% 37% 43% 2012 Male Patients Percentage Percentage 2.1 2.1 2.2 2.3 2.6 2.0 2.0 2.2 2.3 2.1 2.2 2.2 2012 Average Average ASA Grade

25 80% 96% 98% 99% 99% 97% 84% 99% 98% 97% 100% 100% Linkability (%) 2012

24 Rate 84% 64% 92% 97% 32% 88% 99% 78% 36% 99% 98% 92% Consent (%) 2012 8 9 9 17 26 11 18 17 12 10 11 10 2012 No. of Consultants 23 63 588 539 678 824 670 687 340 680 502 239 69% 87% 88% 89% 70% 96% 91% 2012 Only) 1,189 No. of 103% 106% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Sheffield Teaching Hospitals NHS Teaching Sheffield Foundation Trust Northern Hospital General Scarborough General Hospital Scarborough Royal Liverpool and Broadgreen Royal Liverpool and Broadgreen Trust University Hospitals NHS Broadgreen Hospital Broadgreen Royal Hallamshire Hospital Royal Hallamshire The Royal Liverpool University The Royal Liverpool University Hospital Royal National Orthopaedic Hospital Royal National Orthopaedic NHS Trust The Royal National Orthopaedic Hospital (Stanmore) Royal Surrey County Hospital NHS Royal Surrey Foundation Trust County Hospital Royal Surrey Royal United Hospital Bath NHS Trust Royal United Hospital Salford Royal NHS Foundation Trust Salford Royal Salford Salisbury NHS Foundation Trust Salisbury District Hospital Sandwell and West Birmingham Sandwell and West Hospitals NHS Trust City Hospital Sandwell General Hospital Scarborough and North East Scarborough NHS Trust Health Care Yorkshire 26 27 24 25 23

www.njrcentre.org.uk 217 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 98% 99% 84% 93% 75% 78% 83% 95% 34% 84% 43% 94% 99% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 7% 0% 0% 0% 0% 0% 8% 16% 13% 11% 33% 23% 2012 Rated 100% of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 68.2 68.1 67.8 70.5 69.3 76.6 69.4 67.1 70.9 70.5 74.7 70.8 70.5 70.5 2012 Age At Average Average Operation 44% 43% 45% 40% 39% 25% 38% 40% 40% 41% 39% 46% 39% 41% 2012 Male Patients Percentage Percentage 2.0 2.1 2.2 1.8 2.2 2.1 2.2 2.1 2.2 2.1 2.4 2.5 2.0 2.0 2012 Average Average ASA Grade

25 93% 99% 99% 89% 98% 85% 99% 95% 98% 99% 99% 100% 100% 100% Linkability (%) 2012

24 Rate 83% 91% 99% 60% 89% 93% 80% 93% 97% 98% 97% 100% 100% 100% Consent (%) 2012 5 6 3 6 9 8 14 15 13 18 17 10 10 20 2012 No. of Consultants 23 9 693 659 649 142 290 512 572 802 919 150 125 707 524 Only) 97% 85% 62% 93% 89% 74% 96% 93% 2012 No. of 100% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Sherwood Forest Hospitals NHS Sherwood Forest Foundation Trust Kings Mill Hospital The James Cook University Hospital Newark Hospital South Tyneside NHS Foundation South Tyneside Trust District Hospital South Tyneside Shrewsbury and Telford Hospital and Telford Shrewsbury NHS Trust Hospital Royal Shrewsbury The Princess Royal Hospital Southampton University Hospitals Southampton University NHS Trust Southampton General Hospital South Devon Healthcare NHS South Devon Healthcare Foundation Trust Hospital Torbay Southend University Hospital NHS Foundation Trust Southend Hospital South London Healthcare NHS Trust South London Healthcare Princess Royal University Hospital Queen Elizabeth Hospital Woolwich Queen Mary's Hospital Sidcup Southport and Ormskirk Hospital NHS Trust Ormskirk & District General Hospital South Tees Hospitals NHS South Tees Foundation Trust Friarage Hospital

218 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 1 1 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 84% 96% 90% 99% 79% 97% 59% 50% 98% 83% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 0% 0% 1% 6% 6% 2% 0% 20% 12% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 69.7 69.5 66.7 68.2 68.7 71.2 70.7 68.7 71.5 70.1 73.1 2012 Age At Average Average Operation 41% 41% 42% 43% 42% 39% 39% 45% 41% 45% 40% 2012 Male Patients Percentage Percentage 2.0 2.1 2.4 2.1 2.1 2.3 2.2 2.1 2.2 2.0 2.4 2012 Average Average ASA Grade

25 98% 98% 98% 96% 95% 95% 98% 98% 99% 97% 93% Linkability (%) 2012

24 Rate 97% 96% 84% 93% 89% 83% 95% 98% 99% 71% 92% Consent (%) 2012 9 8 8 8 7 9 11 15 11 13 11 2012 No. of Consultants 23 810 157 721 812 194 296 386 707 906 182 512 89% 85% 93% 92% 81% 87% 96% 84% 93% 2012 Only) No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local South Warwickshire NHS NHS South Warwickshire Foundation Trust Warwick Hospital Warwick St George's Healthcare NHS Trust St George's Healthcare St George's Hospital (Tooting) St Helens and Knowsley Hospitals St Helens and Knowsley NHS Trust Whiston Hospital Stockport NHS Foundation Trust Stockport NHS Foundation Stepping Hill Hospital Surrey and Sussex Healthcare NHS and Sussex Healthcare Surrey Trust Hospital East Surrey Redwood Diagnostic Treatment Redwood Diagnostic Treatment Centre Tameside Hospital NHS Foundation Tameside Trust General Hospital Tameside Taunton and Somerset NHS Taunton Foundation Trust Park Hospital Musgrove The Dudley Group of Hospitals NHS The Dudley Group Foundation Trust Russells Hall Hospital The Hillingdon Hospital NHS Foundation Trust Hillingdon Hospital Mount Vernon Treatment Centre Treatment Mount Vernon 26 27 24 25 23

www.njrcentre.org.uk 219 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 56% 90% 50% 97% 99% 64% 63% 83% 28% 99% 2012 Rated 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 6% 2% 2% 5% 14% 50% 67% 20% 35% 59% 16% 30% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 69.5 68.1 71.2 73.1 70.7 70.0 71.1 68.3 71.1 65.8 69.1 68.5 70.1 2012 Age At Average Average Operation 31% 42% 39% 12% 41% 40% 42% 44% 39% 40% 43% 34% 46% 2012 Male Patients Percentage Percentage 2.2 2.2 2.4 2.2 2.0 2.4 2.0 2.1 2.0 2.2 2.1 2.0 2.1 2012 Average Average ASA Grade

