Professional Documents
Culture Documents
KNEES
ANKLES
ELBOWS
SHOULDERS
PROMs
2018
National Joint Registry
for England, Wales,
Northern Ireland and
the Isle of Man
ISSN 2054-1821 (Print) Surgical data to 31 December 2017
Prepared by
NJRSC Members
Mr Martyn Porter (Chairman, Editorial Board)
Mr Peter Howard
Sandra Lawrence
Professor Mike Reed
Jeffrey Stonadge
Professor Mark Wilkinson
Orthopaedic Specialists
Mr Richard Craig
Mr Colin Esler
Mr Andy Goldberg
Professor Jonathan Rees
This document is available in PDF format for download from the NJR Reports website at www.njrreports.org.uk. Additional data and
information can also be found as outlined on pages 18-19.
National Joint Registry | 15th Annual Report
Chairman’s Foreword
Laurel Powers-Freeling, National Joint Registry Chairman
www.njrcentre.org.uk 3
Work on development of NJR information and feedback to them all - Sue for her exemplary work as a patient
systems has included the reclassification and re- ‘champion’ and her support in raising awareness of the
structuring of the NJR’s component database. The patient perspective at many of our NJR events; and
NJR has collaborated with the German registry and Nick and Michael for their advice and input on behalf of
key stakeholders, including suppliers, on this important orthopaedic implant manufacturers.
area of work and reclassification of hip, knee and
shoulder devices will be completed in 2018/19. Also, Members who were granted extensions to their
enhancements to the NJR Clinician Feedback and membership of the NJRSC included: Mr Peter Howard
Management Feedback systems and Consultant Level and Professor Amar Rangan, surgeon members, until
Reports have ensured we continue to improve the way May 2020 and May 2021 respectively; Mr Rob Hurd,
the NJR provides information to our stakeholders. NHS Trust management representative until June 2020;
Professor Mark Wilkinson, public health/epidemiology
Future plans for the coming year representative to January 2022 and Gillian Coward,
patient representative to September 2022.
2018/19
Also Mr Martyn Porter, NJR Vice Chairman and Medical
In addition to our core schedule of activities and
Director, was granted an extension to his membership
continuing development of a number of work streams
until December 2018, when his final term will come to
described above, the NJR will:
an end. Work is now underway with the Department
• Work with Northgate Public Services and the of Health Appointments Team to recruit Martyn’s
University of Sheffield in the final security testing and successor as well as a new patient member to replace
live implementation of the Data Access Portal as a key Sue Musson.
mechanism for data access for NJR stakeholders.
New member appointments included Professor
• Work with the Universities of Sheffield and Bristol to Mike Reed as a surgeon representative, Professor
deliver the web-enabled version of a Patient Decision Karen Barker as Practitioner with Special Interest in
Aid tool to enable shared decision-making between Orthopaedics representative and Sandra Lawrence and
health professionals, patients and their families in Jeff Stonadge, as Orthopaedic Implant Manufacturer
considering hip or knee replacement. representatives. I would like to take the opportunity to
• Redevelop the NJR website to make it more engaging welcome all these new members to the NJRSC and
and easier to navigate, with a specific portal of useful look forward to working with them in the future.
information for patients and their carers.
I would also like to thank Ananda Nanu for his
• Implement the Healthcare Safety Investigation
considerable contribution this year as a co-opted
Branch (HSIB) report recommendations regarding
member of the NJRSC in his capacity as BOA
the implantation of wrong prostheses during joint
President, which has been significant to the NJR in
replacement surgery. The NJR welcomes the HSIB
continuing our valued relationship with the orthopaedic
report and is already working to take forward the
profession. I look forward to welcoming his successor
related action regarding the NJR database alert
who takes up post from September.
mechanism, to further enhance patient safety.
Also I would like to end by thanking all members of
Acknowledgements the NJRSC and sub-committees, for their valuable
contribution, support and enthusiasm. In particular in
During this reporting period, there have been a number
his final term, my special thanks to Mr Martyn Porter,
of changes to the NJRSC membership. Members
for his valuable contribution and unstinting dedication
leaving the NJRSC included patient representative
in undertaking the role of NJR Medical Director as the
member Sue Musson and industry representatives,
first appointee to this post. I have greatly appreciated
Nick Wishart and Michael Green. My sincere thanks
4 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Laurel Powers-Freeling
www.njrcentre.org.uk 5
Contents
Chairman’s Foreword 3
Executive Summary 14
Part 2 Clinical activity 2017 and using the dedicated NJR Reports website 20
2.1 Clinical activity 2017 overview 21
Data source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Linkage between primaries and any associated revisions (the ‘linked files’). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3.3 Revisions after primary hip replacement: effect of head size for selected bearing surfaces/fixation sub-groups . 61
3.3.4 Revisions after primary hip surgery for the main stem/cup brand combinations . . . . . . . . . . . . . . . . . . . . . . . 68
3.3.7 Primary hip replacement for fractured neck of femur compared with other reasons for implantation. . . . . . . . 82
6 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
3.4.3 Revisions after primary knee replacement surgery by main brands for TKR and UKR. . . . . . . . . . . . . . . . . . 124
3.4.4 Revisions for different clinical causes after primary knee replacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.6.3 PROMs Oxford Shoulder Scores (OSS) associated with primary shoulder replacement surgery. . . . . . . . . . 174
www.njrcentre.org.uk 7
3.8 In-depth studies 189
3.8.1 Heart failure after metal-on-metal hip replacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
3.8.2 Revision for Periprosthetic Femoral Fracture in total hip arthroplasty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
3.8.3 Trabecular metal acetabular components in primary and revision total hip arthroplasty. . . . . . . . . . . . . . . . .197
4.3 Outlier units for 90-day mortality and revision rates for the period 2003 to 2017 204
Glossary 210
Glossary 211
Part 3 tables
Table 3.1 Summary description of linked datasets used for main survivorship analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 3.3 Number and percentage of primary hip replacements by fixation and bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Table 3.4 Percentage of primary hip replacements by fixation, bearing and calendar year�������������������������������������������������������40
Table 3.7 KM estimates of cumulative revision (95% CI) by fixation and bearing, in primary hip replacements. . . . . . . . . . . . . 50
Table 3.8 KM estimates of cumulative revision (95% CI) of primary hip replacements by gender, age-group
fixation and bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Table 3.9 KM estimates of cumulative revision (95% CI) of primary hip replacement by fixation, and stem/cup brand . . . . . . . 68
Table 3.10 KM estimates of cumulative revision (95% CI) of primary hip replacement by fixation, stem/cup brand,
and bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Table 3.11 PTIR estimate of indications for hip revision (95% CI) by fixation and bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Table 3.12 PTIR estimate of indications for hip revision (95% CI) by years following primary hip replacement. . . . . . . . . . . . . . 76
Table 3.13 KM estimates of cumulative mortality (95% CI) by age and gender, in primary hip replacement . . . . . . . . . . . . . . . 81
Table 3.15 Fracture NOF vs. OA only by gender, age and fixation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
8 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Table 3.16 Number and percentage of hip revisions by procedure type and year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Table 3.17 Number and percentage of hip revision by indication and procedure type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Table 3.18 (b) KM estimates of cumulative re-revision (95% CI) by years since first failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Table 3.18 (c) KM estimates of cumulative re-revision (95% CI) by fixation and bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 3.20 (b) Number of re-revisions by year, stage, and whether or not primary is in NJR . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Table 3.21 Number and percentage of primary knee replacements by fixation, constraint and bearing . . . . . . . . . . . . . . . . . 105
Table 3.22 Percentage of primary knee replacements by fixation, constraint, bearing and calendar year. . . . . . . . . . . . . . . . 106
Table 3.23 Age at primary knee replacement by fixation, constraint and bearing type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Table 3.25 KM estimates of cumulative revision (95% CI) by fixation, constraint and bearing, in primary
knee replacements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Table 3.26 KM estimates of cumulative revision (95% CI) by gender, age, fixation, constraint and bearing,
in primary knee replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Table 3.27 KM estimates of cumulative revision (95% CI) by total knee replacement brands. . . . . . . . . . . . . . . . . . . . . . . . . 125
Table 3.28 KM estimates of cumulative revision (95% CI) by unicompartmental knee replacement brands . . . . . . . . . . . . . . 126
Table 3.29 KM estimates of cumulative revision (95% CI) by fixation, constraint and brand. . . . . . . . . . . . . . . . . . . . . . . . . . 127
Table 3.30 PTIR estimates of indications for revision (95% CI) by fixation, constraint and bearing type. . . . . . . . . . . . . . . . . .131
Table 3.31 PTIR estimates of indications for revision (95% CI) by years following primary knee replacement. . . . . . . . . . . . . 133
Table 3.32 KM estimates of cumulative mortality (95% CI) by age and gender, in primary knee replacement. . . . . . . . . . . . . 134
Table 3.33 Number and percentage of failures by procedure type and year. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Table 3.34 Number and percentage of knee revision by indication and procedure type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Table 3.35 (b) KM estimates of cumulative re-revision (95% CI) by years since first revision. . . . . . . . . . . . . . . . . . . . . . . . . . 146
Table 3.35 (c) KM estimates of cumulative re-revision (95% CI) by fixation and constraint. . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Table 3.37 (b) Number of re-revisions by year, stage, and whether or not primary is in NJR . . . . . . . . . . . . . . . . . . . . . . . . . 150
Table 3.38 Descriptive statistics of ankle procedures performed by consultant and unit by year of surgery . . . . . . . . . . . . . . 153
Table 3.39 Numbers (%) of primary ankle replacements by ankle brand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Table 3.40 KM estimates of revision (95% CI) after primary ankle replacement, by gender and age. . . . . . . . . . . . . . . . . . . . 155
Table 3.41 Indications for the 211 (first) revisions following primary ankle replacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
www.njrcentre.org.uk 9
Table 3.42 KM estimates of mortality (95% CI) after primary ankle replacement, by gender and age.. . . . . . . . . . . . . . . . . . . 157
Table 3.43 Numbers of primary shoulder replacements (elective and acute trauma), by year with percentages of each type. 159
Table 3.44 Numbers of units and consultant surgeons providing primary shoulder replacements over the last
five years, 2013-2017.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Table 3.45 Reasons for main types of primary shoulder replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Table 3.46 Gender and age at primary for the main types of primary shoulder replacements. . . . . . . . . . . . . . . . . . . . . . . . . 162
Table 3.47 Stemmed brands used for primary shoulder procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Table 3.48 Stemless brands and resurfacing brands used in primary shoulder replacements . . . . . . . . . . . . . . . . . . . . . . . . 166
Table 3.49 Glenoid brands used in total conventional shoulder replacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Table 3.50 KM estimates of cumulative revision (95% CI) for primary shoulder replacement for acute trauma
and elective cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Table 3.51 KM estimates of cumulative revision (95% CI) for elective shoulder primaries by gender and age. . . . . . . . . . . . . 169
Table 3.52 KM estimates of cumulative revision (95% CI) for elective shoulder primary by main type of procedure . . . . . . . . 171
Table 3.53 Numbers of first revisions for each type of primary shoulder replacement and indications for revision. . . . . . . . . . 172
Table 3.54 KM estimates of cumulative mortality (95% CI) for acute trauma and elective cases. . . . . . . . . . . . . . . . . . . . . . . 175
Table 3.55 KM estimates of cumulative mortality (95% CI) for elective cases by age and gender. . . . . . . . . . . . . . . . . . . . . . 176
Table 3.56 Number of primary elbow replacements by year and percentages of each type of procedure. . . . . . . . . . . . . . . . 179
Table 3.57 Types of primary elbow procedures used in acute trauma and elective cases by year . . . . . . . . . . . . . . . . . . . . . 180
Table 3.58 Reasons for main types of primary elbow replacements, by year of primary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Table 3.59 Number of units and consultant surgeons providing primary elbow replacements during each
year from 2015 to 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Table 3.61 Radial head brands used in radial head replacements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Table 3.62 KM estimates of cumulative revision (95% CI) by primary elbow procedures for acute trauma and
elective cases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Table 3.63 Indications for first data linked revision after any primary elbow replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Table 3.64 KM estimates of cumulative mortality (95% CIs) by time from primary elbow replacement, for acute
trauma and elective cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
10 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Table 3.65 Baseline characteristics of patients with implanted metal-on-metal hip prostheses and
non metal-on-metal controls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Table 3.66 Adjusted hazard ratio of heart failure for patients with implanted metal-on-metal hip prostheses and
non metal-on-metal controls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Table 3.67 Cumulative probability of revision at ten years for periprosthetic femoral fracture for most commonly
used stems in primary THA in four patient groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Part 3 figures
Figure 3.1 Initial numbers of procedures for analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Figure 3.4 (a) Cemented primary hip replacement bearing surface by year. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Figure 3.4 (b) Uncemented primary hip replacement bearing surface by year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Figure 3.4 (c) Hybrid primary hip replacement bearing surface by year. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 3.4 (d) Reverse hybrid primary hip replacement bearing surface by year. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 3.5 (a) KM estimates of cumulative revision by year, in primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 3.5 (b) KM estimates of cumulative revision by year, in primary hip replacements plotted by year of primary. . . . . . . . . 49
Figure 3.6 KM estimates of cumulative revision in cemented primary hip replacements by bearing . . . . . . . . . . . . . . . . . . . . . 51
Figure 3.7 KM estimates of cumulative revision in uncemented primary hip replacements by bearing . . . . . . . . . . . . . . . . . . . 52
Figure 3.8 KM estimates of cumulative revision in hybrid primary hip replacements by bearing . . . . . . . . . . . . . . . . . . . . . . . . 53
Figure 3.9 KM estimates of cumulative revision in reverse hybrid primary hip replacements by bearing. . . . . . . . . . . . . . . . . . 54
Figure 3.10 (a) KM estimates of cumulative revision in all primary hip replacements by gender and age . . . . . . . . . . . . . . . . . 55
Figure 3.10 (b) KM estimates of cumulative revision in all primary hip replacements by gender and age,
excluding MoM & resurfacing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Figure 3.11 (a) KM estimates of cumulative revision of primary cemented MoP hip replacement
(monobloc cups) by head size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Figure 3.11 (b) KM estimates of cumulative revision of primary uncemented MoP hip replacement
(metal shells & polyethylene liner) by head size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Figure 3.11 (c) KM estimates of cumulative revision of primary uncemented MoM hip replacement
(monobloc cups or metal shell & liner) by head size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Figure 3.11 (d) KM estimates of cumulative revision of primary cemented CoP hip replacement
(monobloc cups) by head size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Figure 3.11 (e) KM estimates of cumulative revision of primary uncemented CoP hip replacement
(metal shell & polyethylene liner) by head size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Figure 3.11 (f) KM estimates of cumulative revision of primary uncemented CoC hip replacement
(metal shell & ceramic liner) by head size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
www.njrcentre.org.uk 11
Figure 3.12 (a) PTIR estimates of aseptic loosening by fixation & bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Figure 3.12 (f) PTIR estimates of adverse soft tissue reaction by fixation & bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Figure 3.12 (g) PTIR estimates of adverse soft tissue reaction by fixation & bearing, since 2008. . . . . . . . . . . . . . . . . . . . . . . 80
Figure 3.13 KM estimates of cumulative revision for fractured NOF and OA only cases for primary hip replacements . . . . . . . 84
Figure 3.14 KM estimates of cumulative mortality for fractured NOF and OA only in primary hip replacements . . . . . . . . . . . . 85
Figure 3.15 (a) KM estimates of cumulative re-revision in linked primary hip replacements . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Figure 3.15 (b) KM estimates of cumulative re-revision by primary fixation in linked primary hip replacements. . . . . . . . . . . . . 89
Figure 3.15 (c) KM estimates of cumulative re-revision by years since first revision, in linked primary hip replacements. . . . . . 90
Figure 3.16 (a) KM estimates of cumulative re-revision in cemented primary hip replacement by years since first revision,
in linked primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Figure 3.16 (b) KM estimates of cumulative re-revision in uncemented primary hip replacement by years since
first revision, in linked primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Figure 3.16 (c) KM estimates of cumulative re-revision in hybrid primary hip replacement by years since first revision,
in linked primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Figure 3.16 (d) KM estimates of cumulative re-revision in reverse hybrid primary hip replacement by years since
first revision, in linked primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Figure 3.16 (e) KM estimates of cumulative re-revision in resurfacing primary hip replacement by years since
first revision, in linked primary hip replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Figure 3.19 (a) KM estimates of cumulative revision by year, in primary knee replacements . . . . . . . . . . . . . . . . . . . . . . . . . 110
Figure 3.19 (b) KM estimates of cumulative revision by year, in primary knee replacements plotted by year of primary. . . . . 111
Figure 3.20 (a) KM estimates of cumulative revision in primary total cemented knee replacements by constraint and bearing. . 113
Figure 3.20 (b) KM estimates of cumulative revision in primary total uncemented/hybrid knee replacements by
constraint and bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Figure 3.20 (c) KM estimates of cumulative revision in primary unicondylar or patellofemoral knee replacements
by constraint and bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Figure 3.21 (a) KM estimates of cumulative revision in primary total knee replacements by gender and age. . . . . . . . . . . . . 116
Figure 3.21 (b) KM estimates of cumulative revision in primary unicondylar knee replacements by gender and age. . . . . . . . 117
Figure 3.22 (a) KM estimates of cumulative re-revision, in linked revised primary knee replacements . . . . . . . . . . . . . . . . . . 138
Figure 3.22 (b) KM estimates of cumulative re-revision by primary fixation, in linked primary knee replacements. . . . . . . . . . 139
Figure 3.22 (c) KM estimates of cumulative re-revision by years since first revision, in linked primary knee replacements . . . 140
12 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Figure 3.23 (a) KM estimates of cumulative re-revision in primary cemented TKRs by years since first revision. . . . . . . . . . . 141
Figure 3.23 (b) KM estimates of cumulative re-revision in primary uncemented TKRs by years since first revision. . . . . . . . . 142
Figure 3.23 (c) KM estimates of cumulative re-revision in primary hybrid TKRs by years since first revision,
in linked primary knee replacements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Figure 3.23 (d) KM estimates of cumulative re-revision in primary patellofemoral knee replacements by years
since first revision. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Figure 3.23 (e) KM estimates of cumulative re-revision in primary unicondylar knee replacements by years
since first revision. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Figure 3.24 (a) Gender and age distribution of elective shoulder primaries for humeral hemiarthroplasty. . . . . . . . . . . . . . . . 163
Figure 3.24 (b) Gender and age distribution of elective shoulder primaries for total conventional shoulder replacement. . . . . 163
Figure 3.24 (c) Gender and age distribution of elective shoulder primaries for reverse polarity total shoulder replacement. . . 164
Figure 3.25 KM estimates of cumulative revision for primary shoulder replacement by acute trauma and elective cases. . . . 168
Figure 3.26 KM estimates of cumulative revision for primary shoulder replacement, by type of procedure
in elective cases only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Figure 3.27 KM estimates of cumulative revision after primary total prosthetic elbow replacement by acute trauma
and elective cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Figure 3.28 Temporal changes in incidence of revision for PFF and aseptic loosening. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Figure 3.29 Probability of revision for PFF using Flexible Parametric Competing Risks models in gender and
age-based groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Figure 3.30 Cumulative acetabular component survival rate following primary THA in TM and non-TM implants. . . . . . . . . . 198
Figure 3.31 Cumulative acetabular component survival rate following revision THA in TM and non-TM implants. . . . . . . . . . 199
www.njrcentre.org.uk 13
Executive Summary
Mr Martyn Porter, NJR Medical Director and Chairman of the Editorial Board
Within the year, the NJR has worked hard to develop The minimum dataset has also been updated this year.
its Accountability and Transparency Model. Twice One of the changes will now allow for the registration of
a year, the NJR carries out a structured analysis Debridement and Implant Retention (DAIR) procedures
of revision and post-operative mortality to identify used in treating infected joint replacements in the early
unexpected variations of both individual surgeons and periods following surgery as a separate procedure
hospitals. The NJR is now working with the BOA, CQC type, distinct from single stage revision under which
and NHS Improvement to ensure that the appropriate they were previously recorded.. To counter the
governance actions are carried out in a timely way. argument that this may encourage surgeons not to
The NJR’s symposium at the 2018 BOA Congress carry out exchange of the modular components,
entitled “How to use NJR data responsibly to improve simple debridement and washout will also be recorded.
care”, includes the use of data to potentially improve These changes will allow more accurate and detailed
outcomes as well as identifying causes for concern. assessment of these procedures in relation to later
revision surgery that may be required for deep infection.
14 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
The NJR component database is in the process of in outcomes. In particular, it is the young patient who
being redesigned to allow for more detailed analysis has a much higher revision risk compared to more
of the types of components that are implanted. This elderly patients. For example, for hip replacement in
should lead to a better understanding of how the male patients under the age of 55 years at the time
attributes of devices may affect their performance. of surgery, overall, the revision estimate is nearly 16%
at 14 years compared to just 2.5% in male patients
The NJR continues to work closely and has over 75 years. A similar pattern is evident in knee
strengthened its relationships with GIRFT, ODEP and replacement with revision estimates of nearly 15% at 14
Beyond Compliance. The NJR also collaborates with years in female patients less than 55 years at the time of
other international joint replacement registries and is surgery, compared to just 2.5% in female patients over
part of ISAR - the International Society for Arthroplasty 75 years.