25 94% 98% 93% 97% 97% 98% 98% 97% 97% 97% 99% 99% 100% Linkability (%) 2012

24 Rate 82% 84% 54% 97% 63% 89% 98% 97% 87% 99% 85% 99% 97% Consent (%) 2012 7 9 7 5 7 13 12 12 11 12 15 20 13 2012 No. of Consultants 23 46 267 461 701 609 701 859 226 141 272 Only) 77% 93% 81% 93% 97% 73% 71% 73% 74% 2012 1,147 2,142 1,360 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local The Newcastle Upon Tyne Hospitals The Newcastle Upon Tyne NHS Foundation Trust Freeman Hospital Freeman Lincoln County Hospital The Royal Victoria Infirmary Pilgrim Hospital The Princess Alexandra Hospital The Princess Alexandra NHS Trust Princess Alexandra Hospital The Queen Elizabeth Hospital King's The Queen Elizabeth Hospital NHS Foundation Trust Lynn The Queen Elizabeth Hospital The Rotherham NHS Foundation Trust Rotherham District General Hospital The Royal Bournemouth and Hospitals NHS Christchurch Foundation Trust Royal Bournemouth Hospital The Royal Orthopaedic Hospital NHS Foundation Trust Royal Orthopaedic Hospital The Royal Wolverhampton Hospitals The Royal Wolverhampton NHS Trust Hospital New Cross The Whittington Hospital NHS Trust The Whittington Hospital United Lincolnshire Hospitals NHS United Lincolnshire Trust County Hospital Louth Grantham and District Hospital

220 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 0% 99% 29% 97% 93% 79% 95% 63% 84% 92% 2012 Rated 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 4% 0% 1% 2% 2% 0% 6% 7% 0% 30% 68% 11% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 64.7 70.9 73.9 68.0 68.6 69.9 12.2 74.2 67.8 70.6 67.3 69.6 69.5 2012 Age At Average Average Operation 0% 0% 36% 42% 44% 41% 44% 39% 31% 39% 39% 40% 45% 2012 Male Patients Percentage Percentage 2.1 2.3 2.2 1.9 2.2 2.1 1.5 2.6 1.9 2.4 2.0 2.1 2.0 2012 Average Average ASA Grade

25 93% 92% 78% 99% 98% 88% 97% 96% 99% 100% 100% 100% 100% Linkability (%) 2012

24 Rate 57% 99% 13% 84% 90% 50% 39% 97% 96% 94% 98% 100% 100% Consent (%) 2012 9 7 1 1 7 10 11 13 12 11 22 25 23 2012 No. of Consultants 23 2 1 32 33 313 164 369 641 438 929 557 386 96% 80% 80% 82% 61% 86% 85% 73% 2012 Only) 1,798 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local University College London Hospitals University College London NHS Foundation Trust University College Hospital Royal Lancaster Infirmary University Hospital Birmingham NHS University Hospital Birmingham Trust Queen Elizabeth Hospital Birmingham Westmorland General Hospital Westmorland University Hospital of North University Hospital of North NHS Trust Staffordshire City General Hospital University Hospital of South Manchester NHS Foundation Trust Wythenshawe Hospital University Hospitals Bristol NHS Foundation Trust Bristol Royal Hospital for Children Bristol Royal Infirmary University Hospitals Coventry and NHS Trust Warwickshire Hospital of St Cross University Hospital (Coventry) University Hospitals of Leicester NHS Trust Glenfield Hospital Leicester General Hospital University Hospitals of Morecambe University Hospitals of Morecambe Bay NHS Foundation Trust Furness Hospital General 26 27 24 25 23

www.njrcentre.org.uk 221 www.njrcentre.org.uk 1 2 27 Rate - Knee Outliers Revision

2 1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 94% 85% 62% 48% 85% 64% 99% 96% 98% 76% 95% 56% 99% 2012 Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 5% 0% 2% 0% 0% 13% 14% 57% 48% 13% 52% 64% 13% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 67.3 69.8 69.5 69.6 68.6 67.7 71.4 69.2 74.4 69.1 70.6 71.0 70.3 72.2 2012 Age At Average Average Operation 41% 39% 43% 42% 44% 51% 39% 42% 35% 41% 44% 42% 31% 41% 2012 Male Patients Percentage Percentage 2.1 2.1 2.1 2.2 1.7 2.2 2.0 2.7 2.1 2.1 2.1 2.1 2.5 1.9 2012 Average Average ASA Grade

25 98% 94% 98% 99% 95% 88% 94% 98% 61% 99% 99% 100% 100% 100% Linkability (%) 2012

24 Rate 87% 94% 96% 75% 74% 88% 82% 64% 91% 72% 85% 86% 100% 100% Consent (%) 2012 6 7 7 5 9 7 9 9 11 17 12 11 12 15 2012 No. of Consultants 23 89 363 501 572 525 780 177 190 294 777 149 445 629 Only) 92% 81% 72% 82% 97% 69% 78% 92% 2012 1,252 No. of 100% 132% Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Worcestershire Royal Hospital Worcestershire Walsall Healthcare NHS Trust NHS Healthcare Walsall Manor Hospital Worcestershire Acute Hospitals NHS Worcestershire Trust Alexandra Hospital Warrington and Halton Hospitals Warrington NHS Foundation Trust Hospital Warrington Kidderminster Treatment Centre Kidderminster Treatment Western Sussex Hospitals NHS Western Sussex Hospitals Trust Hospital St Richard's West Hertfordshire Hospitals NHS Hospitals Hertfordshire West Trust St Albans City Hospital Watford General Hospital Watford West Middlesex University Hospital West NHS Trust Middlesex University Hospital West Weston Area Health NHS Trust Health Area Weston General Weston West Suffolk Hospitals NHS Trust Hospitals Suffolk West Hospital Suffolk West Whipps Cross University Hospital Whipps Cross NHS Trust University Hospital Whipps Cross Wirral University Teaching Hospital Wirral University Teaching NHS Foundation Trust Park Hospital Arrowe Clatterbridge Hospital

222 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 91% 86% 58% 82% 98% 77% 75% 82% 54% 84% 64% 88% 97% 96% 67% 92% 2012 Rated of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 1% 0% 0% 1% 0% 1% 4% 8% 0% 0% 9% 0% 96% 34% 14% 12% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 66.4 66.9 70.4 69.2 71.6 70.8 70.0 70.3 67.5 69.6 66.4 70.5 67.3 69.7 69.5 69.3 2012 Age At Average Average Operation 45% 37% 41% 44% 40% 44% 39% 47% 43% 43% 45% 46% 41% 37% 52% 33% 2012 Male Patients Percentage Percentage 2.1 1.8 2.0 2.0 2.1 2.0 2.3 2.0 1.9 1.9 1.8 2.0 1.9 1.9 2.1 1.9 2012 Average Average ASA Grade

25 80% 95% 99% 97% 99% 90% 91% 85% 73% 99% 85% 91% 92% 89% 100% 100% Linkability (%) 2012

24 Rate 60% 98% 98% 99% 89% 93% 91% 96% 95% 93% 96% 78% 98% 100% 100% 100% Consent (%) 2012 7 7 8 6 8 5 8 8 25 13 11 20 12 11 30 10 2012 No. of Consultants 23 84 33 695 317 676 153 353 277 596 209 271 611 242 102 124 96% 81% 99% 97% 2012 Only) 2,266 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Wrightington WiganWrightington and Leigh NHS Foundation Trust Wrightington Hospital Wrightington BMI Blackheath Hospital Wye Valley NHS Trust Wye Valley County Hospital Hereford BMI Bury St Edmunds Yeovil District Hospital NHS Yeovil Foundation Trust District Hospital Yeovil BMI Chaucer Hospital York Teaching Hospital NHS Teaching York Foundation Trust Hospital York BMI Chelsfield Park Hospital Aspen Healthcare Limited Aspen Healthcare Hospital Claremont Holly House Hospital Parkside Hospital Benenden Healthcare Society Benenden Healthcare Benenden Hospital BMI Healthcare BMI Alexandra Hospital Cheadle BMI Bath Clinic BMI Beardwood Private Hospital BMI Beardwood BMI Bishops Wood Hospital BMI Bishops Wood 26 27 24 25 23

www.njrcentre.org.uk 223 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 89% 79% 98% 54% 27% 99% 55% 84% 87% 96% 59% 52% 90% 75% 81% 26% 74% 40% 68% 2012 97% 90% Rated 100% 100% 100% 100% 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 2% 1% 4% 6% 0% 0% 0% 5% 0% 0% 0% 0% 0% 0% 0% 0% 39% 45% 60% 12% 36% 60% 25% 64% 41% 2012 19% 15% 30% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 65.0 68.4 68.0 67.0 68.3 49.2 70.0 70.7 67.7 66.7 63.6 69.6 65.5 66.7 69.8 71.0 69.7 67.6 70.3 68.3 67.7 65.8 69.1 68.3 71.0 70.3 67.5 70.8 2012 Age At Average Average Operation 45% 35% 38% 44% 34% 37% 43% 42% 53% 27% 49% 47% 45% 42% 41% 38% 43% 44% 40% 43% 48% 46% 47% 42% 43% 32% 2012 46% Male 100% Patients Percentage Percentage 2.0 1.9 2.0 1.9 3.0 1.7 2.2 1.8 1.9 2.0 2.0 1.8 2.0 1.9 2.0 2.0 2.1 2.0 1.9 2.0 1.9 1.3 2.0 1.8 2.0 1.9 1.9 1.9 2012 Average Average ASA Grade

25 98% 90% 71% 92% 95% 93% 88% 95% 84% 85% 87% 99% 83% 94% 89% 88% 94% 92% 89% 95% 97% 93% 88% 96% 95% 91% 78% 100% Linkability (%) 2012

24 0% Rate 89% 98% 73% 98% 99% 98% 99% 92% 95% 91% 99% 97% 84% 96% 95% 97% 99% 96% 45% 85% 69% 68% 100% 100% 100% 100% 100% Consent (%) 2012 9 5 1 5 5 5 8 9 8 7 6 9 8 9 2 5 4 7 19 13 13 11 13 10 23 10 11 13 2012 No. of Consultants 23 1 91 89 98 39 28 50 55 324 148 217 131 182 357 255 170 620 361 239 158 263 477 192 204 119 163 292 190 Only) 2012 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local BMI Clementine Churchill Hospital BMI Clementine Churchill BMI The Cavell Hospital BMI Manor Hospital BMI Coombe Wing BMI Mount Alvernia Hospital BMI Esperance BMI Princess Margaret BMI Fawkham Manor Hospital BMI Priory Hospital BMI Garden Hospital BMI Garden BMI Ridgeway Hospital BMI Gisburne Park Hospital BMI Runnymede Hospital BMI Goring Hall Hospital BMI Sandringham Hospital BMI Hampshire Clinic BMI Hampshire BMI Sarum Road Hospital BMI Harrogate BMI Highfield Hospital BMI Saxon Clinic BMI Huddersfield BMI Sefton Hospital BMI Kings Oak Hospital BMI Shelburne Hospital BMI Lancaster BMI Shirley Oaks Hospital BMI Lincoln BMI The Beaumont Hospital

224 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 71% 99% 53% 72% 68% 45% 46% 54% 92% 52% 70% 97% 86% 82% 2012 74% 38% Rated 100% 100% 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 1% 0% 0% 0% 0% 5% 0% 51% 29% 29% 27% 59% 29% 48% 47% 24% 24% 10% 2012 28% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 72.8 69.2 72.1 69.4 66.5 62.8 68.9 66.6 69.4 70.4 68.5 64.1 68.0 66.4 68.3 67.0 68.7 68.4 69.4 65.8 68.4 2012 Age At Average Average Operation 38% 48% 42% 48% 43% 55% 44% 47% 41% 44% 43% 32% 40% 52% 45% 45% 42% 47% 44% 48% 2012 46% Male Patients Percentage Percentage 2.0 1.9 2.1 1.9 1.8 1.8 2.0 1.7 2.1 1.8 2.2 1.7 1.9 1.9 2.0 1.9 1.9 1.6 1.9 1.7 2.0 2012 Average Average ASA Grade

25 88% 95% 94% 97% 92% 72% 99% 99% 86% 54% 96% 91% 93% 93% 84% 91% 95% 93% 96% 100% 100% Linkability (%) 2012

24 Rate 96% 97% 96% 96% 87% 99% 94% 95% 57% 93% 84% 55% 79% 92% 99% 97% 98% 100% 100% 100% 100% Consent (%) 2012 7 7 7 6 3 6 6 9 6 5 19 17 24 11 11 16 17 12 10 21 11 2012 No. of Consultants 23 88 35 71 101 277 255 265 279 410 478 145 473 463 486 402 118 110 177 706 390 2012 Only) 1,078 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local BMI The Chiltern BMI Winterbourne Hospital BMI The Droitwich Spa Hospital BMI The Droitwich BMI Woodlands Hospital BMI Woodlands BMI The Edgbaston Hospital Care UK Care Centre NHS Treatment Barlborough BMI The Foscote Hospital North East London NHS Treatment North East London NHS Treatment Centre BMI The Harbour Hospital Southampton NHS Treatment Centre Southampton NHS Treatment BMI The London Independent BMI The London Independent Hospital Circle Bath Hospital Circle BMI The Meriden Hospital Circle Reading Circle BMI The Park Hospital BMI The Sloane Hospital BMI The Somerfield Hospital BMI The South Cheshire Private BMI The South Cheshire Hospital BMI Thornbury Hospital BMI Three Shires Hospital Shires BMI Three BMI Werndale Hospital 26 27 24 25 23

www.njrcentre.org.uk 225 www.njrcentre.org.uk 1 27 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 0% 57% 66% 84% 44% 83% 50% 40% 88% 12% 84% 2012 98% Rated 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 5% 0% 0% 5% 0% 0% 0% 4% 16% 13% 27% 14% 20% 44% 2012 88% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 68.2 70.1 71.1 66.6 69.2 68.9 69.2 61.4 66.0 66.4 69.3 70.2 71.4 69.1 68.8 2012 Age At Average Average Operation 48% 44% 38% 61% 38% 37% 49% 64% 27% 44% 40% 40% 45% 40% 46% 2012 Male Patients Percentage Percentage 1.9 2.0 1.9 1.7 2.0 1.9 1.8 1.9 1.7 1.8 2.0 1.8 2.1 2.0 2.0 2012 Average Average ASA Grade