Registers. The 2018 symposium was held in Reykjavik,
Iceland in May 2018. The NJR was also represented This does not mean that young patients should be
at the EFORT meeting (European Federation of denied joint replacement surgery just because they
Orthopaedics and Traumatology) in Barcelona in June are young and have a high risk of revision, but it
2018. The focus of this international work has been does mean that both surgeons and patients should
on collaboration, sharing registry research findings, be aware of these facts in order to facilitate ‘shared
developing new methodologies, discussing the utility decision-making’. In this regard the development of
of PROMs (Patient Reported Outcome Measures), an NJR ‘risk calculator’ by Mark Wilkinson and his
early signal detection and benchmarking of devices. research team in Sheffield may prove to be timely. Risk
calculators are not a new concept but the NJR model
Main headlines for 2017: Outcomes is nearing the end of its development and should be
available for clinical testing early in 2019. The aim
The data builds on last year’s findings and the main is to be able to enter individual patient information
message of the report is that revision estimates such as age, gender, BMI etc. and for the tool to
following joint replacement surgery remain low. predict the likelihood for revision, mortality or patient
reported outcome measures (PROMs) in a ‘patient
The data is structured to show the effect of patient
like them’. The science and methodology behind this
and implant factors on revision estimates. For
is complex and the current indications are that the
example, patient factors include gender and age at
predictive power is at the best modest but certainly
time of surgery, while implant factors include type of
an improvement on the current general advice on risk
fixation, brand, bearing and head size. The debate has
given to patients. It is envisaged that these tools will
moved on from a comparison of fixation method to
be used in the outpatient setting by the surgeon and
one of comparison of joint replacement ‘constructs’.
patient to facilitate ‘shared decision-making’.
In a hip replacement, for example, a construct
would include a description of the brand of socket, In relation to knee replacement surgery, revision rates
femoral stem and bearing material. It is the ‘whole are higher for partial knee replacement, compared
replacement’ that is important to the patient rather to total knee replacement. However, other data
than the individual components or method of fixation. demonstrates that partial replacements have lower
mortality rates and fewer complications than total knee
Some hip constructs have revision estimates as low as
replacement. The effect of surgeon volume is also
2% at 14 years and some knee brands have revision
important. The NJR’s role is not to discourage partial
estimates as low as 3% at 14 years. These results are
knee replacement surgery, but it would be advisable
quite remarkable in terms of outcomes for patients
to consider the implications of these other factors and
and value for the taxpayer.
perhaps limit surgical activity to surgeons who have
Although there is variation in revision estimates for sufficient volume and can achieve low revision rates.
implants, the patient continues to be a major factor
www.njrcentre.org.uk 15
Much work has been carried out on the classification Members of Data Access Review Group
of components used in shoulder replacement
surgery. This is not a straightforward matter as the Members of the NJR Patient Network
plethora of implants available and the modes in Other organisations:
which they combine are complex. The section on
MHRA
shoulder replacement is very impressive thanks to this
classification work. Revision estimates for all elective CQC
shoulder replacements in male patients over 75 years NHS England
are approximately double those for female patients in NHS Improvement
the same age group. British Orthopaedic Association
There are over 700 ankle replacement procedures BHS
carried out per year and revision estimates are just BASK
under 9% at seven years for all cases, but as for hips BESS
and knees, there is considerable variation according to
BOFAS
the gender and age of the patient.
HQIP
Concluding acknowledgements Northgate Public Services (UK) Ltd
University of Bristol
As mentioned earlier there is considerable additional
University of Oxford
information available online and I would encourage
you to explore the NJR’s dedicated annual report On a personal note I would particularly like to thank
website at www.njrreports.org.uk. The website Laurel Powers-Freeling, Chairman of the NJR and
offers a helpful interactive platform for Part Two of the Elaine Young, NJR Director of Operations.
report, which is the descriptive NJR data; supporting
appendices; and, when published, the latest NJR Northgate Public Services and University of Bristol
Patient and Public Guides to the annual report. teams have, yet again, done an exceptional job.
The NJR continues to work with many stakeholders; Finally, having served on the NJR Steering Committee
the most important, of course, are the patients whom I for over 15 years I will be standing down from the NJR
would like thank for allowing the NJR to use their data. later this year. I have very much enjoyed and learnt
from the experience. Thank you.
Many thanks also to the following without which the
NJR could not function:
16 www.njrcentre.org.uk
Part 1
Annual progress
1.1 Annual Report • Linkability (the ability to link a patient’s primary
procedure to a revision procedure) was recorded as
18 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
www.njrcentre.org.uk 19
Part 2
Clinical activity
2017 and using
the dedicated NJR
Reports website
National Joint Registry | 15th Annual Report
www.njrcentre.org.uk 21
2.2 Navigating the NJR Reports online facility
What can you find at NJR Reports online?
As at 31 March 2018, the total number of procedures recorded in the NJR is now approximately
2.52 million.
The NJR has refreshed its dedicated online annual report website – NJR Reports – to showcase annual report data
and help users easily navigate the growing wealth of information collected about joint replacement procedures.
Part Two of the NJR’s 15th Annual Report presents data on clinical activity during the 2017 calendar year.
Simply navigate the left hand tabs to view information on the volumes and surgical techniques in relation to
procedures submitted to the NJR.
22 www.njrcentre.org.uk
Part 3
Outcomes after
joint replacement
2003 to 2017
3.1 Executive
summary
Part Three of the 15th Annual Report provides performed in 2017. Metal-on-polyethylene bearings
outcome data in relation to hip, knee, ankle, shoulder were still the most widely used but the trend for
and elbow replacements. It describes activity between an increasing number of ceramic-on-polyethylene
1 April 2003 and 31 December 2017. implants continues.
There were 2,526,601 procedures entered in to the Within the entire primary hip dataset available for
NJR across all joint types. After removing procedures analysis of 992,090, the commonest indications
without linkage identifiers (217,484) and those with data for revision remain aseptic loosening, dislocation
errors that hindered linkage (167), 2,308,950 remained. (instability), adverse reaction to particulate debris, pain,
The majority of these exclusions are associated with infection and periprosthetic fracture. The data now
the early years of the registry, where insufficient patient available in the NJR allows for estimation of cumulative
details were collected to enable linkage. revision rates out to 14 years. For the entire cohort
the revision rate at 14 years is 7.27%. Over the same
There were 992,090 primary total hip replacements,
timeframe the revision rate for all cemented constructs
1,087,611 knee replacements, 4,687 ankle
is 4.88%, 8.94% for all uncemented and 5.38% for
replacements, 30,720 shoulder replacements and
hybrid. These findings are influenced by inclusion of
2,872 elbow replacements available for analysis.
patients with metal-on-metal bearings, which for all
categories of fixation have a much higher revision
Hip replacement procedures rate (19.06% – 22.21%). However, it is striking that
The total number of hip replacements recorded in the different combinations of fixation and bearing surface
NJR continues to increase totalling just under one can produce very low failure rates at 14 years, with
million replacements since data was first collected in cemented ceramic-on-polyethylene bearing failure
2003. More women (60%) have undergone surgery rates reported as 3.77% and hybrid ceramic-on-
than men (40%). The mean age at implantation across ceramic 3.52% (with limited data). The most common
the whole group is 68.0, but varies with fixation construct recorded in the NJR is cemented metal-on-
method with all uncemented as the most common polyethylene, which records a reassuringly low failure
fixation method recorded, having a mean age of 64.6, rate of 4.87% at 14 years.
hybrid 69.0 and all cemented 72.8.
Metal-on-metal resurfacings have a revision rate at 14
During 2017, 91,698 hip replacements were entered years of 14.76% and their use, as with all metal-on-
into the NJR which was a slight reduction on the metal bearings, is now very limited.
number performed in 2016. This may represent delayed
Some of the most important data presented in this
data entry for the period by some units, a phenomenon
year’s hip report relate to patient age at implantation
that has also been observed in previous years.
and its relationship to the outcome of different
combinations of fixation and bearing type.
The most commonly performed
hip replacement continues to be all This year’s analysis confirms that
uncemented , but in 2017, for the first across the whole cohort revision
time, the total number of hybrid hip rates increase as age of implantation
replacements implanted was greater decreases, with the effect more
than cemented hip replacements. pronounced in women.
The number of resurfacing metal-on-metal hip In men under the age of 55, hybrid ceramic-
replacements continues to decrease, contributing on-polyethylene and ceramic-on-ceramic hip
only 0.6% of the total number of hip replacements replacements have revision rates of under 4% at
24 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
12 years, less than half the 9.98% seen with The number of hip replacements recorded in the NJR
cemented metal-on-polyethylene constructs which are performed in patients with fractured neck
although, again, limited data suggests some caution of femur continues to increase, with 4,445 entered in
should be employed in interpreting this finding. In 2017. The yearly revision rate of this cohort remains
women under 55 years, hybrid ceramic-on-ceramic very similar to those performed for osteoarthritis after
constructs give the lowest revision rate of 4.11% at the first three months following surgery, with only
12 years and cemented ceramic–on-polyethylene minor elevation in absolute risk resulting from an
gives good results with failure rates of approximately increased risk of failure in the first three months after
4.93% at the same time point. However changing surgery, despite higher mortality rates.
the bearing type in all cemented fixation to metal-
on-polyethylene increases failure rate to 8.16% in To study re-revision of hip replacements, 92,348
women. These data represent important results and records of first revision were available for analysis
can help decision making when selecting constructs leading to 8,528 subsequent re-revisions. The risk of
for the younger patient. However, it is important to revision was strongly related to the time to first revision
recognize the possibility of selection bias with existing surgery and 11.66% of hips revised within a year of the
constructs that may confound this relationship. primary procedure are re-revised within three years.
For patients over the age of 75, the situation is less Knee replacement procedures
complex as all combinations, other than metal-on-
metal bearings, have similarly reliable outcomes. The Within the whole registry there were 1,087,611
lowest failure rates are seen with cemented or hybrid verifiable primary knee replacements recorded
ceramic-on-polyethylene constructs. between 1 April 2003 and 31 December 2017, with
maximum follow-up of 14.75 years. Across the
whole registry the overwhelming majority (97%) of
In addition, combining revision and procedures have been performed for osteoarthritis,
with more women undergoing surgery then men and a
mortality has identified that for the
mean age at implantation of 68.9 (SD 9.6).
majority of patients undergoing hip
replacement over the age of 75, the In 2017, 102,177 primary knee procedures were
replacement will last them for the entered into the NJR, a marginal decrease on the
number performed in 2016. There may be delayed
rest of their lives. data entry for the period by some units, a phenomenon
that has also been observed in previous years. Within
The analysis of head size (bearing diameter) provides a this group, 89% were total knee replacements, 10%
mixed picture with respect to performance. The largest unicondylar procedures and 1% patellofemoral.
head sizes in each class of cemented metal-on-
polyethylene, of uncemented metal-on-polyethylene, Since 2003, cemented TKR has
and cemented ceramic-on-polyethylene tend to
perform poorly. However, the absolute effect of head
remained the most widely used
size is less consistent e.g. a 36mm cemented metal- knee replacement construct in the
on-polyethylene head is the worst in class whereas a UK, representing 85% of all primary
36mm uncemented metal-on-polyethylene head has prostheses recorded in the registry.
comparative performance to the most widely used
bearing size of 28mm. Conversely, a 28mm ceramic-
Over time the total numbers of uncemented or hybrid
on-ceramic bearing tends to have worse performance
total knee implants used has declined, reaching a
than larger bearings therefore, the choice of bearing
current level of 2.2% of all primary knee replacements.
size needs to be carefully considered.
www.njrcentre.org.uk 25
In 2017 of all knee replacements, 63% were cemented However it is reassuring to note that revision is
unconstrained prostheses and 23% cemented still relatively uncommon and for patients around
posterior stabilised, with mobile bearing prosthesis the average age of undergoing surgery, their knee
less commonly used. replacement is likely to last the rest of their life without
the need for revision.
With a maximum follow-up of nearly 15 years the
registry now provides an insight into knee implant Over the last four years, the total number of new
survival into the second decade. The Kaplan-Meier revision operations recorded in the registry has
estimate of revision rate at 14 years for all cemented remained fairly static, with 6,289 entered into the
TKRs is 4.47%. The revision rate for unconstrained registry for 2017. Over 80% of these cases were single
fixed bearing devices (4.07%) was slightly lower than stage procedures, with the most common indications
that recorded for posterior stabilised components for surgery recorded as aseptic loosening, other
(5.39%). The lowest revision rates for cemented TKR indication, instability and pain.
are seen when monobloc tibial components are used,
although again limited data suggests some caution The collection of data over a 14 year time frame
should be employed in interpreting this finding. allows the results of re-revision to be recorded.
The 2017 NJR analysis demonstrates that the risk
Over the last five years there has been an increase in of re-revision following a first revision is higher than
the number of tibio-femoral unicondylar prostheses the risk of revision after a primary procedure. In
entered into the registry, reaching 10% for the first addition the rate of re-revision after
time in 2017. Over the last three years, 809 surgeons unicompartmental knee replacement is broadly
in 358 units now perform this operation. However similar to re-revision after total knee replacement.
revision rates for unicompartmental knee replacement
remain higher when compared to TKR reaching 16.9% Ankle replacement procedures
at 14 years.
A total of 4,687 primary ankle replacements have been
The use of patellofemoral prostheses represented 1% recorded on the NJR between 1 April 2010 and 31
of all primary knee replacements recorded in 2017. This December 2017.
type of knee replacement records the highest revision
rate at 14 years, over 24.4%, which may explain the Within each year of data collection there has been
trend for reduced usage over the last five years. an increase in the number of primary procedures
performed and this trend continued with 734 recorded
in 2017. They were performed by 135 consultants
Age continues to have a significant working in 139 units with a median number of 3
effect on outcome after primary performed per surgeon (IQR 1-8) and per unit (IQR
knee replacement, where younger 1-6). Over the eight years data has been collected, a
total of 69% of consultant surgeons and 74% of units
patients experience a higher have submitted less than 20 procedures in total.
revision rate. This important
observation is most vividly seen in The median age at primary surgery remains at 68 (IQR
61-74) years and 59% of procedures were carried out
patients under the age of 55 and in
in men.
those undergoing unicompartmental
knee replacement. Uncemented fixation has been used in the majority
of implantations recorded (98.2%). Infinity has rapidly
26 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
become the most common prosthesis used in ankle prosthesis for elective indications, followed by total
replacement since its introduction in 2014. shoulder arthroplasty and humeral hemiarthroplasty.
For use in trauma, reverse polarity total shoulder
The 7-year cumulative revision rate for the entire arthroplasty is also the most commonly recorded
cohort is 8.70%. However, in younger patients under implant, with humeral hemiarthroplasty also
65 years, the revision rate is greater than 10%, widely used.
although again, limited data suggests some caution
should be employed in interpreting this finding. Within the entire cohort, there have been 794
revisions of elective procedures and 66 revisions of
Shoulder replacement procedures trauma procedures.
www.njrcentre.org.uk 27
Of the 2,872 elbow replacements within the NJR,
1,938 have been used for elective care. In this group,
total prosthetic replacement has been used in 1,772
(91.4%) cases, treating mainly inflammatory arthritis
(50.6%), osteoarthritis (34.2%) or the sequelae of
trauma (14.6%). In 934 cases elbow replacement has
been used to treat acute trauma, with total prosthetic
replacement (48.9%), radial head replacement
(40.9%) or humeral hemiarthroplasty (8.6%) the three
most commonly used procedure types.
28 www.njrcentre.org.uk
Part 3
3.2 Summary
of data sources,
linkage and
methodology
The main outcome analyses in this section relate outcomes looking better or worse than they actually
to primary and revision joint replacements, unless are. This issue is being addressed by the NJR’s Data
otherwise indicated. We included all patients with Quality Sub-committee. Similarly, the 2014/15 audit
at least one primary joint replacement carried out suggested 9.7% and 9.8% of hip and knee revisions
between 1 April 2003 and 31 December 2017 respectively had been missed during this period. It is
inclusive, whose records had been submitted to the important for all those concerned with and involved with
NJR by 16 February 2018. the NJR to remember that data reporting of all relevant
procedures is mandated by the Department of Health.
Information governance and patient
As of June 2018, 100% of NHS Trusts and Health
confidentiality: Boards contributing data to the NJR had completed
NJR data are collected via a web-based data entry the audit. Although it is possible that some records
application and stored and processed in Northgate may have been missed in the audit process, or
Public Services’ (NPS) data centre. NPS is ISO 27001 subsequently entered, we believe this number is small.
and ISO 9001 accredited, and compliant with the
Whilst the proportion of missing data in the NJR
NHS’s Information Governance Toolkit. Data linkage
is relatively small, the propensity to not record
to other datasets is approved by the Health Research
revision procedures is problematic and will lead to a
Authority under Section 251 of the NHS Act 2006.
reduction in ability to detect trends. From a registry-
Please visit www.hra.nhs.uk/about-the-hra/our-
wide perspective, we believe under-reporting of
committees/section-251 for more details.
revisions would apply across all types of hip and knee
replacements in a random pattern, and therefore
Data source:
would not affect the group comparisons we make.
We know that in the early years of the NJR, when
reporting was not mandated by the Department of Patient level data linkage:
Health, a number of primary procedures were not
Documentation of implant survivorship and mortality
recorded in the NJR, as indicated by discrepancies
requires linkage of person-level identifiers, in order to
between implant levies and procedure rates. In the
identify primary and revision procedures and mortality
subsequent years, selective reporting of primary
events within the same individual.
and revision procedures may partly explain temporal
increases in volume (primary and revision), and Starting with a total of 2,526,601 NJR source records,
revision outcomes for hips and knees replacements 8.6% were lost because no suitable person-level
(see sections 3.3 and 3.4). identifier was found (see Figure 3.1). In around half
of these 217,484 procedures (48.4%), the patient
More recently primary procedures are less likely to
had declined to give consent for their details to be
have been missed. The recent 2014/15 NJR data
held or consent was not obtained, the remainder
completeness and accuracy audit across 149 NHS
being attributable to tracing and linkage difficulties.
trusts reporting to the NJR suggests that about 5.6%
Cases from Northern Ireland were excluded at this
and 5.1% of hip and knee primaries respectively may
step (22,609) because there was no tracing service
not have been recorded on the NJR.
available for them. Patients with longer follow-up might
Our analyses would be more seriously impacted by be less representative of the whole cohort of patients
differential and selective under-reporting of revision undergoing primary joint replacement than those
procedures associated with the primaries that have patients with shorter follow-up, due to difficulties with
been entered. This could lead to reported revision data linkage and differential rates of reporting over time.
30 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Among the linkable procedures with person-level Implant survivorship is first described with respect
identifiers (2,308,950) there were 96,015 (4.2%) to the lifetime of the primary joint only. In sections
revision procedures within the analysis period (2003 to 3.3 and 3.4, we also provide an overview of further
2017) with no associated primary operation recorded revisions following the first hip or knee revision
in the NJR. This would have been either because the procedure. We have also included revisions to a joint
primary had taken place at an earlier point in time replacement where the associated primary had not
(before the NJR data collection period began in 2003) been documented in the NJR.
or was not included for other reasons such as the
operation being performed outside the geographical As in previous years, the unit of observation for all
catchment area of the NJR, or consent for data sets of survivorship analysis has been taken as the
linkage not being provided at the time of the primary individual primary joint replacement. A patient with left
procedure. At the joint level, some further revisions and right replacements of a particular type, therefore,
were excluded because they could not be matched to will have two entries, and an assumption is made
primary joint replacements, i.e. if a primary operation that the survivorship of a replacement on one side
was recorded only for one side and there was only a is independent of the other. In practice, this would
documented revision for the other side, the latter was be difficult to validate, particularly given that some
excluded. However, we have included these ‘unlinked’ patients will have had primary replacements of other
revisions in our general overview of outcomes after joints that were not recorded in the NJR. Established
revision, see Sections 3.3 and 3.4. risk factors, such as age, are recorded at the time of
primary operation and will therefore be different for
Linkage between primaries and any the two procedures unless the two operations are
performed at the same time.
associated revisions (the ‘linked files’):
Within the NJR, a revision is defined as any operation
A total of 1,733,867 patients had at least one record
in which any prosthesis or part of a prosthesis is
of a primary joint replacement within the NJR, i.e.
either removed, exchanged or inserted for any
hip, knee, ankle, shoulder or elbow. At this stage,
reason into a joint in which there is an existing
information about the primary procedures was linked
joint replacement. This therefore not only includes
to subsequent associated revisions (i.e. for the same
complete replacement of one or both of the main
patient-joint-side). Further data cleaning was carried
components of any joint replacement, but also, for
out at this stage, for example, removal of duplicated
example, liner and/or head exchange at surgery
primary information on the same side or revision dates
for suspected infection and secondary patella
that appeared to precede the primary procedure,
resurfacing of an existing total knee replacement.
leading to the final numbers for analysis shown in
Tables 3.1 and 3.2.
www.njrcentre.org.uk 31
Figure 3.1 Initial numbers of procedures for analysis.
linkable procedures
(1 missing date; 1 with unknown operation;
87 with primary prior to 1 April 2003;
78 ‘deaths before procedure’)
2,308,950
linkable procedures
1,733,867
patient identifiers
Table 3.1 Summary description of linked datasets used for main survivorship analyses.