25 58% 99% 93% 58% 91% 97% 99% 61% 66% 65% 58% 99% 97% 73% 100% Linkability (%) 2012

24 Rate 41% 87% 99% 85% 99% 97% 85% 77% 86% 34% 99% 97% 98% 100% 100% Consent (%) 2012 9 6 9 8 2 8 15 16 12 12 13 16 15 12 12 2012 No. of Consultants 23 66 92 64 87 30 367 228 154 100 401 149 272 109 443 Only) 2012 1,963 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Clinicenta Limited Lister Surgicentre Nuffield Health Nuffield Health Bournemouth Nuffield Hospital Cromwell Hospital Cromwell Hospital Bupa Cromwell Nuffield Health Brentwood Hospital Health Brentwood Nuffield East Kent Medical Services Ltd East Kent Medical Services The Spencer Wing - Margate Fairfield Independent Hospital Fairfield Independent Hospital HCA International Ltd London Bridge Hospital The Lister Hospital The Princess Grace Hospital The Wellington Hospital The Wellington Horder Healthcare Horder Healthcare Horder Hospital Management Trust Sancta Maria Hospital St Hugh's Hospital Hospital of St John and St Elizabeth Hospital of St John and St Elizabeth King Edward VII Hospital Sister King Edward Agnes VII Hospital Sister King Edward Agnes

226 www.njrcentre.org.uk 27 Rate - Knee Outliers Revision

2 2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 89% 61% 52% 24% 95% 96% 78% 52% 65% 85% 12% 95% 90% 68% 94% 2012 97% Rated 100% 100% 100% 100% 100% 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 0% 0% 2% 2% 0% 0% 1% 1% 0% 0% 5% 0% 0% 0% 9% 60% 24% 26% 11% 37% 68% 39% 2012 Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 64.7 69.0 68.5 68.2 68.0 70.1 67.5 69.4 69.8 71.2 66.3 68.4 69.4 71.4 67.9 70.9 70.7 69.4 69.3 70.2 68.3 67.9 2012 Age At Average Average Operation 49% 53% 42% 45% 42% 39% 48% 41% 42% 40% 44% 45% 44% 42% 48% 43% 49% 46% 46% 39% 48% 2012 44% Male Patients Percentage Percentage 1.9 2.0 1.8 2.1 1.9 2.0 1.8 1.9 1.9 1.8 1.9 1.9 1.9 1.8 1.8 1.9 1.9 1.9 1.7 2.1 2.0 1.9 2012 Average Average ASA Grade

25 92% 83% 98% 83% 96% 96% 94% 86% 99% 94% 99% 96% 99% 92% 88% 95% 93% 95% 90% 99% 94% 95% Linkability (%) 2012

24 Rate 99% 87% 97% 90% 94% 90% 93% 75% 99% 89% 89% 99% 95% 92% 60% 91% 100% 100% 100% 100% 100% 100% Consent (%) 2012 5 7 8 8 9 8 7 9 9 6 3 4 7 6 13 10 10 11 15 13 16 14 2012 No. of Consultants 23 86 95 521 126 457 302 270 145 202 374 253 511 511 365 490 460 111 128 303 264 640 206 2012 Only) No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Nuffield Health Brighton Hospital Health Nuffield Nuffield Health North Staffordshire Health North Staffordshire Nuffield Hospital Nuffield Health Bristol Hospital Health Nuffield Nuffield Health Plymouth Hospital Nuffield Nuffield Health Cambridge Hospital Health Nuffield Nuffield Health Shrewsbury Hospital Health Shrewsbury Nuffield Nuffield Health Cheltenham Hospital Health Nuffield Nuffield Health Taunton Hospital Health Taunton Nuffield Nuffield Health Chichester Hospital Health Nuffield Nuffield Health Tees Hospital Health Tees Nuffield Nuffield Health Derby Hospital Health Nuffield Nuffield Health The Grosvenor Health The Grosvenor Nuffield Hospital Nuffield Health Exeter Hospital Health Nuffield Nuffield Health The Manor Hospital Nuffield Nuffield Health Guildford Hospital Guildford Health Nuffield Nuffield Health Tunbridge Wells Wells Health Tunbridge Nuffield Hospital Nuffield Health Haywards Heath Health Haywards Nuffield Hospital Nuffield Health Hereford Hospital Health Hereford Nuffield Nuffield Health Ipswich Hospital Nuffield Nuffield Health Leeds Hospital Nuffield Nuffield Health Leicester Hospital Nuffield Nuffield Health Newcastle-upon- Nuffield Hospital Tyne 26 27 24 25 23

www.njrcentre.org.uk 227 www.njrcentre.org.uk 2 2 27 Rate - Knee Outliers Revision

1 2 2 2 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 6% 0% 75% 84% 99% 89% 89% 97% 83% 66% 75% 72% 37% 98% 79% 99% 72% 83% 27% 76% 2012 80% Rated 100% of 10A Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 3% 0% 0% 2% 0% 2% 7% 0% 1% 1% 0% 0% 0% 1% 0% 9% 10% 19% 28% 23% 10% 2012 70% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 67.7 68.0 68.2 68.1 67.5 67.7 68.7 70.1 69.6 69.2 68.2 68.6 68.2 71.6 64.2 69.5 70.4 67.8 69.6 67.3 66.0 67.6 2012 Age At Average Average Operation 46% 43% 41% 45% 45% 39% 42% 43% 45% 48% 40% 37% 43% 18% 41% 37% 39% 46% 40% 42% 40% 45% 2012 Male Patients Percentage Percentage 2.1 2.0 1.9 1.8 1.9 1.8 2.0 1.9 1.9 1.9 1.8 1.9 1.7 2.1 1.6 1.9 1.9 1.8 1.9 1.9 1.9 1.9 2012 Average Average ASA Grade

25 90% 98% 89% 96% 89% 98% 98% 98% 92% 99% 83% 97% 95% 95% 98% 97% 99% 98% 100% 100% 100% 100% Linkability (%) 2012

24 Rate 86% 97% 86% 96% 99% 99% 99% 97% 98% 98% 93% 96% 88% 97% 97% 97% 99% 99% 100% 100% 100% 100% Consent (%) 2012 5 7 8 7 8 9 1 5 7 7 9 9 7 16 10 16 10 10 17 12 13 13 2012 No. of Consultants 23 76 74 38 287 809 160 345 300 310 335 294 235 873 184 359 345 849 419 912 222 526 402 Only) 2012 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Nuffield Health Warwickshire Warwickshire Health Nuffield Hospital Horton NHS Treatment Centre Horton NHS Treatment Nuffield Health Wessex Hospital Wessex Health Nuffield Kendal NHS Treatment Centre Kendal NHS Treatment Nuffield Health Woking Hospital Woking Health Nuffield Mount Stuart Hospital Nuffield Health Wolverhampton Wolverhampton Health Nuffield Hospital New Hall Hospital Nuffield Health York Hospital York Health Nuffield North Downs Hospital Vale Hospital Vale Nottingham Woodthorpe Hospital Nottingham Woodthorpe Orthopaedic Spine and Specialist Orthopaedic Spine and Hospital Orthopaedics and Spine Specialist Hospital Oaklands Hospital Ramsay Health Care Ashtead Hospital Oaks Hospital Clifton Park Hospital Park Hill Hospital Duchy Hospital Euxton Hall Hospital Fitzwilliam Hospital Fulwood Hall Hospital