Time period
1 April 2010* - 31 December 2017 (ankles)
1 April 2012* - 31 December 2017 (shoulders)
Excludes data where person-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
992,218 1,087,696 4,687 30,720 2,872
Number of primary operations
hips knees ankles shoulders elbows
NJR identified primary-linked first revisions
Number of primaries that were
27,608 28,717 211** 860*** 80****
subsequently revised
hips knees ankles shoulders elbows
*These were the dates when data collection formally started, however the analyses in this section include a small number of primaries in the database that took
place before these time points.
**Ankle revisions include 29 conversions to arthrodesis.
***Shoulder revisions include four excisions and two conversions to arthrodesis.
****Elbow revisions includes three excisions.
32 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Joints
Hips Knees Ankles Shoulders Elbows
Number of patients 848,970 884,940 4,486 28,774 2,776
Number (%) of patients with only 705,722 682,184 4,285 26,828 2,680
*Discussed more fully in later sections: the numbers shown include some stage two of two-stage revisions.
**In some cases the first revision was the stage one of a two-stage revision; the numbers in parenthesis exclude cases where subsequent revision procedures
appeared to relate only to that first (i.e. either were another stage one or the respective stage two).
www.njrcentre.org.uk 33
Terminology note: Hip replacements of the bearing and whether the implants are of a
modular design.
There are four distinctive design features reflected
in the analysis of data collected in the NJR and they The knee is made up of three compartments:
include: 1) The type of hip replacement i.e. total hip medial, lateral and patellofemoral. When a total knee
replacements (THR) and hip resurfacings (the NJR does replacement (TKR) is implanted, the medial and
not collect data on hemiarthroplasty); 2) the fixation of lateral compartments are always replaced, and the
the replacement i.e. cemented, uncemented, hybrid patella may be resurfaced. If a single compartment
and reverse hybrid; 3) the bearing surfaces of the is replaced then the term unicompartmental is
hip replacement; 4) the size of femoral head/internal applied to the implant (UKR). The medial, lateral or
diameter of the acetabular bearing. patellofemoral compartments can all be replaced
independently, if clinically appropriate. Medial and
Cemented constructs are fixed using bone
lateral unicompartmental knee replacements are also
cement in both the femoral stem and acetabulum.
referred to as medial or lateral unicondylar
Uncemented constructs rely on press fit and osseous
knee replacements.
integration within the femur and acetabulum that
may be supplemented (e.g. by screw fixation). Hybrid Knee replacements are also characterised by their
constructs contain a cemented femoral stem and an level of constraint (stabilisation). For example, there is
uncemented acetabulum. Reverse hybrid constructs variation in the constraint of the tibial insert’s articulation
contain an uncemented femoral stem and a cemented with the femoral component depending on whether
acetabulum. By convention, the bearing material of the posterior cruciate ligament is preserved (cruciate
the femoral head is listed before the acetabulum. retaining; CR) or sacrificed (posterior stabilised; PS)
Currently, the six main categories of bearing surfaces at the time of surgery. Additional constraint may be
for hip replacements are ceramic-on-ceramic (CoC), necessary to allow the implant to deal with additional
ceramic-on-metal (CoM), ceramic-on-polyethylene ligament deficiency or bone loss (where constrained
(CoP), metal-on-metal (MoM), metal-on-polyethylene condylar (CCK) or hinged knee implants would be used)
(MoP) and resurfacing procedures. The metal-on- in a primary or revision procedure.
metal group in this section refers to patients with a
stemmed prosthesis (THR) and metal bearing surfaces In modular tibial components, the tibial insert may
(a monobloc metal acetabular cup or a metal acetabular be mobile or remain in a fixed position on the
cup with a metal liner). Although they have metal- tibial tray. This also applies to medial and lateral
on-metal bearing surfaces, resurfacing procedures, unicompartmental knees. Many brands of total knee
which have a surface replacement femoral prosthesis implant exist in fixed and mobile forms with options for
combined with a metal acetabular cup, are treated as either CR or PS constraint.
a separate category. Ceramic-on-ceramic resurfacings
are now being implanted and in future reports, these will Tibial elements may or may not be modular
be reported as a new category although the numbers design. Modularity allows some degree of patient
are likely to remain too small for meaningful analysis specific customisation. For example, modular tibial
for a number of years. The size of the femoral head is components are typically composed of a metal tibial
expressed in millimetres. tray and a polyethylene insert which may vary in
thickness. Non-modular tibial components consist
Terminology note: Knee replacements of an all-polyethylene tibial component (monobloc
Knee replacements within the NJR are principally polyethylene tibia) available in different thicknesses.
defined by the number and type of compartments
The NJR now distinguishes between medial and
replaced, the fixation of the components (cemented,
lateral unicondylar knee replacements during the
uncemented or hybrid), level of constraint, the mobility
data collection process; however this was not so in
34 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
earlier versions of the minimum dataset form (MDS). In This assumes that death is unrelated to a failing
addition, there are other possible knee designs, such implant, and can be safely ignored whilst estimating
as combinations of unicondylar and patellofemoral, but implant failure (revision). See Sayers et al. 2018
these are not reported on here, as the numbers are Acta Orthopaedica, 89:3, 256-258, for an extensive
too small. discussion on this problem.
With regard to the use of the word ‘constraint’ The survival tables in this report show ‘Kaplan-Meier’
here, for brevity, total knee replacements are termed (KM) estimates of the cumulative chance (probability)
unconstrained (instead of posterior cruciate-retaining) of failure (revision) or death, at different times from the
or posterior-stabilised (instead of posterior primary operation. In the joint replacement literature
cruciate-stabilised). they are often referred to as KM or simply survival
estimates. We additionally show 95% Confidence
Descriptive statistics Intervals for each estimate (95% CI). Confidence
In simple cases we tend to report simple descriptive intervals illustrate the uncertainty around the estimate,
statistics including; frequencies (N=), percentages with wide confidence intervals indicating greater
(%), minimums (min), maximums (max), inter-quartile uncertainty than narrow ones. Strictly they are
ranges (IQR) (25th centile, 75th centile), means (SD) interpreted in the context of repeated sampling i.e. if
and medians (50th centile) of the data. the data were collected in repeated samples we would
expect 95% CIs generated to contain the true estimate
Survival analysis methods in 95% of samples. However, confidence intervals
In more complex analyses that focus on either implant are strongly influenced by the numbers of prosthesis
failure (denoted revision), recurrent implant failure (re- constructs at risk and can become unreliable when the
revision) or mortality we use ‘survival analysis methods’ numbers at risk become low. In tables, we highlight in
which are also known as ‘time to event’ methods. blue italics all estimates where there are less than 250
prosthesis constructs at risk or remaining at risk at
Survival analysis methods are necessary in joint that particular time point.
replacement data due to a process known as
‘censoring’. There are two forms of censoring which Kaplan-Meier estimates can also be displayed
are important to consider in joint replacement registry graphically using a connected line plot. Figures are
data: administrative censoring and censoring due to joined using a ‘stair-step’ function. Each ‘stair’ is flat,
events, such as death. reflecting the constant nature of the estimate between
the events of interest. When a new event occurs the
Administrative censoring creates differential amounts survival estimate changes, creating a ‘step’. Changes
of follow-up time, i.e. patients from 2003 will have in the numbers at risk because of censoring do not
been followed up for more than ten years, whilst themselves cause a step change but if the numbers
patients collected last year will have one year of at risk become low, when an event does occur, the
follow-up or less. Survival analyses methods allow us stair-step might appear quite dramatic. Whenever
to include all patients in one analysis without being possible, the numbers at risk at each time point have
concerned if patients have one day, one year or one been included in the figures, allowing the reader to
decade of observed follow-up time; these methods more appropriately interpret the data given the number
automatically adjust analyses for the amount of of constructs at risk. The Kaplan-Meier estimates
follow-up time. shown are technically 1 minus the Kaplan-Meier
estimate multiplied by 100, therefore they estimate the
In the case of analyses which estimate implant failure, cumulative percentage probability of construct failure.
death events are also censored, specifically they
are considered non-informative censoring events.
www.njrcentre.org.uk 35
In the case of revisions, no attempt has been made show the numbers of revisions per 1,000 years at
to adjust for the risk of death, as analyses attempt risk. PTIR in other areas of research are often known
to estimate the underlying implant failure rate in as ‘person-time’ incident rates, however, in joint
the absence of death, see Sayers et al. 2018 Acta replacement registers the base unit of analysis is the
Orthopaedica, 89:3, 256-258 for an extensive ‘prosthesis construct’.
discussion on competing risks. Briefly, the Kaplan-
Meier estimator, estimates the probability of implant Note: This method is only appropriate if the hazard
failure (revision) assuming the patient is still alive. rate (the rate at which revisions occur in the unrevised
cases) remains constant across the follow-up period.
Prosthesis (construct) Time Incidence Rates - PTIRs The latter is further explored by sub-dividing the time
interval from the primary operation into intervals and
Prosthesis time incidence rates are used to describe
calculating PTIRs for each interval. We have explored
the incidence (the rate of new events) of specific
temporal changes for hips and knees in this report.
modes of failure in joint replacement. The PTIR
expresses the number of revisions divided by the total
of the individual prosthesis-years at risk. Figures here
36 www.njrcentre.org.uk
Part 3
3.3 Outcomes after
hip replacement
This section looks at revision and mortality outcomes Details of the patient cohort are given in Tables 3.1
for all primary hip operations performed between and 3.2 of the preceding section. Figure 3.2 describes
1 April 2003 and 31 December 2017. Patients the data cleaning applied to produce the total of
operated on at the beginning of the registry therefore 992,090 hips included in our analyses.
have a potential 14.75 years of follow-up.
996,272
Initial hip primaries in NJR
3,823
Excluded duplicate
primary procedures
992,449
© National Joint Registry 2018
992,218
Primary hip replacements with
revision date that did not
precede the primary date 128
Excluded records where
it was not possible to
trace the NHS number.
Also excludes unknown
or missing gender
Within the whole registry, the 992,090 primary hip have just qualified over this period, some may have
replacement procedures contributing to our analyses retired, and some surgeons may have periods of
were carried out by a total of 3,465 consultant inactivity within the coverage of the NJR, therefore
surgeons working across 471 units. Over the last their apparent caseload would be lower.
three years (1 January 2015 to 31 December 2017),
272,496 primary hip procedures (representing 27.5% The majority of hip primary procedures were carried out
of the current registry) were performed by 2,204 on women (females 59.8%: males 40.2%). The median
consultant surgeons working across 415 units. age at primary operation was 69 (IQR 61-76) years and
Looking at caseload over this three-year period, the the overall range was 7-105 years. Osteoarthritis was
median number of primary procedures per consultant given as a documented indication in 911,854 (91.9% of
surgeon was 59 (inter-quartile range (IQR) 4-193.5) the cohort) and was the sole reason given in 880,649
and the median number of procedures per unit was (88.8%) primary hip replacements.
564 (IQR 285-884). A proportion of consultants will
38 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
www.njrcentre.org.uk 39
40
Table 3.4 Percentage* of primary hip replacements by fixation, bearing and calendar year.
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Fixation/bearing n=42,769 n=40,695 n=48,587 n=60,947 n=67,399 n=68,430 n=70,931 n=73,888 n=78,126 n=80,226 n=87,596 n=88,845 n=91,953 n=91,698
All cemented 56.1 48.3 42.4 39.4 34 31.7 31.2 32.2 32.8 33 31.9 30.8 29.4 28.2
Cemented by bearing surface:
MoP 51.2 43.7 38.1 35.5 30.1 28.1 27.1 27.5 28.4 28.3 27 25.8 24.4 23
MoM 0.3 0.4 0.4 0.4 0.4 0.1 <0.1(31) <0.1(9) 0 0 0 <0.1(4) <0.1(1) <0.1(1)
CoP 3 3 2.9 2.5 2.7 2.8 3.2 3.5 4 4.4 4.6 4.8 4.9 5.1
Others/unsure 1.6 1.2 1 1.1 0.8 0.6 0.9 1.2 0.5 0.4 0.3 0.2 0.2 0.1
All uncemented 19.9 25.8 30.2 33.4 39.4 43.2 45.8 45 44.9 42.6 40.9 39.4 38.6 37.8
www.njrcentre.org.uk
Uncemented by bearing surface:
MoP 8.1 9.9 10.3 10.8 13.1 15.1 16.9 17.2 17.8 17.5 17 16.4 16.2 15.8
MoM 1.9 5.5 8.4 10.3 10.9 8 3.2 0.4 0.1 <0.1(6) <0.1(1) <0.1(1) 0 0
CoP 5.1 5.1 4.4 4 3.9 4.7 5.6 6.1 7.3 8.3 9.6 11.5 12.6 14.3
CoC 4 4.4 6.2 7.3 10.1 13.7 18.1 20.1 19.3 16.5 14 11.4 9.6 7.6
CoM <0.1(1) <0.1(1) <0.1(7) 0.1 0.4 0.9 1.1 0.5 0.1 <0.1(27) <0.1(7) <0.1(1) 0 0
Others/unsure 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.7 0.3 0.2 0.2 0.2 0.1 0.1
All hybrid 13.2 14.4 15.6 15.2 15.3 15.9 16.3 17.2 17.8 20.3 23.1 25.7 28.2 30.3
Hybrid by bearing surface:
MoP 8.9 9.5 10 10.1 10 10.5 10.9 11.5 11.6 12.3 13.6 14.6 15.5 16.3
MoM 0.6 0.5 0.7 0.8 0.8 0.4 0.2 <0.1(32) <0.1(4) 0 0 0 0 0
CoP 1.5 1.2 1.2 1 1.3 1.8 2 2.2 3.1 5.1 7.1 9 10.8 12.5
© National Joint Registry 2018
CoC 1.7 2.7 3.2 3 2.7 2.9 3 3.1 2.9 2.7 2.4 2.1 1.7 1.4
Others/unsure 0.5 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.1 0.1 0.1 0.1 0.1 0.1
All reverse hybrid 0.7 0.9 1 1.7 2.4 2.6 2.8 3.1 3.1 3 3.1 3.1 3.2 3.1
Reverse hybrid by bearing surface:
MoP 0.5 0.7 0.8 1.1 1.7 1.8 1.9 2.2 2 2 2 2.1 2.1 2.2
CoP 0.2 0.2 0.2 0.6 0.7 0.8 0.9 0.9 1.1 1 1.1 1 1 0.9
Others/unsure <0.1(7) <0.1(4) <0.1(7) <0.1(10) <0.1(15) <0.1(14) <0.1(17) <0.1(6) <0.1(3) <0.1(6) <0.1(6) <0.1(3) <0.1(4) <0.1(7)
All resurfacing
10 10.6 10.8 10.3 8.9 6.6 3.9 2.5 1.4 1.1 1 0.9 0.7 0.6
(MoM)
All unsure 0.1 <0.1(1) 0 <0.1(1) 0 0 0 0 0 0 0 0 0 0
All types 100 100 100 100 100 100 100 100 100 100 100 100 100 100
Table 3.4 shows the distributions by fixation and Figure 3.3 illustrates the temporal changes in fixation
bearing groups for each year for primary hip of primary hip replacements. Since 2012, the most
replacements. Although the absolute number of marked feature is the increase in the use of hybrid
cemented implants used annually is nearly unchanged primary hip replacements.
between 2004 and 2017, the proportion has nearly
halved. The percentage of hybrid implants used
has tripled over the same period and the use of
uncemented implants doubled.
70
60
Percentage of primaries
50
30
20
10
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
www.njrcentre.org.uk 41
Figure 3.4 (a) Cemented primary hip replacement bearing surface by year
100
90
Percentage of cemented primaries
80
70
© National Joint Registry 2018
60
50
40
30
20
10
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
42 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.4 (b) Uncemented primary hip replacement bearing surface by year.
Figure 3.4 (b) Uncemented primary hip replacement bearing surface by year
60
Percentage of uncemented primaries
50
30
20
10
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
www.njrcentre.org.uk 43
Figure 3.4 (c) Hybrid primary hip replacement bearing surface by year
Figure 3.4 (c) Hybrid primary hip replacement bearing surface by year.
70
Percentage of hybrid primaries
60
50
© National Joint Registry 2018
40
30
20
10
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
44 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.4 (d) Reverse hybrid primary hip replacement bearing surface by year
Figure 3.4 (d) Reverse hybrid primary hip replacement bearing surface by year.
80
Percentage of reverse hybrid primaries
70
60
40
30
20
10
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
www.njrcentre.org.uk 45
Table 3.5 Age at primary hip replacement by fixation and bearing.
Age (years)
By bearing surface Percentage
Fixation within fixation group n Median (IQR*) Mean (SD) (%) males
All cases 992,090 69 (61-76) 68.0 (11.4) 40.2
All cemented 339,220 74 (68-79) 72.8 (9.2) 33.7
Cemented and
MoP 293,839 75 (69-80) 74.0 (8.3) 33.0
MoM 1,108 64 (57-73) 64.4 (11.6) 46.8
CoP 38,128 65 (58-71) 64.2 (10.4) 38.7
Others/unsure 6,145 72 (64-78) 70.6 (11.3) 36.2
All uncemented 386,042 65 (58-72) 64.6 (11.4) 44.4
Uncemented and
MoP 149,785 71 (65-77) 70.1 (9.4) 41.0
MoM 28,903 64 (57-70) 63.0 (11.1) 50.6
© National Joint Registry 2018
*IQR=interquartile range.
46 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
www.njrcentre.org.uk 47
3.3.2 First revisions after primary hip surgery
A total of 27,605 first revisions of a hip prosthesis surgery records of operations undertaken between
have been linked to NJR primary hip replacement 2003 and 2017.
Figure 3.5 (a) KM estimates of cumulative revision by year, in primary hip replacements
Figure 3.5 (a) KM estimates of cumulative revision by year, in primary hip replacements.
2003
7
2004
2005
6
2006
© National Joint Registry 2018
2007
Cumulative revision (%)
5
2008
2009
4
2010
2011
3
2012
2013
2
2014
2015
1
2016
2017
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Years since primary
Figures 3.5 (a) and (b) illustrate temporal changes in 5 and 7 years has also been highlighted. Figure 3.5 (b)
the overall revision rates using Kaplan-Meier estimates; separates each year, allowing changes in failure rates
procedures have been grouped by the year of the over time to be clearly identified. If revision surgery
primary operation. Figure 3.5 (a) plots each Kaplan- and timing of revision surgery were static across time,
Meier survival curve with a common origin, i.e. time zero we would expect all of the failure curves to be the
is equal to the year of operation. This illustrates that same shape and equally spaced; departures from
revision rates increased between 2003 and 2008 and this indicate a change in the number and timing of
then declined between 2008 and 2017. Since 2008, revision procedures. It is also very clear that the three-
the time specific rate of overall revision appears to have and five-year rate of revision increases for operations
changed with increased early revision and decreased occurring between 2003 and 2008 and then reduces
revision in the medium term. for operations occurring between 2009 and 2017. The
differences may be partly a result of under-reporting
Figure 3.5 (b) shows the same curves plotted against in the earlier years of the registry, but most probably
calendar time, where the origin of each curve is the reflects the usage of metal-on-metal bearings, which
year of operation. In addition, the revision rate at 1, 3, peaked in 2008 and then fell (see Table 3.4 on page 40).
48 www.njrcentre.org.uk
Figure 3.5 (b) KM estimates of cumulative revision by year, in primary hip replacements plotted by year of primary.
Figure 3.5 (b) KM estimates of cumulative revision by year, in primary hip replacements
1
© National Joint Registry 2018
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Year of primary
www.njrcentre.org.uk
49
50
Table 3.7 KM estimates of cumulative revision (95% CI) by fixation and bearing, in primary hip replacements.