228 www.njrcentre.org.uk 2 2 27 Rate - Knee Outliers Revision

2 1 2 1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2012 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 68% 96% 71% 85% 52% 54% 2012 86% 95% 98% 95% 67% 65% 89% 99% 72% 41% 51% 86% 96% Rated 100% 100% of 10A 100% 100% 100% Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 4% 8% 8% 3% 0% 5% 0% 0% 1% 3% 0% 2% 0% 49% 11% 19% 2012 21% 10% 11% 74% 14% 12% 34% 16% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 68.8 67.6 68.7 69.5 68.9 68.3 67.6 69.4 2012 69.6 69.1 67.4 67.8 69.5 69.9 68.0 68.1 68.1 66.1 65.7 66.7 63.3 67.1 66.6 67.8 Age At Average Average Operation 45% 36% 52% 40% 46% 42% 44% 45% 2012 Male 35% 40% 43% 47% 32% 37% 45% 42% 43% 45% 38% 42% 47% 43% 48% 38% Patients Percentage Percentage 1.9 1.9 1.9 2.1 2.0 1.8 1.9 1.9 1.8 2.0 1.9 2.0 1.7 1.8 2.1 2.0 1.5 2.0 1.9 1.7 2.0 2.0 1.8 1.8 2012 Average Average ASA Grade

25 99% 97% 87% 99% 95% 88% 93% 95% 98% 95% 95% 96% 92% 94% 97% 93% 92% 92% 90% 98% 90% 88% 94% 94% Linkability (%) 2012

24 Rate 92% 99% 94% 95% 92% 97% 97% 92% 94% 99% 90% 95% 99% 99% 98% 98% 100% 100% 100% 100% 100% 100% 100% 100% Consent (%) 2012 5 5 6 9 8 9 8 9 7 7 8 8 11 10 14 18 14 10 12 20 16 19 14 10 2012 No. of Consultants 23 291 179 205 155 481 202 647 182 291 376 350 122 219 290 184 366 450 571 504 587 306 515 271 303 2012 Only) No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local Pinehill Hospital Woodland Hospital Woodland Methley Park Hospital Renacres Hall Hospital Renacres Spire Healthcare Spire Dunedin Hospital Regency Hospital Rivers Hospital Elland Hospital Spire Alexandra Hospital Spire Fylde Coast Hospital Rowley Hall Hospital Spire Bristol Hospital Spire Hull and East Riding Hospital Spire Bushey Hospital Spire Springfield Hospital Spire Cambridge Lea Hospital Spire The Berkshire Independent Hospital The Berkshire Spire Cardiff Hospital Cardiff Spire The Yorkshire Clinic The Yorkshire Spire Cheshire Hospital Cheshire Spire West Midlands Hospital Midlands West Spire Clare Park Hospital Clare Spire Spire Gatwick Park Hospital Spire Winfield Hospital 26 27 24 25 23

www.njrcentre.org.uk 229 www.njrcentre.org.uk 27 2 Rate - Knee Outliers Revision

1 27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 2012 76% 88% 32% 99% 86% 27% 94% 99% 96% 84% 98% 58% 50% 81% 66% 78% 44% 98% 61% 55% Rated of 10A 100% 100% 100% 100% Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 6% 0% 9% 0% 7% 4% 0% 7% 1% 2% 9% 3% 7% 4% 0% 2% 5% 2012 21% 10% 38% 25% 28% 70% 28% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 2012 70.3 68.4 70.0 67.1 67.8 68.6 71.8 68.4 69.9 70.3 66.1 67.2 68.0 66.5 68.0 67.6 69.1 69.9 69.8 67.8 60.1 70.6 68.7 66.1 Age At Average Average Operation 2012 Male 46% 40% 47% 45% 34% 40% 43% 40% 31% 41% 51% 51% 42% 42% 42% 42% 42% 44% 55% 43% 40% 39% 41% 46% Patients Percentage Percentage 1.9 1.9 1.9 2.0 2.0 1.8 1.9 2.0 1.6 1.9 1.9 2.0 1.8 2.0 2.1 1.9 2.0 1.9 2.0 1.7 1.6 1.7 2.0 1.8 2012 Average Average ASA Grade

25 93% 93% 92% 95% 96% 89% 90% 96% 88% 98% 93% 86% 94% 95% 89% 95% 91% 95% 95% 98% 64% 96% 94% 82% Linkability (%) 2012

24 Rate 96% 50% 99% 99% 98% 99% 96% 74% 97% 85% 94% 98% 95% 95% 99% 96% 90% 99% 90% 98% 100% 100% 100% 100% Consent (%) 2012 9 7 3 7 8 8 16 15 12 10 18 18 17 10 18 13 10 15 11 11 10 15 15 12 2012 No. of Consultants 23 81 232 217 301 343 304 111 376 274 309 201 311 183 269 271 224 942 136 439 168 527 178 719 200 Only) 2012 No. of Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Hospital Trust/Local Health Board/Company Trust/Local Spire Harpenden Hospital Harpenden Spire Spire St Saviours Hospital Spire Spire Hartswood Hospital Hartswood Spire Spire Sussex Hospital Spire Spire Leeds Hospital Leeds Spire Spire Thames Valley Spire Spire Leicester Hospital Leicester Spire Spire Tunbridge Wells Hospital Wells Tunbridge Spire Spire Little Aston Hospital Little Spire Spire Washington Hospital Washington Spire Spire Liverpool Hospital Liverpool Spire Spire Wellesley Hospital Wellesley Spire Spire Manchester Hospital Manchester Spire Spire Yale Hospital Yale Spire Spire Murrayfield Hospital Murrayfield Spire St Anthonys St Anthonys Spire Norwich Hospital Spire St Josephs Hospital St Josephs Hospital Spire Parkway Hospital Spire The London Clinic The London Clinic Spire Portsmouth Hospital Spire Spire Roding Hospital Spire Spire Southampton Hospital Spire Spire South Bank Hospital Spire

230 www.njrcentre.org.uk 27 1 Rate - Knee Outliers Revision

27 - Hip Rate Outliers Revision 26 © National Joint Registry 2013 Knees Outliers - 90 Day Mortality 26 Hips Outliers - 90 Day Mortality 2012 29% 49% Rated of 10A 100% 100% Primary Femoral Percentage Percentage Procedures Procedures Implant Hip 6% 7% 0% 2012 12% Rated of 10A Primary Acetabular Percentage Percentage Procedures Procedures Implant Hip 2012 72.8 69.5 70.1 71.3 Age At Average Average Operation 2012 Male 36% 42% 41% 45% Patients Percentage Percentage 2.0 1.9 2.0 1.9 2012 Average Average ASA Grade

25 85% 100% 100% 100% Linkability (%) 2012

24 Rate 99% 93% 100% 100% Consent (%) 2012 5 8 7 3 2012 No. of Consultants 23 85 868 612 2012 Only) No. of 1,157 Compliance Compliance Health Board Health Board Procedures Procedures (NHS Trust/Local (NHS Trust/Local Dark red if units are outside 99.99% control limits outside 99.99% control if units are Dark red Green if 95% or more if 95% Green deviations (SDs)) 3 standard limits (approx outside 99.8% control if units are Light red Red if lower than 80% than 80% and lower than 95% Amber if equal to or greater 1 2 Compliance figures may be low due to delayed data entry Compliance figures outside England and Wales than expected if they have private patients from Linkability for some hospitals will be lower 2012 and 31 December 2012 carried out between 1 January Part Four data covers procedures Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted to the NJR compared to HES/PEDW. Compliance (NHS Trust/Local Health Board Only) - the percentage of cases submitted Consent Rate - percentage of cases submitted to the NJR with patient consent confirmed compared with the number of procedures recorded on the NJR. Linkability - the proportion of records which include a valid patient’s NHS number l limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark Outliers with respect to 90-day mortality following primary operations performed from 2003 onwards. Units control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in dark red (2). Units outside 99.8% control limits (approx 3 SDs) are flagged in light red (1); those outside 99.99% control limits are flagged in Outliers with respect to revisions (at any time) following primary operations performed from 2003 onwards. Hospital Trust/Local Health Board/Company Trust/Local The New Victoria Hospital The New Victoria Hospital UK Specialist Hospitals Ltd UK Specialist Hospitals NHS Treatment Emersons Green Centre Peninsula NHS Treatment Centre Peninsula NHS Treatment Shepton Mallet Treatment Centre Shepton Mallet Treatment

Outlier analyses are: • • • Please note: Compliance, Consent and Linkability are: 26 27 24 25 23

www.njrcentre.org.uk 231 Glossary National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

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 See Acetabular component. Acetabular prosthesis See Acetabular component. Antibiotic-loaded bone cement See cement. Arthrodesis A procedure where the bones of a natural joint are fused together (stiffened). Arthroplasty A procedure where a natural joint is reconstructed with an artificial prosthesis. ABHI Association of British Healthcare Industries - the UK trade association of medical device suppliers. ASA American Society of Anaesthesiologists scoring system for grading the overall physical condition of the patient, as follows: P1 – fit and healthy; P2 – mild disease, not incapacitating; P3 – incapacitating systemic disease; P4 – life threatening disease; P5 – expected to die within 24 hrs without an operation.

Bearing type The two surfaces that articulate together in a joint replacement. Options include metal-on-polyethylene, metal-on-metal, ceramic-on-polyethylene, ceramic-on-metal and ceramic-on- ceramic. Bilateral operation Operation performed on both sides, e.g. left and right knee procedures, carried out during a single operation. BMI Body mass index. A statistical tool used to estimate a healthy body weight based on an individual’s height. The BMI is calculated by dividing a person’s weight (kg) by the square of their height (m2). BOA British Orthopaedic Association - the professional body representing orthopaedic surgeons. Bone cement See cement. Brand (of prosthesis) The brand of a prosthesis (or implant) is the manufacturer's product name, e.g. the Exeter V40 brand for hips, the PFC Sigma brand for knees, the Mobility brand for ankles, the Delta Xtend brand for shoulders and the Coonrad Morrey for elbows.

CQC Care Quality Commission. Regulators of care provided by the NHS, local authorities, private companies and voluntary organisations. Case ascertainment Proportion of all relevant joint replacement procedures performed in England and Wales that are entered into the NJR. Case mix Term used to describe variation in surgical practice, relating to factors such as indications for surgery, patient age and gender. Cement The material used to fix cemented joint replacements to bone - polymethyl methacrylate (PMMA). Antibiotic can be added to bone cement to try and reduce the risk of infection. Cemented Prostheses designed to be fixed into the bone using cement. Cementless Prostheses designed to be fixed into the bone by bony ingrowth or ongrowth, without using cement. Compliance The percentage of all total joint procedures that have been entered into the NJR within any given period compared with the expected number of procedures performed. The expected number of procedures can be the number of levies returned, or for the NHS Sector only, the number of procedures submitted to HES and PEDW. Competing risks survival analysis An alternative to standard survival analysis methods (such as Kaplan-Meier estimation or the Cox proportional hazards model) when there are competing risks. A competing risk can prevent the event of interest from occurring (in this case, death is a competing risk to the risk of revision as patients who die will never experience revision). A competing-risks survival analysis adjusts the results accordingly. Confidence interval (CI) A confidence interval (CI) gives an estimated range of values which is likely to include the unknown population parameter (e.g. a revision rate) being estimated from the given sample. If independent samples are taken repeatedly from the same population, and a confidence interval calculated for each sample, then a certain percentage (confidence level: e.g. 95%) of the intervals will include the unknown population parameter.

www.njrcentre.org.uk 233 Confounding Systematic variation due to the presence of factors not on the causal pathway, which affect the outcome, which are unequally distributed amongst interventions being compared which leads to inaccurate inferences about the results. Cox proportional hazards model A semi-parametric survival analysis model commonly used to model time-to-event data as it does not require the underlying hazard function to take a particular shape. As it is a multi-variable model, it can be used to explore the effects of covariates on the outcome of interest and reduce the impact of confounding. Cross-linked polyethylene See modified polyethylene. Cumulative hazard An ‘accumulated’ hazard rate over time; the hazard rate at a particular point in time is the rate of occurrence of the event (revision or death) amongst those in whom the event has not yet occurred. Estimated using the Nelson-Aalen estimate. Cup See Acetabular component.

Data collection periods for annual The NJR Annual Report Part One reports on data collected between 1 April 2012 and 31 March report analysis 2013 – the 2012/13 financial year. The NJR Annual Report Parts Two and Four analyse data on hip, knee, ankle, elbow, and shoulder procedures undertaken between 1 January and 31 December 2012 inclusive – the 2012 calendar year. The NJR Annual Report Part Three reports on hip, knee and ankle joint replacement revision rates for procedures that took place between 1 April 2003 and 31 December 2012. DDH Developmental dysplasia of the hip. A condition where the hip joint is malformed, usually with a shallow socket (acetabulum), which may cause instability. DH Department of Health. DVT Deep vein thrombosis. A blood clot that can form in the veins of the leg, and is recognised as a significant risk after joint replacement surgery.

Excision arthroplasty A procedure where the articular ends of the bones are simply excised, so that a gap is created between them, or when a joint replacement is removed and not replaced by another prosthesis.

Femoral component (hip) Part of a 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 Spherical portion of the femoral component of the artificial hip replacement. Femoral prosthesis Portion of a total joint replacement used to replace damaged parts of the femur (thigh bone). Femoral stem The part of a modular femoral component inserted into the femur (thigh bone). Has a femoral head mounted on it to form the complete femoral component. Flexible parametric proportional Developed by Royston and Parmar, this model extends the standard Cox proportional hazards hazards model approach by modelling the baseline distribution parametrically using a restricted cubic spline function. This allows more flexibility in modelling the shape of the baseline hazard function than using standard parametric distributions. Funnel plot A graph of a performance measure for each unit plotted against the unit’s number of cases. Control limits are shown to indicate acceptable performance.