Blue italics signify that fewer than 250 cases remained at risk at these time points.
www.njrcentre.org.uk
CoP 38,128 0.47 (0.41-0.55) 0.98 (0.87-1.09) 1.35 (1.22-1.49) 2.40 (2.15-2.68) 3.05 (2.68-3.46) 3.77 (3.20-4.43)
Others/unsure 6,145 0.65 (0.47-0.88) 1.23 (0.97-1.54) 1.70 (1.39-2.08) 3.32 (2.78-3.95) 4.80 (3.96-5.80) 4.98 (4.09-6.06)
All uncemented 386,042 0.98 (0.95-1.01) 1.88 (1.84-1.93) 2.85 (2.79-2.91) 6.29 (6.15-6.42) 7.55 (7.36-7.75) 8.94 (8.55-9.35)
Uncemented and
MoP 149,785 1.05 (1.00-1.11) 1.76 (1.69-1.83) 2.22 (2.14-2.31) 4.09 (3.91-4.27) 5.19 (4.91-5.49) 6.40 (5.87-6.99)
MoM 28,903 1.03 (0.92-1.16) 3.41 (3.20-3.63) 7.54 (7.23-7.85) 17.71 (17.23-18.21) 20.23 (19.60-20.88) 22.21 (20.86-23.63)
CoP 78,777 0.84 (0.77-0.90) 1.44 (1.35-1.54) 1.93 (1.82-2.06) 3.32 (3.09-3.56) 3.97 (3.66-4.31) 5.28 (4.54-6.14)
CoC 121,382 0.95 (0.89-1.00) 1.79 (1.72-1.87) 2.36 (2.27-2.45) 3.81 (3.64-4.00) 4.59 (4.30-4.90) 6.12 (5.30-7.07)
CoM 2,156 0.65 (0.39-1.10) 2.82 (2.20-3.62) 4.86 (4.02-5.87) 8.51 (6.62-10.91)
Others/unsure 5,039 1.34 (1.06-1.70) 2.28 (1.90-2.74) 3.12 (2.66-3.66) 5.26 (4.58-6.03) 6.48 (5.52-7.60) 7.29 (6.03-8.81)
All hybrid 200,706 0.77 (0.74-0.82) 1.32 (1.26-1.37) 1.86 (1.79-1.93) 3.56 (3.41-3.70) 4.45 (4.24-4.67) 5.38 (4.97-5.83)
Hybrid and
© National Joint Registry 2018
MoP 121,818 0.82 (0.77-0.87) 1.36 (1.29-1.44) 1.85 (1.76-1.94) 3.37 (3.20-3.56) 4.34 (4.07-4.63) 5.32 (4.77-5.93)
MoM 2,191 0.82 (0.52-1.31) 3.03 (2.39-3.85) 6.54 (5.56-7.68) 16.29 (14.68-18.07) 18.92 (16.92-21.12) 20.91 (18.31-23.83)
CoP 49,592 0.73 (0.65-0.81) 1.19 (1.08-1.31) 1.55 (1.41-1.71) 2.42 (2.12-2.77) 3.18 (2.65-3.82) 4.61 (3.42-6.19)
CoC 24,709 0.61 (0.52-0.71) 1.08 (0.96-1.23) 1.61 (1.45-1.79) 2.82 (2.55-3.13) 3.25 (2.91-3.64) 3.52 (3.03-4.08)
Others/unsure 2,396 1.19 (0.82-1.72) 1.56 (1.13-2.16) 1.95 (1.45-2.62) 3.39 (2.61-4.39) 3.82 (2.89-5.03) 3.82 (2.89-5.03)
All reverse hybrid 25,929 0.83 (0.72-0.95) 1.56 (1.41-1.73) 2.07 (1.88-2.29) 3.77 (3.30-4.30) 5.08 (4.12-6.27) 6.78 (5.00-9.16)
Reverse hybrid and
MoP 17,662 0.88 (0.75-1.03) 1.54 (1.36-1.75) 2.05 (1.81-2.31) 4.07 (3.44-4.82) 5.86 (4.51-7.59) 8.63 (5.90-12.56)
CoP 8,158 0.71 (0.55-0.93) 1.55 (1.28-1.87) 2.09 (1.76-2.48) 3.12 (2.53-3.84) 3.43 (2.65-4.45) 3.43 (2.65-4.45)
Others/unsure 109** 1.89 (0.48-7.35) 5.20 (2.19-12.08) 5.20 (2.19-12.08) 9.37 (4.75-18.04)
All resurfacing
40,154 1.24 (1.14-1.36) 3.06 (2.90-3.24) 5.46 (5.24-5.70) 11.18 (10.84-11.53) 13.07 (12.66-13.48) 14.76 (14.19-15.35)
(MoM)
*Includes 39 with unsure fixation/bearing surface; **Wide CI because based on very small group size (n=109).
Note: Blank cells indicate the number at risk is below ten and thus estimates have been omitted as they are highly unreliable.
National Joint Registry | 15th Annual Report | Hips
Table 3.7 provides Kaplan-Meier estimates of the Further revisions in these groups would be highly
cumulative percentage probability of first revision, for unlikely and, when they do occur, they may appear to
any cause, firstly for all cases combined and then by have a disproportionate impact on the Kaplan-Meier
type of fixation and by bearing surface within each estimate, i.e. the step upwards may seem steeper.
fixation group. The table shows updated estimates Furthermore, the upper 95% CI at these time points
at 1, 3, 5, 10, 12 and 14 years from the primary may be underestimated. Although a number of
operation together with 95% Confidence Intervals statistical methods have been proposed to deal with
(95% CI). Results at 14 years have been added, but in this, they typically give different values and, as yet,
general, the group sizes are too small for meaningful there is no clear consensus for the large datasets we
sub-division, hence many of these estimates are have here. Kaplan-Meier estimates are not shown at
shown in blue italics. Estimates in blue italics indicate all when the numbers at risk fell below ten cases.
time points where fewer than 250 cases remained
at risk, meaning that the estimates are less reliable.
Figure 3.6. KM estimates of cumulative revision in cemented primary hip replacements by bearing
Figure 3.6 KM estimates of cumulative revision in cemented primary hip replacements by bearing.
20
Cumulative revision (%)
15
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figures 3.6 to 3.9 illustrate the differences between perform worse than all other options regardless of
the various bearing surface sub-groups for cemented, fixation. The failure rates for ceramic-on-polyethylene
uncemented, hybrid and reverse hybrid hips, bearings remain particularly low and it is encouraging
respectively. Metal-on-metal bearings continue to that these are becoming more widely used with time.
www.njrcentre.org.uk 51
Figure 3.7. KM estimates of cumulative revision in uncemented primary hip replacements by bearing
Figure 3.7 KM estimates of cumulative revision in uncemented primary hip replacements by bearing.
25
Cumulative revision (%)
20
15
© National Joint Registry 2018
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
52 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.8 KM estimates of cumulative revision in hybrid primary hip replacements by bearing.
Figure 3.8. KM estimates of cumulative revision in hybrid primary hip replacements by bearing
20
Cumulative revision (%)
15
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
www.njrcentre.org.uk 53
Figure 3.9 KM estimates of cumulative revision in reverse hybrid primary hip replacements by bearing.
Figure 3.9. KM estimates of cumulative revision in reverse hybrid primary hip replacements
by bearing
8
Cumulative revision (%)
7
© National Joint Registry 2018
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
54 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.10 (a) KM estimates of cumulative revision in all primary hip replacements by gender and age.
Figure 3.10 (a) KM estimates of cumulative revision in all primary hip replacements by gender and age
Males Females
14 14 <55 y
55−59 y
12 12 60−64 y
10 10
70−74 y
75−79 y
8 8
80+ y
6 6
4 4
2 2
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
In Figures 3.10 (a) and 3.10 (b), the whole cohort has the age groups was greater in women than in men.
been sub-divided by age at primary operation and by Thus, for example, women under 55 years had higher
gender. Across the whole group, there was an inverse revision rates than their male counterparts in the same
relationship between the probability of revision and age band, whereas women aged 80 years and older
the age of the patient. A closer look at both genders had a lower revision rate than their male counterparts.
(Figure 3.10 (a)) shows that the variation between
www.njrcentre.org.uk 55
Figure 3.10 (b) KM estimates of cumulative revision in all primary hip replacements by gender and age,
excluding MoM & resurfacing.
Figure 3.10 (b) KM estimates of cumulative revision in all primary hip replacements by gender and age,
excluding MoM & resurfacing
Males Females
8 8 <55 y
55−59 y
© National Joint Registry 2018
7 7
60−64 y
Cumulative revision (%)
6 6 65−69 y
70−74 y
5 5
75−79 y
4 4 80+ y
3 3
2 2
1 1
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
In Figure 3.10 (b), primary total hip replacements with Where group sizes permitted (overall group
metal-on-metal (or uncertain) bearing surfaces and size>10,000), Table 3.8 further expands Table 3.7 to
resurfacings have been excluded. The revision rates show separate estimates for males and females within
for the younger women are much reduced compared each of four age bands, <55, 55-64, 65-74 and 75+
to the data in Figure 3.10 (a) which includes metal-on- years. Estimates are shown at 1, 3, 5, 10, 12 and 14
metal bearings; an age trend is seen in both genders years after the primary operation. These refine results
but rates for women are lower than for men across the shown for the first time in the 2015 12th Annual
entire age spectrum. Report, but now with larger numbers of cases and
therefore generally narrower Confidence Intervals.
56 www.njrcentre.org.uk
Table 3.8 KM estimates of cumulative revision (95% CI) of primary hip replacements by gender, age group, fixation and bearing.
Blue italics signify that fewer than 250 cases remained at risk at these time points.
Males Females
Fixation Age at
Time since primary Time since primary
group/ primary
bearing (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.91 2.19 3.65 7.84 9.36 10.66 0.93 2.33 4.06 9.67 11.72 14.00
All cases <55 59,754 59,927
(0.84-1.00) (2.06-2.32) (3.48-3.83) (7.52-8.17) (8.95-9.80) (10.04-11.32) (0.85-1.01) (2.20-2.46) (3.88-4.25) (9.31-10.05) (11.24-12.22) (13.19-14.87)
All 0.73 1.88 2.81 6.36 9.06 11.24 0.74 1.67 2.49 6.00 7.64 9.26
<55 4,840 7,313
cemented (0.52-1.03) (1.50-2.34) (2.32-3.40) (5.34-7.57) (7.59-10.80) (9.07-13.88) (0.57-0.97) (1.38-2.02) (2.11-2.94) (5.17-6.94) (6.53-8.94) (7.61-11.25)
0.92 2.29 3.35 6.72 9.98 12.92 0.99 1.97 2.59 6.03 8.16 9.53
MoP <55 2,041 3,493
(0.58-1.45) (1.69-3.10) (2.57-4.35) (5.30-8.51) (7.93-12.53) (10.01-16.58) (0.70-1.38) (1.54-2.53) (2.06-3.25) (4.97-7.29) (6.69-9.93) (7.52-12.04)
0.61 1.28 1.81 3.41 4.29 4.29 0.51 1.28 2.03 4.11 4.93 7.39
CoP <55 2,444 3,480
(0.36-1.02) (0.87-1.88) (1.28-2.56) (2.36-4.90) (2.86-6.40) (2.86-6.40) (0.32-0.82) (0.93-1.77) (1.52-2.70) (3.06-5.52) (3.52-6.89) (4.70-11.55)
All 0.96 2.35 3.82 8.50 10.12 11.61 0.95 2.21 3.72 8.24 9.98 12.62
<55 32,285 34,764
uncemented (0.86-1.08) (2.18-2.54) (3.58-4.07) (7.98-9.06) (9.40-10.88) (10.33-13.05) (0.85-1.06) (2.05-2.38) (3.49-3.96) (7.76-8.76) (9.29-10.71) (11.13-14.29)
0.93 2.13 3.14 6.11 7.21 7.21 1.20 2.05 2.88 4.62 6.16 9.85
MoP <55 3,843 4,733
(0.66-1.30) (1.68-2.69) (2.53-3.89) (4.93-7.55) (5.70-9.09) (5.70-9.09) (0.92-1.57) (1.65-2.53) (2.37-3.50) (3.75-5.69) (4.77-7.95) (6.65-14.48)
0.68 3.52 7.42 17.44 20.19 22.34 1.78 5.70 12.55 26.35 29.62 33.42
MoM <55 3,247 2,368
(0.45-1.03) (2.93-4.21) (6.56-8.39) (16.07-18.92) (18.44-22.07) (19.22-25.87) (1.32-2.40) (4.83-6.71) (11.27-13.96) (24.54-28.25) (27.46-31.91) (29.32-37.93)
1.14 2.04 2.94 3.75 4.69 5.33 0.87 1.55 2.36 4.58 5.39 9.33
CoP <55 6,614 7,166
(0.90-1.44) (1.68-2.49) (2.42-3.57) (3.00-4.68) (3.51-6.25) (3.77-7.50) (0.68-1.13) (1.26-1.91) (1.94-2.87) (3.60-5.81) (4.14-7.02) (5.84-14.72)
0.96 2.17 3.08 4.93 5.83 8.50 0.81 1.84 2.57 4.68 5.56 5.98
CoC <55 17,992 19,752
(0.82-1.11) (1.95-2.41) (2.81-3.38) (4.41-5.50) (5.04-6.75) (6.03-11.93) (0.69-0.95) (1.65-2.05) (2.34-2.83) (4.19-5.23) (4.85-6.38) (5.10-7.00)
© National Joint Registry 2018
0.89 1.59 2.45 5.90 6.62 7.07 0.76 1.39 2.09 4.64 5.50 6.86
All hybrid <55 8,265 10,972
(0.70-1.12) (1.32-1.91) (2.07-2.89) (5.04-6.90) (5.58-7.84) (5.89-8.47) (0.61-0.95) (1.17-1.65) (1.79-2.44) (4.02-5.36) (4.70-6.43) (5.50-8.53)
1.25 2.21 3.29 7.57 8.97 9.90 0.92 1.80 2.54 5.41 6.55 9.34
MoP <55 1,512 2,172
(0.79-1.97) (1.54-3.18) (2.36-4.57) (5.53-10.34) (6.36-12.57) (6.91-14.07) (0.59-1.43) (1.28-2.51) (1.87-3.43) (4.10-7.13) (4.83-8.85) (6.25-13.84)
1.00 1.50 2.15 3.59 3.59 3.59 0.79 1.40 1.69 3.56 3.56 5.09
CoP <55 3,377 4,180
(0.70-1.41) (1.09-2.07) (1.52-3.05) (2.28-5.64) (2.28-5.64) (2.28-5.64) (0.55-1.12) (1.04-1.88) (1.25-2.28) (2.41-5.23) (2.41-5.23) (2.66-9.61)
0.65 1.33 1.91 3.53 3.53 3.53 0.58 0.99 1.55 3.19 4.11 4.60
CoC <55 2,985 4,243
(0.41-1.01) (0.97-1.84) (1.44-2.53) (2.69-4.62) (2.69-4.62) (2.69-4.62) (0.39-0.86) (0.72-1.35) (1.19-2.03) (2.51-4.06) (3.17-5.32) (3.37-6.27)
All reverse 0.99 2.04 2.30 5.10 5.10 0.93 1.69 2.62 4.57 7.01
<55 723 1,022
hybrid (0.47-2.07) (1.18-3.49) (1.35-3.88) (2.71-9.50) (2.71-9.50) (0.48-1.77) (1.02-2.81) (1.67-4.10) (2.86-7.26) (3.33-14.46)
0.76 3.29 3.29 3.29 3.29 0.75 4.45 9.00
MoP <55 147 217 0 0
(0.11-5.26) (1.24-8.54) (1.24-8.54) (1.24-8.54) (1.24-8.54) (0.11-5.22) (1.61-12.00) (3.00-25.32)
All
0.88 2.23 4.07 8.15 9.57 10.58 1.32 4.97 9.27 19.89 22.83 25.53
resurfacing <55 13,637 5,852
(0.74-1.05) (1.99-2.50) (3.74-4.43) (7.63-8.69) (8.94-10.24) (9.77-11.45) (1.06-1.65) (4.44-5.56) (8.55-10.05) (18.84-20.99) (21.63-24.07) (23.87-27.29)
(MoM)
www.njrcentre.org.uk
57
58
Table 3.8 (continued)
Males Females
Fixation Age at
Time since primary Time since primary
group/ primary
bearing (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.89 1.89 2.91 6.24 7.66 9.03 0.73 1.64 2.74 6.35 7.79 9.39
All cases 55-64 98,593 119,685
(0.83-0.95) (1.80-1.98) (2.79-3.03) (6.02-6.47) (7.36-7.97) (8.54-9.54) (0.68-0.78) (1.56-1.72) (2.63-2.84) (6.14-6.56) (7.51-8.08) (8.89-9.92)
All 0.63 1.46 2.08 4.63 5.98 7.34 0.45 1.09 1.75 3.97 5.09 6.64
55-64 16,357 26,507
cemented (0.52-0.77) (1.28-1.67) (1.85-2.34) (4.17-5.13) (5.38-6.65) (6.44-8.36) (0.37-0.54) (0.96-1.23) (1.58-1.94) (3.64-4.33) (4.66-5.56) (5.92-7.43)
0.66 1.68 2.39 5.08 6.55 8.09 0.48 1.19 1.83 3.99 5.04 6.69
MoP 55-64 10,458 18,106
(0.52-0.84) (1.44-1.96) (2.09-2.74) (4.53-5.70) (5.83-7.35) (7.04-9.29) (0.39-0.59) (1.03-1.37) (1.63-2.06) (3.61-4.40) (4.56-5.57) (5.89-7.60)
www.njrcentre.org.uk
0.58 0.98 1.26 2.45 2.92 3.30 0.33 0.74 1.16 2.53 3.57 4.43
CoP 55-64 5,259 7,624
(0.41-0.84) (0.73-1.31) (0.96-1.66) (1.80-3.34) (2.06-4.14) (2.25-4.81) (0.22-0.49) (0.55-0.98) (0.90-1.49) (1.95-3.28) (2.68-4.76) (3.22-6.07)
All 0.92 1.99 3.07 7.13 8.86 10.45 0.82 1.78 2.91 6.98 8.34 9.75
55-64 52,056 60,507
uncemented (0.84-1.01) (1.86-2.12) (2.90-3.24) (6.76-7.51) (8.33-9.42) (9.42-11.59) (0.75-0.90) (1.67-1.90) (2.77-3.07) (6.66-7.32) (7.90-8.80) (8.84-10.74)
0.99 2.03 2.65 5.18 6.89 9.09 0.81 1.74 2.26 4.51 5.96 6.99
MoP 55-64 12,509 16,133
(0.83-1.18) (1.78-2.31) (2.34-2.99) (4.57-5.88) (5.96-7.95) (7.46-11.06) (0.68-0.97) (1.54-1.97) (2.01-2.53) (4.03-5.04) (5.21-6.82) (5.83-8.38)
0.86 3.07 6.60 16.72 19.93 19.93 0.85 3.58 8.93 22.09 24.89 27.44
MoM 55-64 5,125 4,820
(0.64-1.16) (2.63-3.59) (5.94-7.32) (15.63-17.88) (18.43-21.55) (18.43-21.55) (0.63-1.16) (3.09-4.14) (8.15-9.77) (20.87-23.36) (23.36-26.50) (23.64-31.70)
0.87 1.47 2.04 3.53 4.85 7.10 0.66 1.33 1.86 3.58 4.02 4.98
CoP 55-64 12,189 14,545
(0.72-1.06) (1.24-1.73) (1.75-2.39) (2.97-4.20) (3.95-5.96) (4.76-10.53) (0.54-0.81) (1.13-1.55) (1.61-2.15) (3.05-4.21) (3.38-4.78) (3.81-6.50)
0.92 1.85 2.43 3.91 4.56 5.61 0.89 1.59 2.13 3.43 4.12 6.25
CoC 55-64 21,290 23,726
(0.80-1.06) (1.67-2.05) (2.22-2.67) (3.50-4.37) (3.95-5.27) (4.47-7.02) (0.78-1.02) (1.44-1.77) (1.94-2.33) (3.09-3.81) (3.60-4.73) (4.15-9.37)
© National Joint Registry 2018
0.80 1.50 2.17 4.20 5.40 6.57 0.62 1.22 1.83 3.63 4.61 5.89
All hybrid 55-64 16,298 24,831
(0.67-0.95) (1.30-1.71) (1.91-2.45) (3.72-4.75) (4.72-6.18) (5.47-7.87) (0.53-0.73) (1.08-1.37) (1.64-2.04) (3.28-4.01) (4.12-5.15) (4.87-7.11)
1.04 1.78 2.41 4.43 5.70 6.97 0.78 1.31 1.96 3.68 4.75 6.18
MoP 55-64 5,585 9,476
(0.80-1.35) (1.45-2.19) (2.00-2.91) (3.68-5.32) (4.67-6.95) (5.53-8.77) (0.61-0.98) (1.09-1.57) (1.67-2.31) (3.18-4.26) (4.05-5.57) (4.71-8.08)
0.66 1.17 1.52 2.57 4.07 5.95 0.58 1.12 1.36 2.58 4.14 5.75
CoP 55-64 6,179 8,824
(0.48-0.90) (0.89-1.53) (1.14-2.01) (1.74-3.80) (2.51-6.57) (2.96-11.79) (0.44-0.77) (0.89-1.41) (1.07-1.72) (1.90-3.49) (2.80-6.11) (3.56-9.21)
0.61 1.10 1.76 2.89 3.61 4.43 0.40 0.93 1.40 2.31 2.51 2.51
CoC 55-64 4,003 5,840
(0.41-0.90) (0.82-1.49) (1.37-2.26) (2.28-3.66) (2.74-4.74) (2.89-6.76) (0.26-0.60) (0.71-1.23) (1.11-1.76) (1.87-2.86) (2.00-3.15) (2.00-3.15)
All reverse 0.99 2.18 3.08 5.10 6.68 0.85 1.80 2.48 4.18 6.59 6.59
55-64 2,026 3,226
hybrid (0.63-1.55) (1.58-3.00) (2.29-4.14) (3.51-7.38) (3.89-11.37) (0.58-1.24) (1.36-2.37) (1.92-3.19) (3.13-5.58) (4.16-10.36) (4.16-10.36)
0.82 1.66 2.94 6.02 8.56 1.23 1.98 3.03 5.91 8.85
MoP 55-64 786 1,378
(0.37-1.81) (0.92-2.99) (1.77-4.86) (3.33-10.78) (4.21-17.00) (0.75-1.99) (1.33-2.95) (2.13-4.31) (3.97-8.73) (5.23-14.76)
All
1.22 2.42 3.94 7.41 8.63 9.82 1.63 4.50 8.58 17.48 20.48 23.49
resurfacing 55-64 11,852 4,610
(1.04-1.44) (2.15-2.71) (3.60-4.32) (6.91-7.95) (8.02-9.27) (9.00-10.70) (1.30-2.04) (3.94-5.14) (7.80-9.43) (16.37-18.66) (19.16-21.87) (21.55-25.57)
(MoM)
Males Females
Fixation Age at
Time since primary Time since primary
group/ primary
bearing (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.84 1.55 2.20 4.52 5.66 6.61 0.67 1.29 1.93 3.91 4.76 5.39
All cases 65-74 137,683 211,769
(0.80-0.89) (1.48-1.62) (2.11-2.29) (4.35-4.70) (5.41-5.91) (6.24-7.00) (0.63-0.71) (1.24-1.35) (1.86-2.00) (3.78-4.04) (4.59-4.95) (5.12-5.67)
All 0.57 1.17 1.68 3.61 4.90 5.92 0.41 0.99 1.47 2.87 3.68 4.27
65-74 44,977 82,671
cemented (0.51-0.65) (1.07-1.28) (1.55-1.81) (3.37-3.88) (4.55-5.28) (5.40-6.49) (0.37-0.46) (0.92-1.07) (1.38-1.56) (2.71-3.03) (3.46-3.91) (3.96-4.60)
0.60 1.22 1.72 3.70 5.00 6.11 0.40 0.98 1.47 2.89 3.72 4.33
MoP 65-74 38,860 72,650
(0.53-0.69) (1.11-1.34) (1.58-1.87) (3.43-3.98) (4.62-5.41) (5.55-6.73) (0.35-0.45) (0.90-1.06) (1.37-1.57) (2.73-3.07) (3.49-3.97) (4.00-4.69)
0.39 0.84 1.21 2.23 2.87 2.87 0.51 1.10 1.32 1.94 2.47 2.97
CoP 65-74 5,177 8,485
(0.25-0.61) (0.61-1.16) (0.90-1.63) (1.67-2.99) (2.04-4.05) (2.04-4.05) (0.38-0.69) (0.88-1.38) (1.07-1.63) (1.53-2.44) (1.86-3.26) (2.13-4.13)
All 0.96 1.77 2.52 5.37 6.55 7.80 0.89 1.65 2.51 5.50 6.34 6.76
65-74 58,346 75,152
uncemented (0.89-1.05) (1.66-1.89) (2.38-2.67) (5.06-5.70) (6.08-7.04) (7.03-8.65) (0.83-0.96) (1.56-1.75) (2.38-2.64) (5.23-5.79) (5.97-6.72) (6.31-7.25)
0.94 1.67 2.07 4.26 5.53 6.73 0.93 1.55 1.97 3.61 4.48 4.71
MoP 65-74 25,678 36,627
(0.83-1.07) (1.52-1.85) (1.89-2.28) (3.82-4.74) (4.84-6.32) (5.60-8.07) (0.83-1.03) (1.42-1.69) (1.82-2.14) (3.30-3.95) (4.00-5.00) (4.15-5.35)
1.06 2.97 5.98 13.47 15.07 17.20 1.10 3.45 8.55 19.47 20.81 21.49
MoM 65-74 4,565 4,668
(0.80-1.40) (2.51-3.51) (5.32-6.73) (12.39-14.64) (13.67-16.61) (14.98-19.71) (0.83-1.44) (2.96-4.02) (7.77-9.40) (18.26-20.75) (19.43-22.28) (19.85-23.25)
0.77 1.26 1.54 2.46 3.11 4.37 0.75 1.36 1.76 3.05 3.57 3.72
CoP 65-74 12,817 15,666
(0.63-0.94) (1.07-1.49) (1.31-1.81) (2.02-3.00) (2.44-3.96) (2.97-6.41) (0.62-0.90) (1.17-1.58) (1.53-2.03) (2.61-3.57) (2.98-4.26) (3.08-4.49)
1.12 1.82 2.28 3.37 4.57 5.37 0.87 1.50 1.83 2.60 3.09 4.16
CoC 65-74 14,241 16,847
(0.96-1.31) (1.61-2.07) (2.03-2.56) (2.91-3.90) (3.56-5.85) (3.99-7.20) (0.74-1.02) (1.32-1.70) (1.63-2.06) (2.26-3.00) (2.41-3.96) (2.95-5.86)
0.88 1.48 2.02 3.93 4.56 5.00 0.75 1.21 1.73 3.08 4.03 5.00
All hybrid 65-74 27,545 46,557
© National Joint Registry 2018
(0.77-1.00) (1.33-1.64) (1.83-2.22) (3.55-4.36) (4.07-5.11) (4.32-5.78) (0.68-0.84) (1.11-1.32) (1.60-1.88) (2.82-3.36) (3.62-4.47) (4.08-6.13)
0.89 1.49 2.01 3.88 4.63 5.08 0.76 1.26 1.76 3.08 4.06 4.87
MoP 65-74 17,173 30,888
(0.75-1.04) (1.30-1.69) (1.78-2.27) (3.42-4.40) (4.03-5.31) (4.25-6.07) (0.66-0.86) (1.14-1.40) (1.60-1.94) (2.78-3.41) (3.59-4.59) (3.85-6.15)
0.84 1.39 1.61 2.66 2.66 2.66 0.72 1.07 1.46 1.62 2.23 4.78
CoP 65-74 7,172 11,357
(0.64-1.09) (1.11-1.75) (1.27-2.05) (1.86-3.78) (1.86-3.78) (1.86-3.78) (0.57-0.90) (0.87-1.30) (1.18-1.79) (1.30-2.02) (1.48-3.36) (2.04-11.00)
0.79 1.41 2.01 3.05 3.05 0.79 0.99 1.43 2.78 3.29
CoC 65-74 2,596 3,502
(0.51-1.22) (1.01-1.97) (1.50-2.70) (2.26-4.13) (2.26-4.13) (0.54-1.14) (0.70-1.39) (1.06-1.92) (2.08-3.69) (2.38-4.55)
All reverse 1.07 1.90 2.39 4.40 5.69 0.59 1.06 1.51 3.36 3.99 3.99
65-74 3,718 6,544
hybrid (0.78-1.47) (1.48-2.45) (1.88-3.03) (3.20-6.04) (3.43-9.38) (0.43-0.81) (0.82-1.36) (1.20-1.90) (2.49-4.55) (2.68-5.93) (2.68-5.93)
1.35 2.19 2.82 4.92 6.69 0.60 1.03 1.42 3.94 4.80
MoP 65-74 2,575 4,854
(0.96-1.90) (1.65-2.89) (2.16-3.68) (3.34-7.23) (3.71-11.88) (0.42-0.87) (0.77-1.39) (1.08-1.87) (2.79-5.54) (3.06-7.50)
All
1.96 3.09 4.57 7.84 8.95 9.45 1.92 3.74 6.61 15.54 17.72 19.87
resurfacing 65-74 3,093 836
(1.52-2.51) (2.53-3.77) (3.87-5.40) (6.86-8.95) (7.78-10.29) (7.99-11.17) (1.18-3.12) (2.65-5.28) (5.10-8.55) (13.10-18.39) (14.81-21.12) (15.84-24.76)
(MoM)
www.njrcentre.org.uk
59
60
Table 3.8 (continued)
Males Females
Fixation Age at
Time since primary Time since primary
group/ primary
bearing (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.93 1.53 2.00 3.51 4.07 4.98 0.69 1.12 1.49 2.52 2.95 3.36
All cases 75+ 102,428 202,251
(0.87-0.99) (1.45-1.62) (1.90-2.10) (3.30-3.72) (3.77-4.40) (4.27-5.80) (0.65-0.73) (1.07-1.17) (1.43-1.55) (2.41-2.64) (2.79-3.12) (3.09-3.66)
0.73
All 1.27 1.70 2.94 3.51 4.21 0.43 0.83 1.15 1.99 2.34 2.72
75+ 48,198 (0.66- 108,357
cemented (1.17-1.39) (1.57-1.84) (2.69-3.21) (3.14-3.91) (3.47-5.12) (0.39-0.47) (0.78-0.89) (1.08-1.22) (1.87-2.13) (2.16-2.53) (2.43-3.03)
0.82)
0.73 1.28 1.70 2.97 3.54 4.29 0.43 0.82 1.14 1.98 2.33 2.69
www.njrcentre.org.uk
MoP 75+ 45,526 102,705
(0.66-0.82) (1.17-1.40) (1.57-1.85) (2.72-3.25) (3.16-3.97) (3.51-5.23) (0.39-0.47) (0.76-0.88) (1.07-1.21) (1.85-2.12) (2.15-2.52) (2.40-3.01)
0.73 1.33 1.60 1.94 1.94 0.41 0.74 1.06 1.46 1.46 2.71
CoP 75+ 1,863 3,796
(0.42-1.26) (0.86-2.04) (1.04-2.45) (1.18-3.18) (1.18-3.18) (0.25-0.68) (0.49-1.10) (0.73-1.54) (1.00-2.15) (1.00-2.15) (1.08-6.74)
All 1.28 1.93 2.50 4.53 4.65 6.18 1.25 1.79 2.25 3.97 4.72 5.80
75+ 28,892 44,040
uncemented (1.15-1.42) (1.77-2.11) (2.30-2.72) (4.06-5.05) (4.13-5.23) (4.31-8.81) (1.15-1.36) (1.66-1.92) (2.10-2.41) (3.65-4.31) (4.23-5.28) (4.60-7.30)
1.35 2.01 2.45 3.62 3.62 6.34 1.25 1.71 2.10 3.47 4.01 4.88
MoP 75+ 19,340 30,922
(1.19-1.53) (1.81-2.23) (2.22-2.72) (3.18-4.13) (3.18-4.13) (3.45-11.50) (1.13-1.38) (1.56-1.87) (1.93-2.28) (3.11-3.88) (3.46-4.65) (3.38-7.02)
1.08 1.97 3.80 9.09 9.09 1.29 3.04 4.86 9.06 12.52
MoM 75+ 1,698 2,412
(0.68-1.71) (1.39-2.77) (2.95-4.89) (7.40-11.14) (7.40-11.14) (0.91-1.83) (2.41-3.82) (4.05-5.84) (7.78-10.54) (9.78-15.96)
1.10 1.59 2.13 3.07 3.07 1.03 1.49 1.73 3.10 3.10 3.67
CoP 75+ 4,084 5,696
(0.82-1.48) (1.22-2.06) (1.65-2.75) (2.07-4.53) (2.07-4.53) (0.79-1.33) (1.19-1.87) (1.39-2.16) (2.39-4.02) (2.39-4.02) (2.52-5.32)
1.21 1.90 2.05 4.63 6.12 1.50 1.89 2.13 3.57 4.38
CoC 75+ 3,279 4,255
(0.89-1.65) (1.48-2.45) (1.60-2.63) (3.07-6.96) (3.47-10.68) (1.17-1.91) (1.51-2.35) (1.72-2.63) (2.39-5.31) (2.69-7.10)
0.88 1.54 1.96 3.59 4.82 6.24 0.74 1.13 1.52 2.42 2.97 2.97
All hybrid 75+ 22,217 44,021
(0.76-1.02) (1.37-1.73) (1.75-2.20) (3.09-4.16) (3.84-6.03) (3.87-9.98) (0.66-0.83) (1.03-1.25) (1.39-1.67) (2.15-2.71) (2.52-3.51) (2.52-3.51)
0.85 1.57 1.99 3.53 4.97 6.77 0.78 1.18 1.52 2.43 3.07 3.07
© National Joint Registry 2018
www.njrcentre.org.uk 61
Figure 3.11 (a) KM estimates of cumulative revision of primary cemented MoP hip replacement
(monobloc cups) by head size.
Figure 3.11 (a) KM estimates of cumulative revision of primary cemented MoP hip replacement
(monobloc cups) by head size
7
Cumulative revision (%)
5
© National Joint Registry 2018
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
62 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.11 (b) KM estimates of cumulative revision of primary uncemented MoP hip replacements
Figure 3.11 (b) KM (metal shells & polyethylene line) by head size
estimates of cumulative revision of primary uncemented MoP hip replacements
(metal shells & polyethylene liner) by head size.
12
Cumulative revision (%)
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
www.njrcentre.org.uk 63
Figure 3.11 (c) KM estimates of cumulative revision of primary uncemented MoM hip replacement
(monobloc cups or metal shell liner) by head size.
Figure 3.11 (c) KM estimates of cumulative revision of primary uncemented MoM hip replacement
(monobloc cups or metal shell liner) by head size
35
30
Cumulative revision (%)
© National Joint Registry 2018
25
20
15
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Years since primary
64 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.11 (d) KM estimates of cumulative revision of primary cemented CoP hip replacement
(monobloc cups) by head size.
Figure 3.11 (d) KM estimates of cumulative revision of primary cemented CoP hip replacement
(monobloc cups) by head size
5
Cumulative revision (%)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
www.njrcentre.org.uk 65
Figure 3.11 (e) KM estimates of cumulative revision of primary uncemented CoP hip replacement
(metal shell & polyethylene liner) by head size.
Figure 3.11 (e) KM estimates of cumulative revision of primary uncemented CoP hip replacement
(metal shell & polyethylene liner) by head size
5
Cumulative revision (%)
© National Joint Registry 2018
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
66 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.11 (f) KM estimates of cumulative revision of primary uncemented CoC hip replacement
(metal shell & ceramic liner) by head size.
Figure 3.11 (f) KM estimates of cumulative revision of primary uncemented CoC hip replacement
(metal shell & ceramic liner) by head size
6
Cumulative revision (%)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
www.njrcentre.org.uk 67
3.3.4 Revisions after primary hip at risk; no results are shown at all where the number
had fallen below ten cases. Given that the sub-groups
surgery for the main stem/cup may differ in composition with respect to age and
brand combinations gender, the percentage of males and the median (IQR)
of the ages are also shown in these tables.
As in previous reports, we have only included stem/
cup brand combinations with more than 2,500 Table 3.9 shows Kaplan-Meier estimates of the
procedures for cemented, uncemented, hybrid and cumulative percentage probability of revision of
reverse hybrid hips or more than 1,000 procedures in primary hip replacement (for any reason) for the main
the case of resurfacings. The figures in blue italics are stem/cup brands.
at time points where fewer than 250 cases remained
Table 3.9 KM estimates of cumulative revision (95% CI) of primary hip replacement by fixation, and stem/cup
brand. Blue italics indicate that fewer than 250 cases remained at risk at these time points.
5,178 72 (66-77) 40
Stem / Elite Plus Ogee (0.23-0.57) (0.64-1.18) (0.88-1.52) (1.27-2.08) (2.09-3.31) (3.01-5.12)
C-Stem Cemented 0.41 0.94 1.30 1.93
7,185 68 (59-75) 41
Stem / Marathon (0.28-0.59) (0.72-1.24) (1.00-1.67) (1.45-2.56)
MS-30 / Original ME 0.20 0.48 0.75 1.03 1.72 2.26
3,534 74 (68-80) 32
Muller Low Profile C (0.10-0.42) (0.29-0.79) (0.49-1.16) (0.68-1.55) (1.14-2.60) (1.29-3.95)
Muller Straight Stem /
0.42 0.80 1.08 1.92 2.51 3.97
Original ME Muller Low 2,877 74 (69-80) 31
(0.24-0.75) (0.52-1.23) (0.73-1.59) (1.37-2.69) (1.81-3.50) (2.57-6.09)
Profile C
Stanmore Modular
0.43 1.08 1.54 1.86 2.42 4.07
Stem / Stanmore- 5,382 75 (70-80) 29
(0.29-0.65) (0.83-1.41) (1.23-1.94) (1.49-2.30) (1.95-3.01) (3.12-5.30)
Arcom Cup
0.64 1.46 2.01 2.69 3.67 4.53
CPT / Elite Plus Ogee 2,992 73 (67-79) 36
(0.41-1.00) (1.08-1.97) (1.55-2.62) (2.10-3.43) (2.86-4.69) (3.18-6.42)
0.74 1.35 2.02 2.57 3.62 4.56
CPT / ZCA 14,872 76 (71-81) 30
(0.61-0.89) (1.16-1.56) (1.76-2.31) (2.26-2.92) (3.16-4.13) (3.82-5.45)
Exeter V40 / Exeter 0.45 0.91 1.29 1.63 2.34 3.62
77,380 74 (68-79) 35
Contemporary Flanged (0.41-0.50) (0.84-0.98) (1.20-1.38) (1.52-1.75) (2.16-2.52) (3.11-4.20)
Exeter V40 / Elite Plus 0.37 0.81 1.15 1.56 2.22 2.98
24,563 74 (69-80) 35
Ogee (0.30-0.45) (0.70-0.93) (1.02-1.31) (1.39-1.75) (1.97-2.49) (2.57-3.46)
Exeter V40 / Exeter 0.59 1.19 1.65 2.44 3.75 5.65
16,978 73 (67-79) 32
Duration (0.48-0.72) (1.04-1.37) (1.46-1.87) (2.20-2.72) (3.39-4.15) (4.93-6.46)
0.39 0.85 1.17 1.56 3.04 5.98
Exeter V40 / Opera 2,824 74 (68-80) 32
(0.22-0.71) (0.56-1.27) (0.82-1.66) (1.13-2.15) (2.24-4.12) (4.16-8.55)
Exeter V40 / Cenator 0.59 1.37 2.01 2.25 2.86 4.76
2,556 75 (69-80) 32
Cemented Cup (0.36-0.98) (0.98-1.92) (1.52-2.66) (1.71-2.95) (2.19-3.72) (3.53-6.42)
Exeter V40 / Elite Plus 0.34 0.67 0.84 1.15 1.55 2.10
5,085 73 (66-79) 33
Cemented Cup (0.21-0.54) (0.47-0.94) (0.62-1.16) (0.85-1.54) (1.14-2.10) (1.48-2.98)
Note: Blank cells indicate that the number at risk at the time shown has fallen below ten and thus estimates have been omitted as they are highly unreliable.
68 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Note: Blank cells indicate that the number at risk at the time shown has fallen below ten and thus estimates have been omitted as they are highly unreliable.
www.njrcentre.org.uk 69
Table 3.9 (continued)
Note: Blank cells indicate that the number at risk at the time shown has fallen below ten and thus estimates have been omitted as they are highly unreliable.
70 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Table 3.10 KM estimates of cumulative revision (95% CI) of primary hip replacement by fixation, stem/cup
brand, and bearing. Blue italics signify that fewer than 250 cases remained at risk at these time points.
Note: Blank cells indicate that the number at risk at the time shown has fallen below ten and thus estimates have been omitted as they are highly unreliable.
www.njrcentre.org.uk 71
Table 3.10 (continued)
Note: Blank cells indicate that the number at risk at the time shown has fallen below ten and thus estimates have been omitted as they are highly unreliable.
Sub-groups with more than 10,000 procedures were made for Charnley Cemented Stem/Charnley
in Table 3.9 have been further divided by bearing Cemented Cup, as all the procedures described in
surface. Table 3.10 shows the estimated cumulative Table 3.9 were cemented MoP. Similarly, the majority
percentage probability of revision for the resulting of the cemented CPT/ZCA and Exeter V40/Exeter
fixation/bearing sub-groups provided there were more Duration combinations shown in Table 3.9 were MoP.
than 1,000 procedures. Note: no further sub-divisions
72 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
3.3.5 Revisions for different causes (MDSv3 onwards) was staggered over time and so
revisions associated with a few primaries as late as
after primary hip replacement 2011 had revisions reported on MDSv1 and MDSv2 of
Overall, 27,605 (2.8%) of the 992,090 primary hip the data collection forms. Restricting our analyses to
replacements had an associated first revision. The primaries from 2008 onwards however, as we did in
most common indications for revision were aseptic our previous four annual reports, ensures that 99.4%
loosening (6,705), adverse soft tissue reaction to of revisions were recorded on later forms (MDSv3
particulate debris (4,619, a figure that is likely to be an onwards). We noted, however, that only 2,055 of the
underestimate due to changes in MDS collection, see 4,619 instances of adverse reactions to particulate
later), dislocation/subluxation (4,663), pain (4,507), and debris would thus be included, i.e. we are thereby
infection (3,872). Pain was not usually cited alone; in missing 2,564 of the earlier cases. Therefore, as we
3,117 out of the 4,507 instances, it was cited together did last year, we present two sets of PTIRs, one set
with one or more other indications. Associated PTIRs for all primary hip replacements, which are likely to be
for these, and the other indications are shown in underestimates, and the other set for all primary hip
Table 3.11. Here, implant wear denotes either wear of replacements performed since the beginning of 2008,
the polyethylene component, wear of the acetabular which has better ascertainment but does not include
component or dissociation of the liner. the cases with longer term follow-up.
The number of adverse reactions to particulate debris Table 3.11, overleaf, reports revision by indication
is likely to be under-estimated because this was not with further breakdowns by hip fixation and bearing.
solicited (i.e. it was not available as an indication for Metal-on-metal (irrespective of type of fixation) and
revision) on the revision data collection forms in the resurfacings seem to have the highest PTIRs for both
early phase of the registry, i.e. was missing for MDSv1 aseptic loosening and pain. Metal-on-metal bearings
and MDSv2. Some of these cases may have recorded have the highest incidence of adverse reaction to
the indication for revision as ‘other’ but we simply do particulate debris.
not know. Adoption of the later revision report forms
www.njrcentre.org.uk 73
74
Table 3.11 PTIR estimates of indications for hip revision (95% CI) by fixation and bearing.