Glenoid component The portion of a total shoulder replacement prosthesis that is inserted into the scapula – the socket part of a ball and socket joint in conventional shoulder replacement or the ball part in reverse shoulder replacement. Glenoid head Domed head portion of the glenoid component of the reverse shoulder replacement attached to the scapula.

234 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

Head See Femoral head and/or Humeral head. Healthcare provider NHS or independent sector organisation that provides healthcare; in the case of the NJR, orthopaedic hip, knee, ankle, elbow or shoulder replacement surgery. HES Hospital Episode Statistics. Data on case mix, procedures, length of stay and other hospital statistics collected routinely by NHS hospitals in England. HQIP Healthcare Quality Improvement Partnership. Manages the NJR on behalf of the Department of Health. Promotes quality in health and social care services and works to increase the impact that clinical audit has nationally. Humeral component (elbow) Part of a total elbow joint that is inserted into the humerus (upper arm bone) of the patient to replace the articulating surface of the humerus. Humeral component (shoulder) Part of a total or partial shoulder joint that is inserted into the humerus (upper arm bone) of the patient. It normally consists of a humeral stem and head (ball) in conventional shoulder replacement or a humeral stem and a humeral cup in a reverse shoulder replacement. Humeral cup The shallow socket of a reverse shoulder replacement attached to the scapula. Humeral head Domed head portion of the humeral component of the artificial shoulder replacement attached to the humeral stem. Humeral prosthesis Portion of a total joint replacement used to replace damaged parts of the humerus (upper arm bone). Humeral stem The part of a modular humeral component inserted into the humerus (upper arm bone). Has a humeral head or humeral cup mounted on it to form the complete humeral component. Hybrid procedure Joint replacement procedure in which cement is used to fix one prosthetic component while the other is cementless. For hip procedures, the term hybrid covers both reverse hybrid (cementless stem, cemented socket) and hybrid (cemented stem, cementless socket).

Image/computer-guided surgery Surgery performed by the surgeon, using real-time images and data computed from these to assist alignment and positioning of prosthetic components. Independent hospital A hospital managed by a commercial company that predominantly treats privately-funded patients but does also treat NHS-funded patients. Index joint The primary joint replacement that is the subject of an NJR entry. Indication (for surgery) The reason for surgery. The NJR system allows for more than one indication to be recorded. ISTC Independent sector treatment centre (see Treatment centre).

Kaplan-Meier Estimates of the cumulative probability of failure (revision or death) that properly take into account ‘censored’ data. Censorings arise from incomplete follow up; for revision, for example, a patient may have died or reached the end the analysis period (end of 2012) without having been revised. The estimates do not adjust for any confounding factors.

Lateral resurfacing (elbow) Partial resurfacing of the elbow with a humeral surface replacement component used with a lateral resurfacing head inserted with or without cement. Levy Additional payment placed on the sales of specific hip, knee, ankle, elbow, and shoulder implants to cover the costs associated with the ongoing operation and development of the NJR. Linkable percentage Linkable percentage is the percentage of all relevant procedures that have been entered into the NJR, which 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. Linked total elbow Where the humeral and ulnar parts of a total elbow replacement are physically connected.

www.njrcentre.org.uk 235 LHMoM Large head metal-on-metal. Where a metal femoral head of 36mm diameter or greater is used in conjunction with a femoral stem, and is articulating with either a metal resurfacing cup or a metal liner in a modular acetabular cup. Resurfacing hip replacements are excluded from this group. LMWH Low molecular weight Heparin. A blood-thinning drug used in the prevention and treatment of deep vein thrombosis (DVT).

MDS Minimum dataset, 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 dataset version one, used to collect data from 1 April 2003. MDS 1 closed to new data entry on 1 April 2005. MDS 2 (MDSv2) Minimum dataset version two, introduced on 1 April 2004. MDS 2 replaced MDS 1 as the official data set on 1 June 2004. MDS 3 (MDSv3) Minimum dataset version three, introduced on 1 November 2007 replacing MDS 2 as the new official data set. MDS 4 (MDSv4) Minimum dataset version four, introduced on 1 April 2010 replacing MDS 3 as the new official dataset. This dataset has the same hip and knee MDS 3 dataset but includes the data collection for total ankle replacement procedures. MDS 5 (MDSv5) Minimum dataset version five, introduced on 1 April 2012 replacing MDS 4 as the new official dataset. This dataset has the same hip, knee and ankle MDS 4 dataset but includes the data collection for total elbow and total shoulder replacement procedures. MHRA Medicines and Healthcare products Regulatory Agency – the UK regulatory body for medical devices. Minimally-invasive surgery Surgery performed using small incisions (usually less than 1Ocm). This may require the use of special instruments. Mixing and matching Also known as 'cross breeding'. Hip replacement procedure in which a surgeon chooses to implant a femoral component from one manufacturer with an acetabular component from another. Modified Polyethylene Any component made of polyethylene which has been modified in some way in order to improve its performance characteristics. Some of these processes involve chemical changes, such as increasing the cross-linking of the polymer chains or the addition of vitamin E and/or other antioxidants . Others are physical processes such as heat pressing or irradiation in a vacuum or inert gas. Modular Component composed of more than one piece, e.g. a modular acetabular cup shell component with a modular cup liner, or femoral stem coupled with a femoral head. Monobloc Component composed of, or supplied as, one piece, e.g. a monobloc knee tibial component.

Nelson-Aalen estimator An estimate of the cumulative hazard rate. NHS National Health Service. NICE National Institute for Health and Care Excellence. NICE benchmark See ODEP ratings. NJR National Joint Registry for England, Wales and Northern Ireland. The NJR has collected and analysed data on hip and knee replacements since 1 April 2003, on ankle replacements since 1 April 2010 and on elbow replacements and shoulder replacements since April 2012. It covers both the NHS and independent healthcare sectors to ensure complete recording of national activity in England, Wales and Northern Ireland. NJR Centre National coordinating centre for the NJR. NJR StatsOnline Web facility for viewing and downloading NJR statistics on www.njrcentre.org.uk

ODEP Orthopaedic Data Evaluation Panel of the NHS Supply Chain. www.odep.org.uk

236 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

ODEP ratings ODEP ratings are the criteria for product categorisation of prostheses for primary total hip replacement against benchmarks. The letter represents the strength of evidence and the number the length of time in years during which the implant has been studied. The full benchmark is 10A and the entry is at 3 years with progression through 5 and 7 years. Pre-entry submissions are also recorded. “A” represents strong supporting evidence for the use of the prosthesis, “B” less strong but acceptable evidence. All implants that are used without a 10-year benchmark should be followed up closely. OPCS-4 Office of Population, Censuses and Surveys: Classification of Surgical Operations and Procedures, 4th Revision – a list of surgical procedures and codes. Outlier Data for a surgeon, unit or implant brand that falls outside of the defined control limits.