Adverse
Prosthesis- reaction to
Pros- years at risk particulate
thesis- Adverse (x1,000) for debris for
years Peri- Head/ reaction to primaries primaries
Fixation/ at risk Aseptic Dislocation/ prosthetic Mal- Implant Implant socket size Other particulate from from
bearing type (x1,000) loosening Pain subluxation Infection fracture alignment Lysis wear fracture mismatch indication debris*** 1.1.2008**** 1.1.2008****
1.25 0.84 0.87 0.72 0.69 0.37 0.28 0.28 0.16 0.04 0.48 0.86 0.59
www.njrcentre.org.uk
All cases* 5,373.0 3,454.1
(1.22-1.28) (0.81-0.86) (0.84-0.89) (0.70-0.74) (0.67-0.71) (0.36-0.39) (0.27-0.29) (0.27-0.29) (0.15-0.17) (0.03-0.04) (0.46-0.50) (0.84-0.88) (0.57-0.62)
All 1.03 0.33 0.82 0.67 0.48 0.21 0.20 0.16 0.08 0.01 0.16 0.08 0.06
1,941.2 1,088.7
cemented (0.99-1.08) (0.31-0.36) (0.78-0.86) (0.63-0.71) (0.45-0.51) (0.19-0.23) (0.18-0.22) (0.15-0.18) (0.07-0.09) (0.01-0.02) (0.14-0.18) (0.07-0.10) (0.05-0.08)
Cemented and
1.05 0.32 0.84 0.66 0.49 0.21 0.20 0.17 0.07 0.01 0.15 0.03 0.03
MoP 1,702.0 935.9
(1.00-1.10) (0.29-0.34) (0.80-0.89) (0.62-0.70) (0.46-0.52) (0.19-0.24) (0.18-0.22) (0.15-0.19) (0.06-0.08) (0.01-0.02) (0.13-0.17) (0.02-0.04) (0.02-0.04)
3.58 3.06 0.92 1.02 1.43 0.20 1.84 0.20 0.82 0.10 2.55 8.68 10.25
MoM 9.8 2.8
(2.57-4.98) (2.14-4.38) (0.48-1.77) (0.55-1.90) (0.85-2.41) (0.05-0.82) (1.16-2.92) (0.05-0.82) (0.41-1.63) (0.01-0.73) (1.73-3.78) (7.02-10.74) (7.12-14.74)
0.78 0.30 0.65 0.73 0.33 0.16 0.11 0.13 0.09 0.01 0.14 0.05 0.06
CoP 186.2 128.6
(0.66-0.92) (0.23-0.39) (0.54-0.78) (0.61-0.86) (0.26-0.43) (0.11-0.23) (0.07-0.17) (0.09-0.19) (0.06-0.15) (0.00-0.04) (0.10-0.21) (0.03-0.09) (0.03-0.12)
Others/ 0.99 0.44 0.60 0.93 0.44 0.16 0.23 0.12 0.16 0.02 0.23 0.37 0.19
43.2 21.4
unsure (0.74-1.34) (0.28-0.69) (0.41-0.88) (0.68-1.26) (0.28-0.69) (0.08-0.34) (0.12-0.43) (0.05-0.28) (0.08-0.34) (0.00-0.16) (0.12-0.43) (0.23-0.60) (0.07-0.50)
All 1.53 1.02 0.93 0.75 0.72 0.51 0.30 0.42 0.23 0.06 0.63 1.38 0.94
2,034.1 1,491.5
uncemented (1.47-1.58) (0.98-1.07) (0.89-0.97) (0.72-0.79) (0.68-0.76) (0.48-0.54) (0.28-0.33) (0.39-0.45) (0.21-0.25) (0.05-0.07) (0.59-0.66) (1.33-1.43) (0.89-0.99)
Uncemented and
© National Joint Registry 2018
1.19 0.56 1.17 0.71 0.91 0.47 0.22 0.46 0.10 0.05 0.30 0.18 0.20
MoP 742.4 555.3
(1.11-1.27) (0.51-0.62) (1.10-1.25) (0.65-0.77) (0.85-0.99) (0.42-0.52) (0.19-0.25) (0.42-0.52) (0.08-0.13) (0.04-0.07) (0.26-0.34) (0.15-0.21) (0.16-0.24)
3.60 4.00 0.90 1.41 0.71 0.88 1.27 0.69 0.18 0.09 2.46 10.11 9.60
MoM 245.4 118.5
(3.37-3.85) (3.75-4.26) (0.79-1.02) (1.27-1.57) (0.61-0.82) (0.77-1.00) (1.14-1.42) (0.59-0.80) (0.13-0.24) (0.06-0.14) (2.27-2.67) (9.72-10.52) (9.06-10.18)
1.09 0.46 0.99 0.64 0.60 0.41 0.15 0.38 0.11 0.04 0.30 0.08 0.08
CoP 333.1 239.8
(0.98-1.21) (0.39-0.54) (0.89-1.10) (0.56-0.73) (0.52-0.69) (0.35-0.49) (0.12-0.20) (0.32-0.46) (0.08-0.15) (0.02-0.07) (0.25-0.37) (0.05-0.11) (0.05-0.13)
1.31 0.72 0.65 0.62 0.56 0.46 0.11 0.27 0.45 0.06 0.47 0.15 0.15
CoC 659.6 542.3
(1.23-1.40) (0.66-0.78) (0.59-0.71) (0.57-0.69) (0.51-0.62) (0.41-0.52) (0.09-0.14) (0.24-0.32) (0.40-0.50) (0.04-0.08) (0.42-0.52) (0.13-0.19) (0.12-0.19)
2.60 1.59 0.76 1.21 0.38 0.76 0.51 0.70 0.19 0.19 1.14 1.84 1.84
CoM 15.8 15.2
(1.92-3.53) (1.07-2.35) (0.43-1.34) (0.77-1.89) (0.17-0.85) (0.43-1.34) (0.25-1.01) (0.39-1.26) (0.06-0.59) (0.06-0.59) (0.72-1.81) (1.28-2.65) (1.27-2.67)
Others/ 1.88 0.98 0.85 0.50 0.82 0.56 0.34 0.37 0.34 0.08 0.53 1.03 0.78
37.7 20.5
unsure (1.49-2.37) (0.71-1.35) (0.60-1.20) (0.32-0.79) (0.58-1.17) (0.36-0.85) (0.20-0.59) (0.22-0.63) (0.20-0.59) (0.03-0.25) (0.34-0.82) (0.76-1.41) (0.48-1.28)
Adverse
Prosthesis- reaction to
Pros- years at risk particulate
thesis- Adverse (x1,000) for debris for
years Peri- Head/ reaction to primaries primaries
Fixation/ at risk Aseptic Dislocation/ prosthetic Mal- Implant Implant socket size Other particulate from from
bearing type (x1,000) loosening Pain subluxation Infection fracture alignment Lysis wear fracture mismatch indication debris*** 1.1.2008**** 1.1.2008****
0.60 0.43 1.03 0.81 0.89 0.29 0.18 0.23 0.14 0.03 0.28 0.23 0.15
All hybrid 925.3 639.2
(0.55-0.65) (0.39-0.47) (0.97-1.10) (0.75-0.87) (0.83-0.95) (0.26-0.33) (0.15-0.20) (0.20-0.26) (0.12-0.17) (0.02-0.04) (0.25-0.32) (0.20-0.27) (0.13-0.19)
Hybrid and
0.59 0.33 1.13 0.83 0.99 0.28 0.17 0.25 0.11 0.02 0.23 0.06 0.06
MoP 578.9 396.3
(0.53-0.66) (0.28-0.38) (1.05-1.22) (0.76-0.91) (0.91-1.07) (0.24-0.33) (0.14-0.21) (0.21-0.30) (0.08-0.14) (0.01-0.04) (0.19-0.27) (0.05-0.09) (0.04-0.09)
3.56 3.72 1.39 1.19 1.81 0.62 1.44 0.36 0.26 0.10 2.37 7.74 7.80
MoM 19.4 7.3
(2.81-4.51) (2.95-4.68) (0.96-2.03) (0.79-1.79) (1.30-2.52) (0.35-1.09) (1.00-2.09) (0.17-0.76) (0.11-0.62) (0.03-0.41) (1.78-3.17) (6.60-9.08) (6.02-10.11)
0.35 0.27 1.14 0.91 0.87 0.19 0.10 0.17 0.07 0.03 0.22 0.04 0.03
CoP 154.3 128.8
(0.27-0.46) (0.20-0.36) (0.98-1.32) (0.77-1.08) (0.73-1.03) (0.14-0.28) (0.06-0.16) (0.12-0.26) (0.04-0.13) (0.01-0.08) (0.16-0.31) (0.02-0.09) (0.01-0.08)
0.51 0.51 0.53 0.54 0.47 0.37 0.09 0.17 0.32 0.03 0.28 0.12 0.13
CoC 155.6 98.7
(0.41-0.63) (0.41-0.64) (0.42-0.65) (0.44-0.67) (0.37-0.59) (0.29-0.48) (0.05-0.15) (0.12-0.25) (0.24-0.42) (0.01-0.08) (0.20-0.37) (0.08-0.19) (0.08-0.23)
Others/ 0.59 0.82 0.88 0.94 0.35 0.18 0.23 0.29 0.18 0.23 0.18 0.12
17.1 0 8.0
unsure (0.32-1.09) (0.49-1.39) (0.53-1.46) (0.57-1.53) (0.16-0.78) (0.06-0.55) (0.09-0.62) (0.12-0.70) (0.06-0.55) (0.09-0.62) (0.06-0.55) (0.02-0.89)
© National Joint Registry 2018
All reverse 1.32 0.45 0.94 0.86 0.66 0.31 0.15 0.24 0.04 0.03 0.29 0.07 0.05
118.9 98.1
hybrid (1.13-1.54) (0.34-0.58) (0.78-1.13) (0.71-1.04) (0.53-0.82) (0.23-0.43) (0.10-0.24) (0.17-0.35) (0.02-0.10) (0.01-0.09) (0.21-0.41) (0.03-0.13) (0.02-0.12)
Reverse and
1.25 0.31 1.04 0.84 0.78 0.30 0.15 0.21 0.04 0.03 0.29 0.08 0.05
MoP 79.7 65.8
(1.03-1.53) (0.21-0.46) (0.84-1.29) (0.66-1.07) (0.61-1.00) (0.20-0.45) (0.09-0.27) (0.13-0.34) (0.01-0.12) (0.01-0.10) (0.19-0.43) (0.03-0.17) (0.01-0.14)
1.43 0.70 0.75 0.86 0.42 0.31 0.16 0.31 0.05 0.05 0.29 0.03 0.03
CoP 38.5 31.9
(1.10-1.86) (0.48-1.02) (0.52-1.09) (0.61-1.21) (0.25-0.68) (0.18-0.55) (0.07-0.35) (0.18-0.55) (0.01-0.21) (0.01-0.21) (0.16-0.52) (0.00-0.18) (0.00-0.22)
Others/ 2.84 1.42 2.84 1.42 1.42 1.42 2.30
0.7 0 0 0 0 0 0 0.4
unsure** (0.71-11.36) (0.20-10.09) (0.71-11.36) (0.20-10.09) (0.20-10.09) (0.20-10.09) (0.32-16.32)
All
2.49 3.76 0.32 0.54 1.17 0.70 0.92 0.29 0.27 0.07 1.93 4.03 3.60
resurfacing 353.2 136.5
(2.33-2.66) (3.57-3.97) (0.26-0.38) (0.47-0.62) (1.06-1.28) (0.62-0.80) (0.82-1.02) (0.24-0.35) (0.22-0.33) (0.05-0.10) (1.79-2.08) (3.83-4.25) (3.29-3.93)
(MoM)
www.njrcentre.org.uk
75
76
Table 3.12 PTIR estimates of indications for hip revision (95% CI) by years following primary hip replacement.
Prosthesis-
Number of failures per 1,000 prosthesis-years for: years
at risk Adverse
Pros- (x1,000) for reaction to
Time thesis- Adverse primaries particulate
period years Peri- Head/ reaction to from debris for
since at risk Aseptic Dislocation/ prosthetic Implant Implant socket size Other particulate 1.1.2008 primaries from
primary (x1,000) loosening Pain subluxation Infection fracture Malalignment Lysis wear fracture mismatch indication debris** only*** 1.1.2008****
1.25 0.84 0.87 0.72 0.69 0.37 0.28 0.28 0.16 0.04 0.48 0.86 0.59
All cases 5,373.0 3,454.1
(1.22-1.28) (0.81-0.86) (0.84-0.89) (0.70-0.74) (0.67-0.71) (0.36-0.39) (0.27-0.29) (0.27-0.29) (0.15-0.17) (0.03-0.04) (0.46-0.50) (0.84-0.88) (0.57-0.62)
1.11 0.61 2.39 1.63 1.69 0.76 0.08 0.35 0.23 0.11 0.72 0.09 0.11
www.njrcentre.org.uk
<1 year 934.9 744.5
(1.04-1.18) (0.56-0.66) (2.30-2.50) (1.55-1.71) (1.61-1.78) (0.71-0.82) (0.06-0.10) (0.31-0.39) (0.21-0.27) (0.09-0.13) (0.66-0.77) (0.07-0.11) (0.09-0.13)
1.07 0.84 0.63 0.73 0.37 0.34 0.15 0.14 0.13 0.03 0.40 0.26 0.32
1-3 years 1,552.0 1,182.5
(1.02-1.12) (0.79-0.88) (0.59-0.67) (0.69-0.77) (0.34-0.40) (0.31-0.37) (0.13-0.17) (0.13-0.17) (0.11-0.15) (0.02-0.04) (0.37-0.43) (0.24-0.29) (0.29-0.36)
1.07 0.96 0.46 0.46 0.41 0.26 0.24 0.22 0.12 0.02 0.45 0.91 0.80
3-5 years 1,155.8 805.0
(1.01-1.13) (0.91-1.02) (0.42-0.50) (0.42-0.50) (0.38-0.45) (0.24-0.29) (0.22-0.27) (0.20-0.25) (0.10-0.14) (0.01-0.03) (0.41-0.49) (0.85-0.96) (0.74-0.87)
© National Joint Registry 2018
1.31 1.04 0.44 0.39 0.54 0.26 0.39 0.29 0.15 0.02 0.48 1.70 1.19
5-7 years 809.8 482.3
(1.23-1.39) (0.97-1.11) (0.40-0.49) (0.35-0.44) (0.49-0.59) (0.23-0.30) (0.35-0.43) (0.26-0.33) (0.13-0.18) (0.01-0.03) (0.44-0.53) (1.62-1.80) (1.10-1.30)
1.70 0.84 0.58 0.41 0.63 0.27 0.62 0.41 0.17 0.02 0.45 1.97 1.56
7-10 years 680.6 239.8
(1.61-1.80) (0.77-0.91) (0.53-0.64) (0.36-0.46) (0.57-0.69) (0.24-0.32) (0.57-0.69) (0.36-0.46) (0.14-0.20) (0.01-0.03) (0.40-0.50) (1.87-2.08) (1.41-1.73)
2.31 0.45 0.65 0.40 0.85 0.22 0.75 0.78 0.23 0.01 0.25 1.51
10+ years* 240.0
(2.13-2.51) (0.38-0.55) (0.56-0.76) (0.33-0.49) (0.74-0.98) (0.17-0.28) (0.65-0.87) (0.68-0.90) (0.17-0.29) (0.00-0.04) (0.20-0.33) (1.36-1.68)
In Table 3.12, the PTIRs for each indication are shown pain both increased with time from surgery, whereas
separately for different time periods from the primary the rates due to subluxation/dislocation, infection,
hip replacement, within the first year, and between periprosthetic fracture, and malalignment were all
1-3, 3-5, 5-7, 7-10 and 10+ years after surgery (note: higher in the first year and then fell. Adverse reaction
the maximum follow-up for any implant is now 14.75 to particulate debris increased with time, as did lysis,
years). The same overall time trends are seen as although the PTIRs for the latter were low.
before – revision rates due to aseptic loosening and
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
0 1 2 3 4 5 6
Figures 3.12 (a) to 3.12 (g) show how PTIRs for dislocation/subluxation in all fixation/bearing groups
aseptic loosening, pain, dislocation/subluxation, which later fell (Figure 3.12 (c)). Revision rates for
infection, lysis and adverse soft tissue reaction to infection were initially high and then fell in all groups
particulate debris changed with time in an arbitrary apart from uncemented metal-on-metal primary total
selection of the cemented/uncemented bearing sub- hip replacement (Figure 3.12 (d)).
groups from Table 3.12. Only sub-groups with a total
overall prosthesis-years at risk of more than 150,000 Revision rates due to adverse reaction to particulate
have been included. With time from operation, debris increased with time up to five years in
PTIRs for aseptic loosening and pain tended to rise uncemented metal-on-metal primary total hip
in uncemented metal-on-metal primary total hip replacement and resurfacings (Figures 3.12 (f) and
replacements and resurfacings. These trends were (g)). Confidence Intervals have not been shown here
not seen in the other groups shown (Figures 3.12 (a) for simplicity, but could be quite wide; these trends
and (b)). Conversely, there was a high initial rate for require more in-depth investigation.
www.njrcentre.org.uk 77
Figure 3.12 (b) PTIR estimate of pain by fixation & bearing
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
© National Joint Registry 2018
7−10y
10+y
0−1y
1−3y
(iii) Uncemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(iv) Uncemented MoM 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(v) Uncemented CoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vi) Uncemented CoC 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vii) Resurfacing 3−5y
5−7y
7−10y
10+y
0 1 2 3 4 5 6
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
© National Joint Registry 2018
7−10y
10+y
0−1y
1−3y
(iii) Uncemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(iv) Uncemented MoM 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(v) Uncemented CoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vi) Uncemented CoC 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vii) Resurfacing 3−5y
5−7y
7−10y
10+y
0 1 2 3 4 5 6
78 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
0 1 2 3 4 5 6
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
7−10y
© National Joint Registry 2018
10+y
0−1y
1−3y
(iii) Uncemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(iv) Uncemented MoM 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(v) Uncemented CoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vi) Uncemented CoC 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vii) Resurfacing 3−5y
5−7y
7−10y
10+y
0 1 2 3 4 5 6
www.njrcentre.org.uk 79
Figure 3.12 (f) PTIR estimate of adverse soft tissue reaction by fixation &
estimates of adverse soft tissue reaction by fixation & bearing.
Figure 3.12 (f) PTIR bearing
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(ii) Cemented CoP 3−5y
5−7y
7−10y
© National Joint Registry 2018
10+y
0−1y
1−3y
(iii) Uncemented MoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(iv) Uncemented MoM 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(v) Uncemented CoP 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vi) Uncemented CoC 3−5y
5−7y
7−10y
10+y
0−1y
1−3y
(vii) Resurfacing 3−5y
5−7y
7−10y
10+y
0 5 10 15 20
Figure 3.12 (g) PTIR estimate of adverse soft tissue reaction by fixation &
bearing, since 2008
Figure 3.12 (g) PTIR estimates of adverse soft tissue reaction by fixation & bearing, since 2008.
0−1y
1−3y
(i) Cemented MoP 3−5y
5−7y
7−10y
0−1y
1−3y
(ii) Cemented CoP 3−5y
© National Joint Registry 2018
5−7y
7−10y
0−1y
1−3y
(iii) Uncemented MoP 3−5y
5−7y
7−10y
0−1y
1−3y
(iv) Uncemented MoM 3−5y
5−7y
7−10y
0−1y
1−3y
(v) Uncemented CoP 3−5y
5−7y
7−10y
0−1y
1−3y
(vi) Uncemented CoC 3−5y
5−7y
7−10y
0−1y
1−3y
(vii) Resurfacing 3−5y
5−7y
7−10y
0 5 10 15 20
80 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
3.3.6 Mortality after primary hip Personal Demographic Service. Amongst the 992,090
primary hip replacements there were 4,617 bilateral
replacement surgery operations, with the left and right side operated on
This section describes the mortality of the cohort up the same day; here the second of the two has been
to 14 years from primary hip replacement, according excluded, leaving 987,473 primary hip replacements,
to gender and age group. Deaths were updated of whom 138,481 had died before the end of 2017.
on 16 February 2018 using data from the NHS
Table 3.13 KM estimates of cumulative mortality (95% CI) by age and gender, in primary hip replacement.
Blue italics indicate that fewer than 250 cases remained at risk at these time points.
*Some patients had operations on the left and right side on the same day. The second of 4,617 pairs of simultaneous bilateral operations were excluded.
www.njrcentre.org.uk 81
Table 3.13 (continued)
<55 59,331
(0.04-0.08) (0.17-0.25) (0.60-0.73) (2.31-2.59) (4.78-5.30) (6.89-8.16)
0.07 0.19 0.58 3.02 6.95 10.98
55-59 46,819
(0.05-0.10) (0.15-0.23) (0.52-0.66) (2.85-3.21) (6.61-7.30) (10.22-11.80)
0.07 0.17 0.60 3.76 9.36 16.06
60-64 72,117
(0.05-0.09) (0.14-0.21) (0.55-0.66) (3.60-3.93) (9.04-9.69) (15.23-16.93)
0.08 0.23 0.76 4.77 13.78 25.17
65-69 100,281
(0.07-0.10) (0.20-0.26) (0.70-0.81) (4.62-4.93) (13.44-14.13) (24.23-26.14)
0.12 0.28 0.96 7.15 21.52 39.52
70-74 110,778
(0.10-0.14) (0.25-0.32) (0.91-1.02) (6.98-7.34) (21.12-21.92) (38.47-40.59)
0.23 0.46 1.50 11.57 34.83 59.65
75-79 100,404
(0.20-0.26) (0.42-0.50) (1.43-1.58) (11.33-11.81) (34.33-35.33) (58.47-60.83)
0.36 0.82 2.54 18.64 53.50 79.29
80-84 66,861
(0.32-0.41) (0.75-0.89) (2.42-2.66) (18.29-18.99) (52.85-54.16) (78.05-80.51)
0.83 1.78 4.85 32.25 74.40 92.84
85+ 34,710
(0.74-0.93) (1.65-1.93) (4.63-5.09) (31.66-32.85) (73.57-75.22) (91.65-93.92)
*Some patients had operations on the left and right side on the same day. The second of 4,617 pairs of simultaneous bilateral operations were excluded.
Table 3.13 shows Kaplan-Meier estimates of 3.3.7 Primary hip replacement for
cumulative percentage mortality at 30 days, 90 days
and at 1, 5, 10 and 14 years from the primary hip
fractured neck of femur compared
replacement, for all cases and by age and gender. with other reasons for implantation
Note: These cases were not censored when further As total hip replacement is an increasingly popular
revision surgery was undertaken. Whilst such surgery treatment option for fractured neck of femur, this
may have contributed to the overall mortality, the section further updates results from last year’s annual
impact of this is not investigated here. report on revision and mortality rates for primary total
hip replacements performed as a result of fractured
neck of femur compared to cases implanted for other
indications. A total of 29,689 (3.0%) of the primary
total hip replacements were performed for a fractured
1
neck of femur (#NOF) .
1 These comprised 2,231 cases with indication for primary hip replacement including fractured neck of femur in the early phase of the registry (i.e. 185,001
implants entered using MDSv1 and v2) and 27,458 cases with reasons including acute trauma neck of femur in the later phase (i.e. 725,337 entered using
MDSv3 and v6). 39 cases were omitted as no indication for primary hip replacement was given.
82 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Table 3.14 shows that the proportion of primary hip neck of femur has continued to increase with time to a
replacements performed for an indication of fractured maximum of 5.5% in 2016 and 5.3% recorded in 2017.
www.njrcentre.org.uk 83
Table 3.15 compares the #NOF group with The #NOF cases were significantly older (median
the remainder with respect to gender and age age 73 years versus 69 years at operation: P<0.001
composition together and type of hip replacement by Mann-Whitney U-test). Cemented and hybrid
received. A significantly larger percentage of the #NOF hips were used more commonly in #NOF than in hip
cases, compared with the remainder, were women replacements performed for other indications.
(72.9% versus 59.3%: P<0.001, Chi-squared test).
Figure 3.13. KM estimates of cumulative revision by fractured NOF and OA only cases for
primary hip replacements
Figure 3.13 KM estimates of cumulative revision for fractured NOF and OA only cases for primary
hip replacements.
10
Cumulative revision (%)
8
© National Joint Registry 2018
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.13 shows that the cumulative revision rate as stratification by these variables left the result
was higher in the #NOF group compared with the unchanged (P<0.001 using stratified logrank test: 14
remainder (P<0.001, logrank test). This effect was sub-groups of age <55, 55-59, 60-64, 65-69, 70-74,
not fully explained by differences in age and gender, 75-79, 80+ for each gender).
84 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
60
Cumulative mortality (%)
50
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.14 shows a markedly worse overall survival together with 4,617 cases that were the second
in the #NOF cases compared to cases implanted of simultaneous bilateral procedures. Gender/age
for other reasons (P<0.001, logrank test). As in the differences did not fully explain the difference seen as
overall mortality section, 208 cases with untraced a stratified analysis still showed a difference (P<0.001)
NHS numbers or missing age have been excluded, but the results warrant further exploration.
www.njrcentre.org.uk 85
3.3.8 Overview of hip revision there were 64,743 revisions for which no primary hip
replacement had been recorded in the NJR.
procedures
Revisions are classified as single stage, stage one
This section looks at all hip revision procedures
and stage two of two-stage revisions. Information on
performed since the start of the registry, 1 April 2003,
stage one and stage two revisions are entered into
up to 31 December 2017, for all patients with valid
the database separately, whereas stage one and
patient identifiers (i.e. whose data could therefore
stage two revisions in practice have to be linked.
be linked).
Although not all patients who undergo a stage one
In total, there were 106,200 revisions on 92,348 of two revision will undergo a stage two of two
2
individual patient-sides (87,038 actual patients). revision, in some cases stage one revisions have been
In addition to the 27,605 revised primary hip entered without a stage two, and vice versa, making
replacements described in section 3.3.2 of this report, identification of individual revision episodes difficult. An
attempt has been made to do this later in this section.
Table 3.16 Number and percentage of hip revisions by procedure type and year.
*Incomplete year.
Note: MDSv1, in use in 2003, only defined operations as primary or revision. All revisions using MDSv1 have been listed as single stage procedures in this table.
Single stages include DAIRs (debridement and implant retention).
Table 3.16 gives an overview of all hip replacement The incidence of revision hip replacement peaked in
revision procedures carried out each year since 2012 and has steadily declined since then, despite
April 2003. There were up to a maximum of nine the increasing number of at-risk implants prevailing in
documented revision procedures associated with any the database.
individual patient-side (discussed later in this section).
2 For 129 patient-sides, multiple procedures had been entered on the same operation date; 128 had two on the same date and one had three. Details of the
components that had been entered for these cases were reviewed. As a result of this, 237 of the 258 revision procedures have been dropped and 21 have
been reclassified.
86 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Table 3.17 Number and percentage of hip revision by indication and procedure type.
*Not recorded in the early phase of the registry; MDSv3 and v6 only.
Table 3.17 shows the stated indication for the revision episode was undertaken. For this purpose, we
revision hip replacement surgery. Please note that, as regarded an initial stage one followed by either a stage
several reasons can be stated, the reasons are not one or a stage two as being the same revision episode
mutually exclusive and therefore column percentages and these were disregarded, looking instead for the
may not add up to 100%. Aseptic loosening is the start of a second revision episode. (We counted the
commonest indication for revision. maximum number of distinct revision episodes for any
patient-side to be eight).
3.3.9 Rates of hip re-revision
Kaplan-Meier estimates of the cumulative
For a given patient-side, we have looked at the percentage probability of having a subsequent revision
survival following the first documented revision hip (re-revision) were calculated. There were 8,528
replacement procedure in the NJR (n=92,348). In re-revisions and, for 18,945 cases, the patient died
most instances (91.4%), the first revision procedure without having been revised. The censoring date for
was a single stage revision, however in the remaining the remainder was the end of 2017.
8.6% it was part of a two-stage procedure. We have
looked at the time from the first documented revision
procedure (of any type) to the time at which a second
www.njrcentre.org.uk 87
Figure 3.15 (a) KM estimates of cumulative re-revision in linked primary hip replacements (shaded
area indicate point-wise 95% CI).
Figure 3.15 (a) KM estimates of cumulative re−revision in linked primary hip replacements
20
Cumulative re−revision (%)
© National Joint Registry 2018
15
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
88 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.15 (b) KM estimates of cumulative re-revision by primary fixation in linked primary
hip replacements.
Figure 3.15 (b) KM estimates of cumulative re−revision by primary fixation in linked primary hip
replacements
30
Cumulative re−revision (%)
25
15
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Number at risk
Cemented 6,347 4,964 4,034 3,227 2,526 1,941 1,384 974 652 412 246 131 54 23 2
Uncemented without MoM 8,272 6,698 5,547 4,405 3,437 2,618 1,813 1,255 846 517 295 148 70 25 5
Uncemented MoM 4,600 4,051 3,548 3,041 2,479 1,812 1,108 580 334 173 78 22 10 1 1
Hybrid 3,725 2,896 2,281 1,749 1,339 1,041 759 537 351 206 130 67 34 15 3
Reverse hybrid 509 390 299 247 180 134 99 63 37 17 7 4 3 0 0
Resurfacing 4,152 3,736 3,337 2,978 2,566 2,069 1,465 970 648 418 239 108 49 12 4
www.njrcentre.org.uk 89
Figure 3.15 (c) KM estimates of cumulative re-revision by years since first revision, in linked primary
hip replacements. Figure 3.15 (c) KM estimates of cumulative re−revision by years since first revision, in linked
primary hip replacements
20
Cumulative re−revision (%)
15
© National Joint Registry 2018
10
0
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 7,533 6,182 5,245 4,364 3,573 2,957 2,325 1,808
First rev. 1−3y 5,962 5,099 4,503 3,887 3,349 2,840 2,244 1,616
First rev. 3−5y 4,662 4,088 3,663 3,226 2,790 2,173 1,383 716
First rev. 5+y 9,448 7,366 5,635 4,170 2,815 1,645 676 239
Figure 3.15 (c) shows the relationship between time subluxation and pain were more prevalent in the early
to first revision and risk of subsequent revision. period after the primary hip replacement and aseptic
The earlier the primary hip replacement is revised, loosening and pain later on. The relationship between
the higher the risk of a second revision. There is a (i) the time to first revision and the subsequent time to
relationship between the indication for first revision and re-revision, and (ii) the indication for the first revision
time to first revision; earlier in this report (section 3.3.5) and the time to re-revision require further investigation.
we showed, for example, that revisions for dislocation/
90 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
25
Cumulative re−revision (%)
20
10
0
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 1,677 1,327 1,110 918 734 597 459 345
First rev. 1−3y 1,523 1,255 1,088 922 756 627 486 377
First rev. 3−5y 960 777 665 552 475 369 261 174
First rev. 5+y 2,187 1,605 1,171 835 561 348 178 78
For those with a documented primary hip replacement those who had their first revision within one year of
within the NJR, Figures 3.16 (a) to (e) show cumulative the initial primary hip replacement, experienced the
re-revision rates following the first revision hip worst re-revision rates. However, for reverse hybrid
replacement, according to the main fixation used in the hip replacements, the worst re-revision rates were
primary. Each sub-group has been further sub-divided experienced by those who had their first revision within
according to the time interval from the primary hip 3 to 5 years of the initial primary hip replacement;
replacement to the first revision, i.e. less than 1 year, however, the numbers were small and therefore the
1 to 3, 3 to 5 and more than 5 years. For cemented, results should be interpreted with caution.
uncemented, hybrid and resurfacing hip replacements,
www.njrcentre.org.uk 91
Figure 3.16 (b) KM estimates of cumulative re-revision in uncemented primary hip replacement by
years since first revision, in linked primary hip replacements.
Figure 3.16 (b) KM estimates of cumulative re−revision in uncemented primary hip replacement by
years since first revsion, in linked primary hip replacements
25
Cumulative re−revision (%)
20
© National Joint Registry 2018
15
10
0
0 1 2 3 4 5 6 7
Number at risk
First rev. <1y 3,666 3,085 2,653 2,226 1,827 1,492 1,130 856
First rev. 1−3y 2,808 2,437 2,181 1,888 1,621 1,350 1,037 706
First rev. 3−5y 2,231 1,980 1,784 1,561 1,340 1,002 557 222
First rev. 5+y 4,167 3,247 2,477 1,771 1,128 586 197 51
92 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.16 (c) KM estimates of cumulative re-revision in hybrid primary hip replacement by years
since first revision, in linked primary hip replacements.
Figure 3.16 (c) KM estimates of cumulative re−revision in hybrid primary hip replacement by
years since first revision, in linked primary hip replacements
25
Cumulative re−revision (%)
20
10
0 1 2 3 4 5 6 7
Number at risk
First rev. <1y 1,489 1,142 905 682 523 428 342 259
First rev. 1−3y 784 626 506 401 332 270 207 154
First rev. 3−5y 521 433 366 306 234 186 137 86
First rev. 5+y 931 695 504 360 250 157 73 38
www.njrcentre.org.uk 93
Figure 3.16 (d) KM estimates of cumulative re-revision in reverse hybrid primary hip replacement
by years since first revision, in linked primary hip replacements.
Figure 3.16 (d) KM estimates of cumulative re−revision in reverse hybrid primary hip replacement by
years since first revision, in linked primary hip replacements
25
Cumulative re−revision (%)
20
© National Joint Registry 2018
15
10
0
0 1 2 3 4 5 6 7
Number at risk
First rev. <1y 206 158 128 116 89 63 47 29
First rev. 1−3y 144 116 88 70 57 49 37 26
First rev. 3−5y 68 53 42 33 17 12 8 5
First rev. 5+y 91 63 41 28 17 10 7 3
94 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Figure 3.16 (e) KM estimates of cumulative re-revision in resurfacing primary hip replacement by
years since first revision, in linked primary hip replacements.
Figure 3.16 (e) KM estimates of cumulative re−revision in resurfacing primary hip replacement by
years since first revision, in linked primary hip replacements
25
Cumulative re−revision (%)
20
10
0 1 2 3 4 5 6 7
www.njrcentre.org.uk 95
Table 3.18 (a) KM estimates of cumulative re-revision (95% CI).
© National Joint Registry 2018
Table 3.18 (b) KM estimates of cumulative re-revision (95% CI) by years since first failure.
Primary in the NJR Number of first Time since first revision
© National Joint Registry 2018
Table 3.18 (b) shows that primary hip replacements at each time point compared with primaries that last
that fail within the first year after surgery have more than five years.
approximately twice the chance of needing re-revision
96 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Table 3.18 (c) KM estimates of cumulative re-revision (95% CI) by fixation and bearing.
Table 3.18 (c) shows cumulative re-revision rates at 1, The failure rates for resurfacings were comparatively
3, 5 and 7 years following the first revision for those with low, but Figure 3.15 (b) on page 89 shows that after
documented primary hip replacements within the NJR, eight years the failure rate of re-revisions following
broken down by fixation types and bearing surfaces. resurfacing is higher than alternatives.
www.njrcentre.org.uk 97
3.3.10 Reasons for hip re-revision
Table 3.19 Number of failures by indication for re-revision.
*Adverse reaction to particulate debris was only recorded using MDSv3 onwards and as such was only a potential reason for revision among a total of 85,818
revisions as oppose to 106,200 revisions for the other reasons.
Table 3.19 shows a breakdown of the stated first revision of a hip replacement compared to that of
indications for the first revision and for any second primary hip replacement.
revision (note the indications are not mutually
exclusive). Column (i) shows the indications for Tables 3.20 (a) and (b) show that the numbers of
all revisions recorded in the NJR, (ii) reports the revisions and the relative proportion of revisions
indications for the first recorded revision, (iii) records with a linked primary in the NJR increased with
the number and percentage of first recorded time. Approximately 46% of revisions performed
revisions that were revised, and column (iv) reports in 2017 had a linked primary in the NJR. This is
the indications for the second recorded revision. likely to reflect improved data capture over time,
It is interesting to note that both dislocation and improved linkability of records and the longevity of
infection are much more common indications for hip replacements with a proportion of primaries being
second revision than first revision. This shows the revised being performed before NJR data capture
increased risk of instability and infection following the began or outside the coverage of the NJR.
98 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
Table 3.20 (b) Number of re-revisions by year, stage, and whether or not primary is in NJR.
www.njrcentre.org.uk 99
3.3.11 90-day mortality after For the first time, in 2017, hybrid fixation (30.3%) was
more common than cemented fixation (28.2%). Since
hip revision 2011, the use of ceramic-on-ceramic bearings has
The overall cumulative percentage mortality at 90 declined whilst the use of ceramic-on-polyethylene
days after hip revision was lower in the cases with bearings has increased at roughly the same rate, with
their primary hip replacement recorded in the NJR ceramic-on-polyethylene bearings now being the
compared with the remainder (Kaplan-Meier estimates second most commonly chosen bearing after metal-
1.05 (95% CI 0.94-1.18) versus 1.66 (1.56-1.76)), on-polyethylene.
which may reflect the fact that this patient group
Since the 12th Annual Report in 2015, we have
were younger at the time of their first revision, median
presented data by age and gender comparing
age of 68 (IQR 60-76) years compared to the group
combinations of fixation and bearing. This assists
without primaries documented in the NJR who had a
clinicians and patients in choosing classes of
median age of 73 (IQR 65-80) years. The percentage
prostheses that are the most appropriate for particular
of males was similar in both groups (43.8% versus
types of patients. For example, in males under 55 years
42.3% respectively).
of age, at twelve years post-surgery hybrid ceramic-
on-polyethylene and ceramic-on-ceramic constructs
3.3.12 Conclusions have revision rates of less than 4%, whilst cemented
As in previous annual reports, we have analysed metal-on-polyethylene constructs have revision rates of
implants by revision of the construct, rather than 9.98% (95% CI 7.93-12.53) and uncemented ceramic-
revision of a single component, as the mechanisms of on-ceramic bearings 5.83% (95% CI 7.93-12.53).
failure (such as wear, adverse reaction to particulate In contrast, in women under 55 years, cemented
debris and dislocation) are interdependent between ceramic-on-polyethylene constructs give excellent
different parts of the construct. We have also results with a 4.93% (95% CI 3.52-6.89) revision
stratified revision by age and gender. The highest rate at twelve years. However, cemented metal-
failure rates are among young women and the lowest on-polyethylene has a higher revision rate, whilst
among older women. When data on metal-on-metal results with uncemented constructs with metal-on-
is excluded, young women have similar revision rates polyethylene, ceramic-on-polyethylene and ceramic-
to young men. Once again we must emphasise that on-ceramic are not statistically different to those
implant survivorship is only one measure of success achieved by cemented ceramic-on-polyethylene. For
and cannot be used as an indication of satisfaction, patients over 75 years old, all combinations except
relief of pain, improvement in function and greater those with metal-on-metal bearings have good
participation in society. The data clearly show that outcomes, with cemented and hybrid ceramic-on-
constructs fail at different rates depending on the age polyethylene possibly having the lowest failure rates.
and gender of the recipients.
Both male and female patients aged over 75 years
Overall the number of primary hip replacements have a less than 5% risk of revision at 14 years. The
recorded annually in the NJR continues to increase 14-year mortality rate in men aged 75-79 years is
with 996,272 now recorded of which 992,090 were 71.54% (95% CI 69.91-73.14) and in women aged
available for analysis. 75-79 years is 59.65% (95% CI 58.47-60.83). This
clearly shows that in older patients the vast majority
Since 2003 the types of implants utilised have of treatment strategies will last the rest of the patients’
changed dramatically and these changes continue. lives. Even in those aged 65 to 69 years at the time
Between 2003 and 2007 cemented fixation was the of surgery, 66% of males and 75% of females are still
most common, followed by uncemented fixation. alive 14 years later.
Between 2008 and 2016 uncemented fixation was
the most common followed by cemented fixation, We have examined head sizes (bearing diameters) with
with hybrid fixation increasing steadily since 2012. different fixation and bearing types and again these
100 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Hips
results are interesting. With metal-on-polyethylene and have received metal-on-metal bearings. Analysis of
ceramic-on-polyethylene, large head sizes appear to stemmed metal-on-metal bearings by head size shows
be associated with higher failure rates particularly with that 28mm heads have the best survivorship, but this is
36mm heads used with cemented fixation and heads still poor compared to alternatives.
>36mm used with hybrid and uncemented fixation.
Ceramic-on-ceramic bearings have lower failure rates Revision rates by year of surgery for the entire
with larger bearings as predicted by Alison Smith’s cohort increased dramatically from 2003 to 2008
flexible parametric survival models published in the and then declined until 2013. This matches the use
Lancet in 2012 .
3 of resurfacing arthroplasty and stemmed metal-on-
metal with the peak usage of these devices in 2008
With regard to specific branded stem/cup corresponding with the highest failure rates by year
combinations, some of the best implant survivorships of primary surgery. This demonstrates the profoundly
are still achieved by “mix and match” cemented hard- negative effect metal-on-metal has had on hip
on-soft bearing constructs, although this practice replacement outcomes.
remains contrary to MHRA and manufacturers’
guidelines for usage. Consistent with results from previous years’ reports,
similar revision rates were observed for total hip
It is encouraging that the most commonly used replacement performed as a result of fractured neck of
constructs by brand in cemented and hybrid femur and those done for other causes. As expected,
fixation have good results. This does not hold true mortality rates were higher for the fractured neck of
for uncemented fixation, but further breakdown by femur group.
bearing type for commonly used uncemented implants
shows that results are acceptable if metal-on-metal The number of revision total hip replacements
bearings are excluded. recorded in the NJR increased to a peak of 10,463 in
2012 and since then has declined steadily to 8,425 in
Metal-on-metal stemmed and resurfacing implants 2016 and 8,073 in 2017. Please note that there may
continue to fail at higher than expected rates and be a small amount of late registrations for 2017 and
their use is now extremely rare. The best performing thus the figure for this year may be revised upward
brand of resurfacing has a failure rate of 7.95% (95% slightly in the next annual report. Aseptic loosening is
CI 7.56-8.37) at ten years. The use of metal-on-metal the most common reason for revision, accounting for
bearings has undoubtedly led to a large excess of nearly half of all cases, followed by pain and instability.
revisions which would not have occurred if alternate
bearings had been used. This has been modelled and Risk of re-revision rate is strongly associated with
published in the Journal of Bone and Joint Surgery. time to first revision; 11.66% (95% CI 10.90-12.46%)
For every 100 MoM hip-resurfacing procedures, we of hips revised within a year of primary surgery are
estimate that there would be 7.8 excess revisions by re-revised within three years. In contrast when the
ten years, and similarly for every 100 stemmed MoM primary lasts at least five years the re-revision rate is
THR procedures that there would be 15.9, which 6.78% (95% CI 6.23-7.38). Re-revision rates up to
4
equates to 8,021 excess first revisions . seven years appear to be independent of the fixation
and bearing of the primary hip replacement.