Pantalar (ankle) Affecting the whole talus, i.e. the ankle (tibio talar) joint, the subtalar (talo calcaneal) joint and the talonavicular joint. Patella resurfacing Replacement of the surface of the patella (knee cap) with a prosthesis. Patello-femoral knee Procedure involving replacement of the trochlear and replacement resurfacing of the patella. Patello-femoral prosthesis Two-piece knee prosthesis that provides a prosthetic (knee) articulation surface between the patella and trochlear. Patient consent Patient personal details may only be submitted to the NJR where explicit informed patient consent has been given or where patient consent has not been recorded. If a patient declines to 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. Patient-time The summation of time (in years) for a cohort of primary procedures where the time is measured from the primary date to either date of revision, date of patient's death or analysis date (last observation date). PDS The NHS Personal Demographics Service is the national electronic database of NHS patient demographic details. The NJR uses the PDS Demographic Batch Service (DBS) to source missing NHS numbers and to determine when patients recorded on the NJR have died. PEDW Patient Episode Database for Wales. The Welsh equivalent to Hospital Episode Statistics (HES) in England. Poisson distribution This distribution expresses the probability of a number of relatively rare events occurring in a fixed time if these events occur with a known average rate and are independent of the time since the last event. It is a special case of the binomial distribution in that it models discrete events. Primary hip/knee/ankle/elbow/ The first time a total joint replacement operation is performed on any individual joint in a patient. shoulder replacement Prosthesis Orthopaedic implant used in joint replacement procedures, e.g. a total hip, a unicondylar knee, a total ankle, a reverse shoulder or a radial head replacement. PROMs Patient Reported Outcome Measures. PTIR Patient-Time Incidence Rate. This is the rate of occurrences of an event (i.e. revision) for a given total time at risk.

Radial head component (elbow) Part of a partial elbow joint that is inserted into the radius (outer lower arm bone) of the patient to replace the articulating surface of the radial head. May be monobloc or modular. Resurfacing (hip) Resurfacing of the femoral head with a surface replacement femoral prosthesis and insertion of a monobloc acetabular cup, with or without cement. Resurfacing (shoulder) Resurfacing of the humeral head with a surface replacement humeral prosthesis inserted, with or without cement. Reverse shoulder replacement Replacement of the shoulder joint where a glenoid head is attached to the scapula and the humeral cup to the humerus. Revision burden The proportion of revision procedures carried out as a percentage of the total number of surgeries on that particular joint. Revision hip/knee/ankle/elbow/ Operation performed to remove (and usually replace) one or more components of a total joint prosthesis shoulder replacement for whatever reason.

www.njrcentre.org.uk 237 S

Shoulder hemiarthroplasty Replacement of the humeral head with a humeral stem and head or shoulder resurfacing component which articulates with the natural glenoid. Single-stage revision A revision carried out in a single operation. Subtalar joint (ankle) The joint between the talus and the calcaneum. Surgical approach Method used by a surgeon to gain access to, and expose, the joint. Survivorship analysis A statistical method that is used to determine what fraction of a population, such as those who have had a particular hip implant, has survived unrevised past a certain time. See Kaplan-Meier.

Talar component Portion of an ankle prosthesis that is used to replace the articulating surface of the talus at the ankle joint. TAR Total ankle replacement (total ankle arthroplasty). Replacement of both tibial and talar surfaces, with or without cement. TED stockings Thrombo embolus deterrent (TED) stockings. Elasticised stockings that can be worn by patients following surgery and which may help reduce the risk of deep vein thrombosis (DVT). THR Total hip replacement (total hip arthroplasty). Replacement of the femoral head with a stemmed femoral prosthesis and 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, usually deep vein thrombosis (DVT), in the post-operative period. Tibial component (knee) Portion of a knee prosthesis that is used to replace the articulating surface of the tibia (shin bone) at the knee joint. May be modular or monobloc (one piece). Tibial component (ankle) Portion of an ankle prosthesis that is used to replace the articulating surface of the tibia (shin bone) at the ankle joint. TKR Total knee replacement (total knee arthroplasty). Replacement of both tibial and femoral condyles (with or without resurfacing of the patella), 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’s knee. Treatment centre Treatment centres are dedicated units that offer elective and short-stay surgery and diagnostic procedures in specialties such as ophthalmology, orthopaedic and other conditions. These include hip, knee, ankle, elbow, and shoulder replacements. Treatment centres may be privately funded (independent sector treatment centre – ISTC). NHS Treatment Centres exist but their data is included in those of the English NHS Trusts and Welsh Local Health Boards to which they are attached. Trochanter Bony protuberance of the femur, found on its upper outer aspect. Trochanteric osteotomy Temporary incision of the trochanter, used to aid exposure of hip joint during some types of total hip replacement. Two-stage revision A revision procedure carried out as two operations, often used in the treatment of deep infection. Type (of prosthesis) Type of prosthesis is the generic description of a prosthesis, e.g. modular cemented stem (hip), patello- femoral joint (knee), talar component (ankle), reverse shoulder (shoulder) and radial head replacement (elbow).

Ulnar component (elbow) Part of a total elbow joint that is inserted into the ulna (inner lower arm bone) of the patient to replace the articulating surface of the ulna. May be linked or unlinked. Uncemented See cementless. Unicondylar arthroplasty Replacement of one tibial condyle and one femoral condyle in the knee, with or without resurfacing of the patella. Unicondylar knee replacement See Unicondylar arthroplasty. Unilateral operation Operation performed on one side only, e.g. left hip. Unlinked total elbow Where the humeral and ulnar parts of a total elbow replacement are not physically connected.

238 www.njrcentre.org.uk National Joint Registry for England, Wales and Northern Ireland | 10th Annual Report

The National Joint Registry (NJR) produces this report using data The NJR shall have no liability (including but not limited to liability by reason of collected, collated and provided by third parties. As a result of this the negligence) for any loss, damage, cost or expense incurred or arising by reason NJR takes no responsibility for the accuracy, currency, reliability and of any person using or relying on the data within this report and whether caused correctness of any data used or referred to in this report, nor for the by reason of any error, omission or misrepresentation in the report or otherwise. accuracy, currency, reliability and correctness of links or references to other This report is not to be taken as advice. Third parties using or relying on the information sources and disclaims all warranties in relation to such data, data in this report do so at their own risk and will be responsible for making their links and references to the maximum extent permitted by legislation. own assessment and should verify all relevant representations, statements and information with their own professional advisers.

NJR Centre contact details

National Joint Registry based at Northgate Information Solutions (UK) Ltd Peoplebuilding 2 Peoplebuilding Estate Maylands Avenue Hemel Hempstead Hertfordshire HP2 4NW

Telephone: 0845 345 9991 Fax: 0845 345 9992

Email: [email protected] www.njrcentre.org.uk 239 Produced by Pad Creative Ltd www.padcreative.co.uk by Pad Creative Produced Website: www.njrcentre.org.uk www.njrcentre.org.uk

Every effort was made at the time of publication to ensure that the information contained in this report was accurate. If amendments or corrections are required after publicaton, they will be published on the NJR website at www. njrcentre.org.uk where the document is available to download in PDF format.