It is striking to note the high rates of revision for adverse
soft tissue reaction to particulate debris in patients who
3 Smith AJ, Dieppe P, Vernon K, Porter M, Blom AW; National Joint Registry of England and Wales. Failure rates of stemmed metal-on-metal hip
replacements: analysis of data from the National Joint Registry of England and Wales. Lancet. 2012 Mar 31;379(9822):1199-204
4 Hunt LP, Whitehouse MR, Beswick A, Porter ML, Howard P, Blom AW; Implications of Introducing New Technology: Comparative Survivorship Modeling of
Metal-on-Metal Hip Replacements and Contemporary Alternatives in the National Joint Registry. J Bone Joint Surg Am. 2018 Feb 7;100(3):189-196
www.njrcentre.org.uk 101
Part 3
3.4 Outcomes after
knee replacement
National Joint Registry | 15th Annual Report | Knees
This section looks at revision and mortality outcomes orthopaedic surgery involved for each form of
for all primary knee operations performed between 1 replacement can be found in section 3.2. Of special note
April 2003 and 31 December 2017 (inclusive). Patients here is that the NJR data collection process now collects
operated on at the beginning of the registry therefore separate information on medial and lateral unicondylar
had a potential 14.75 years of follow-up. replacements, although this was not the case in the past.
The outcomes of total and partial knee replacement Details of the patient cohort are given in Tables 3.1
procedures are discussed throughout this section, and 3.2 of section 3.2. Figure 3.17 describes the data
hereon referred to as total (TKR) and unicompartmental cleaning applied to produce the total of 1,087,611
(UKR) replacement. Brief details of the type of primary knee procedures included in our analyses.
1,091,636
Initial knee primaries in NJR
3,818
Excluded duplicate
primary procedures
1,087,818
1,087,696
Primary knee replacements with
revision dates that did not
precede the primary date 85
Excluded records where
it was not possible to
trace the NHS number.
Also excludes unknown
or missing gender
www.njrcentre.org.uk 103
Within the whole registry, the 1,087,611 primary IQR 4-39) in 358 units (median=42 cases per
knee joint replacement procedures contributing to unit; IQR 16-92). The number of procedures per
our analyses were carried out by a total of 3,208 consultant over this period may be lower for newly
consultant surgeons working across 461 units. Over qualified consultants, those who may have retired
the last three years (1 January 2015 to 31 December during this period and surgeons with periods of
2017), 303,960 primary knee procedures were inactivity within the coverage of the NJR.
performed by 1,985 consultant surgeons working
across 403 units. Looking at caseload over this The majority of primary knee replacements were
three-year period, the median number of primary carried out on women (females 56.8%; males
procedures per consultant surgeon was 112 (IQR 30- 43.2%). The median age at primary operation was
220) and the median number of procedures per unit 69 (IQR 63-76) years and the overall range was
was 655 (IQR 318-1,020). 7-102 years. For unicompartmental primary knee
replacements, patients were typically six (unicondylar;
Over this three-year period, there have been 272,133 median age 64 years; IQR 57-70) and twelve years
primary total knee replacements performed by younger (patellofemoral; median age 58 years; IQR
1,976 surgeons (median=103; IQR 29-202.5) in 50-67). Osteoarthritis was given as a documented
403 separate units (median=599 cases per unit; reason in 1,058,623 procedures (97.3% of the
IQR 278-944). In the same time period, there have cohort) and was the sole reason given in 1,046,439
been 28,573 primary unicondylar knee procedures (96.2%) of primary knee procedures.
performed by 809 consultant surgeons (median=15;
104 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Table 3.21 shows the breakdown of cases by the followed by 20.7% which were all cemented, posterior
method of fixation, constraint and bearing used. A stabilised and fixed. Within each method of fixation,
breakdown within each method of fixation of the it can be seen that uncemented/hybrid prostheses
percentage of constraint and bearing types used in are mostly unconstrained (cruciate retaining) but
surgery is shown in a separate column. Cemented almost equally likely to have a mobile or fixed bearing.
TKR is the most commonly used operation type Approximately two-thirds (67.0%) of cemented
(85.1% of all primary knee replacements). A further implants are unconstrained (cruciate retaining) and
4.8% were either all uncemented or hybrid TKRs. have a fixed bearing. Unicondylar knee surgery
Most UKRs were unicondylar (8.9% of the total) with typically involves the use of a mobile type of bearing/
the remainder being patellofemoral (1.2%). constraint. A number of primary knee replacements
could not be classified according to their bearing/
More than half of all operations (57.0%) were TKRs constraint (approximately 1.1% of the total cohort).
which were all cemented, unconstrained and fixed,
www.njrcentre.org.uk 105
106
Table 3.22 Percentage* of primary knee replacements by fixation, constraint, bearing and calendar year.
Fixation/bearing/ 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
constraint n= 41,630 n=42,507 n=50,453 n= 67,084 n= 74,505 n=76,522 n= 79,152 n= 82,746 n=86,567 n=86,269 n=96,216 n= 99,123 n=102,660 n=102,177
Total knee replacement
All cemented 81.0 81.7 81.3 81.9 81.9 82.6 84.0 85.4 86.7 87.7 87.4 87.4 87.2 86.7
Cemented and
unconstrained fixed 52.9 52.9 50.4 50.3 51.2 52.8 54.2 56.3 59.0 59.7 60.7 61.6 62.0 61.4
unconstrained mobile 4.2 5.4 6.5 6.4 5.7 4.8 4.1 3.0 2.4 2.2 2.0 1.7 1.8 1.7
posterior-stabilised fixed 20.7 19.6 20.1 20.4 21.0 21.4 21.8 21.7 21.1 21.1 20.5 20.2 19.8 20.0
posterior-stabilised mobile 1.0 1.6 1.9 1.6 1.4 1.4 1.4 1.2 1.1 1.2 1.0 0.8 0.6 0.4
constrained condylar 0.4 0.4 0.3 0.3 0.3 0.3 0.4 0.4 0.6 0.8 1.1 1.2 1.3 1.4
www.njrcentre.org.uk
monobloc polyethylene tibia 0.3 0.4 0.6 0.9 0.8 0.7 1.0 1.6 2.0 2.2 1.9 1.5 1.5 1.6
bearing/constraint
1.5 1.4 1.6 1.9 1.5 1.2 1.2 1.2 0.6 0.5 0.3 0.3 0.3 0.2
unknown
All uncemented 6.6 6.2 6.5 6.5 6.2 5.7 4.7 4.1 3.3 2.5 2.6 2.3 2.0 2.0
Uncemented and
unconstrained fixed 2.5 2.4 2.7 3.0 2.8 2.6 1.8 1.4 1.0 0.7 0.6 0.7 0.7 0.8
unconstrained mobile 3.3 3.2 3.1 3.0 3.0 2.6 2.6 2.4 2.0 1.6 1.6 1.4 1.1 1.0
posterior-stabilised fixed 0.6 0.5 0.5 0.4 0.4 0.4 0.2 0.2 0.2 0.2 0.3 0.2 0.2 0.2
other constraint <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
constraint unknown 0.3 0.2 0.1 0.1 0.1 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
All hybrid 2.8 2.4 1.7 1.4 1.4 1.2 0.9 0.5 0.4 0.4 0.4 0.4 0.4 0.2
Hybrid and
unconstrained fixed 2.3 1.9 1.2 1.0 1.1 1.0 0.7 0.3 0.2 0.2 0.1 0.1 0.1 0.1
© National Joint Registry 2018
unconstrained mobile 0.3 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.3 0.1
posterior-stabilised fixed 0.1 0.1 0.1 0.1 0.1 0.1 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
other constraint <0.1 0.2 0.2 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
constraint unknown <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
Unicompartmental knee replacement
All unicondylar 8.5 8.7 9.3 8.9 9.1 9.0 9.0 8.6 8.2 8.1 8.5 8.9 9.3 10.0
Unicondylar and
fixed 1.7 2.1 2.3 2.0 2.1 2.3 2.7 2.6 3.0 3.3 3.5 3.7 4.1 4.5
mobile 6.7 6.5 6.9 6.7 6.8 6.6 6.3 5.8 5.2 4.8 5.0 5.1 5.2 5.5
constraint unknown 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 <0.1 0.1 <0.1 <0.1 <0.1 <0.1
All patellofemoral 1.0 1.0 1.1 1.4 1.5 1.5 1.4 1.5 1.4 1.2 1.1 1.1 1.1 1.0
All unknown 0.1 0.01
All types 100 100 100 100 100 100 100 100 100 100 100 100 100 100
* Percentage of all primary operations in a particular year which used one of the five fixation methods: cemented, uncemented, hybrid, patellofemoral or unicondylar. Percentages shown represent percentage of
total procedures.
Note: Data from 2003 has been included in 2004, since 2003 was not a complete year.
National Joint Registry | 15th Annual Report | Knees
Table 3.22 shows the annual change in the usage of carried out has been in the use of all uncemented and
primary knee replacements. Overall, more than 80% hybrid total knee replacements over time (now 2.2% of
of all primaries utilised an all cemented fixation method all knee replacements). Each implant of this type now
and since 2003, the share of all implant replacements used has decreased proportionally to less than a third
of this type has increased by about 6%. The main of those figures reported for 2003 (when they were
decline in the type of Figure
primary
3.18.knee replacements
Fixation by year of procedure in primary9.4% of all knee replacements).
knee replacements
90
80
Percentage of primaries
70
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Year of primary
www.njrcentre.org.uk 107
Table 3.23 Age at primary knee replacement by fixation, constraint and bearing type.
Table 3.23 shows the age and gender distribution of Women are also more likely to have a primary TKR;
patients undergoing primary knee replacement. The 57.7%, 51.8% and 55.5% of cemented, uncemented
median age of a person receiving a cemented TKR and hybrid type procedures respectively are carried out
was 70 years (IQR 64-76 years). Patients receiving on female patients. Conversely, unicondylar surgery
UKRs were typically six (unicondylar; median age is performed on a higher proportion of males (53.1%).
64 years; IQR 57-70) and twelve years younger Patellofemoral surgery is predominantly carried out on
(patellofemoral; median age 58 years; IQR 50-67). females (77.5% of patients) who are typically younger
than a TKR or unicondylar patient with a median age at
Over all operation types, a higher percentage of females operation of 58.
(56.8%) than males have had a knee replacement.
108 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Table 3.24 shows the ASA grade and reason for replacement. The majority of cases are performed
knee replacement by gender for all primary knee for osteoarthritis. 1,046,439 (96.2%) of all 1,087,468
replacements. A greater number of females than knee replacements with a reason for primary surgery
males undergo knee replacement and ASA 2 is the recorded in the NJR are performed for osteoarthritis as
most common ASA grade. Only a small number of the sole indication.
patients with a grade greater than ASA 3 undergo knee
www.njrcentre.org.uk 109
3.4.2 First revision after primary knee surgery
A total of 28,717 first revisions of a knee prosthesis surgery records of operations undertaken between
have been linked to NJR primary knee replacement 2003 and 2017.
Figure 3.19 (a) KM estimates of cumulative revision by year, in primary knee replacements
6
2003
2004
5 2005
2006
© National Joint Registry 2018
2007
4
2008
2009
3 2010
2011
2012
2
2013
2014
1 2015
2016
2017
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Years since primary
Figures 3.19 (a) and (b) illustrate temporal changes in the number and timing of revision procedures. The
the overall revision rates using Kaplan-Meier estimates; cumulative probability of a joint being revised at three
procedures have been grouped by the year of the and five years increased for each operative year group
primary operation. Figure 3.19 (a) plots each Kaplan- between 2003 and 2008; the probability of being
Meier survival curve with a common origin, i.e. time revised at three and five years reduced for operations
zero is equal to the year of operation. This illustrates performed between 2009 and 2017. From the peak in
that there was a small increase in revision rates up 2008, the yearly survivorship curves are less divergent,
until 2008 followed by a small decline. i.e. a slowing in the increasing trend.
Figure 3.19 (b) shows the same curves plotted against Possible reasons for a peak in the probability of
calendar time, where the origin of each curve is the revision in the 2008 cohort are: 1) the registry was
year of operation. Figure 3.19 (b) separates each year not capturing the full range and number of operations
allowing changes in failure rates to be clearly identified. taking place in units in England and Wales until
In addition, the revision rate at 1, 3, 5 and 7 years 2008, and 2) there could be bias in terms of the
has been highlighted. If revision rates and timing of general overall health, risk of revision, and other key
revision rates were static across time we would expect characteristics of the patients on record in the NJR in
all failure curves to be the same shape and equally the early years.
spaced; a departure from this indicates a change in
110 www.njrcentre.org.uk
Figure 3.19Figure estimates
(b) KM3.19 cumulative
(b) KMofestimates of revision by year,
cumulative in primary
revision knee
by year, replacements
in primary plotted by year of primary.
knee replacements
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Year of primary
www.njrcentre.org.uk
111
112
Table 3.25 KM estimates of cumulative revision (95% CI) by fixation, constraint and bearing, in primary knee replacements. Blue italics signify that
fewer than 250 cases remained at risk at these time points.
www.njrcentre.org.uk
posterior-stabilised, fixed 224,861 0.46 (0.43-0.49) 1.70 (1.64-1.76) 2.49 (2.42-2.57) 3.96 (3.84-4.09) 4.56 (4.40-4.73) 5.13 (4.85-5.44)
posterior-stabilised, mobile 12,155 0.66 (0.53-0.82) 2.15 (1.90-2.44) 2.91 (2.61-3.25) 4.33 (3.89-4.83) 4.85 (4.28-5.48) 5.83 (4.34-7.82)
constrained, condylar 7,967 0.87 (0.68-1.11) 2.18 (1.83-2.60) 2.88 (2.42-3.43) 4.10 (3.18-5.29) 4.56 (3.37-6.15) 6.64 (3.48-12.46)
monobloc polyethylene tibia 14,554 0.35 (0.26-0.46) 1.35 (1.15-1.57) 1.85 (1.61-2.14) 2.57 (2.17-3.04) 2.90 (2.33-3.60) 2.90 (2.33-3.60)
bearing type unknown 9,336 0.77 (0.61-0.97) 2.40 (2.10-2.75) 3.41 (3.04-3.83) 5.15 (4.63-5.73) 5.89 (5.22-6.65) 6.28 (5.44-7.23)
All uncemented 43,011 0.57 (0.50-0.65) 2.09 (1.96-2.24) 2.87 (2.70-3.04) 4.15 (3.92-4.39) 4.85 (4.55-5.17) 5.61 (5.11-6.16)
unconstrained, fixed 16,471 0.65 (0.54-0.79) 2.37 (2.14-2.62) 3.08 (2.81-3.37) 4.37 (4.01-4.76) 5.13 (4.65-5.66) 6.05 (5.19-7.04)
unconstrained, mobile 22,759 0.51 (0.42-0.61) 1.86 (1.68-2.05) 2.63 (2.41-2.86) 3.72 (3.43-4.04) 4.27 (3.88-4.69) 4.69 (4.11-5.34)
posterior-stabilised, fixed 3,062 0.58 (0.36-0.93) 2.42 (1.90-3.07) 3.46 (2.81-4.26) 6.02 (5.00-7.23) 7.52 (6.15-9.19) 9.96 (7.47-13.22)
other constraint 249 0.81 (0.20-3.21) 2.63 (1.19-5.77) 3.14 (1.51-6.49) 3.14 (1.51-6.49)
bearing type unknown 470 0.64 (0.21-1.98) 1.60 (0.76-3.33) 3.35 (2.00-5.60) 4.74 (2.97-7.51) 5.24 (3.31-8.24) 5.24 (3.31-8.24)
© National Joint Registry 2018
All hybrid 9,157 0.57 (0.43-0.75) 1.84 (1.57-2.15) 2.43 (2.12-2.79) 3.53 (3.12-4.00) 4.08 (3.57-4.66) 4.56 (3.84-5.42)
unconstrained, fixed 6,168 0.48 (0.33-0.69) 1.66 (1.36-2.02) 2.19 (1.85-2.60) 3.18 (2.73-3.70) 3.75 (3.19-4.41) 4.32 (3.50-5.33)
unconstrained, mobile 1,933 0.95 (0.60-1.51) 1.74 (1.22-2.50) 2.26 (1.61-3.17) 3.42 (2.35-4.95) 3.98 (2.61-6.07) 3.98 (2.61-6.07)
posterior-stabilised, fixed 539 0.19 (0.03-1.34) 2.65 (1.54-4.52) 4.01 (2.58-6.23) 5.49 (3.68-8.14) 6.48 (4.13-10.10) 6.48 (4.13-10.10)
other constraint 391 0.26 (0.04-1.85) 2.21 (1.11-4.38) 3.14 (1.75-5.62) 4.90 (2.94-8.11) 4.90 (2.94-8.11)
bearing type unknown 126 1.59 (0.40-6.20) 6.59 (3.35-12.75) 6.59 (3.35-12.75) 10.12 (5.67-17.72) 10.12 (5.67-17.72)
All unicondylar 96,700 1.07 (1.01-1.14) 4.02 (3.89-4.16) 6.11 (5.94-6.29) 11.27 (10.96-11.58) 14.04 (13.59-14.50) 16.89 (16.04-17.78)
fixed 32,870 0.73 (0.64-0.83) 3.60 (3.38-3.84) 5.63 (5.32-5.95) 9.92 (9.37-10.51) 12.82 (11.92-13.79) 14.52 (13.02-16.17)
mobile 62,966 1.25 (1.17-1.34) 4.23 (4.07-4.41) 6.36 (6.14-6.58) 11.76 (11.39-12.14) 14.51 (13.99-15.05) 17.70 (16.70-18.74)
bearing type unknown 864 0.82 (0.39-1.71) 3.97 (2.84-5.54) 5.54 (4.16-7.35) 11.59 (9.19-14.56) 13.66 (10.53-17.62) 13.66 (10.53-17.62)
All patellofemoral 13,378 1.15 (0.98-1.35) 5.95 (5.53-6.40) 9.75 (9.19-10.35) 18.70 (17.70-19.75) 22.44 (21.04-23.92) 24.44 (22.62-26.38)
Note: Blank cells indicate the number at risk is below ten and thus estimates have been omitted as they are highly unreliable.
National Joint Registry | 15th Annual Report | Knees
Table 3.25 shows Kaplan-Meier estimates of the Intervals (95% CI). Results at 14 years have been
cumulative percentage probability of first revision, added, but in general, the group sizes are too small
for any cause, for the cohort of all primary knee for meaningful sub-division, hence many of these
replacements. This is broken down for TKR by knee estimates are shown in blue italics. Estimates in blue
fixation type and sub-divided further within each italics indicate time points where fewer than 250 cases
fixation type by bearing/constraint type and for UKR, remained at risk, meaning that the estimates are less
by bearing/constraint type. The table shows updated reliable. Kaplan-Meier estimates are not shown at all
estimates at 1, 3, 5, 10, 12 and 14 years from the when the numbers at risk fell below ten.
primary operation together with 95% Confidence
Figure 3.20 (a) KM estimates of cumulative revision in primary total cemented knee replacements by
constraint and bearing
Figure 3.20 (a) KM estimates of cumulative revision in primary total cemented knee replacements
by constraint and bearing.
7
6
Cumulative revision (%)
Figures 3.20 (a) to 3.20 (c) illustrate the differences in monobloc polyethylene tibias. The revision rates in
revision rates between the types of knee replacement, cemented TKRs that are posterior stabilised and those
fixation and constraint. It is worth noting the different that have mobile bearings remain higher. The revision
vertical scales between the figures. The results show rates for UKRs remain substantially higher, this is most
the lowest revision rates for unconstrained fixed marked in the patellofemoral replacement group.
bearing cemented TKR and cemented TKR with
www.njrcentre.org.uk 113
Figure 3.20 (b) KM estimates of cumulative revision in primary total uncemented/hybrid
knee replacements by constraint and bearing
Figure 3.20 (b) KM estimates of cumulative revision in primary total uncemented/hybrid knee
replacements by constraint and bearing.
10
Cumulative revision (%)
8
© National Joint Registry 2018
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Years since primary
Number at risk
Unconstrained, fixed 22,639 21,362 20,091 18,924 17,801 16,581 15,164 13,448 11,334 8,767 6,214 3,962 2,435 1,144 351
Unconstrained, mobile 24,692 23,200 21,373 19,313 17,236 15,387 13,208 11,006 8,879 6,885 4,849 3,130 1,853 884 261
Posterior−stabilised, fixed 3,601 3,337 3,082 2,808 2,463 2,205 1,970 1,745 1,473 1,160 858 577 320 149 55
114 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
26
24
22
Cumulative revision (%)
20
18
www.njrcentre.org.uk 115
Figure 3.21 (a) KM estimates of cumulative revision in primary total knee replacements by gender
and age
Figure 3.21 (a) KM estimates of cumulative revision in primary total knee replacements by gender
and age.
Males Females
12
12
© National Joint Registry 2018
10
10
Cumulative revision (%)
8
8
6 6
4 4
Under 55 y
55−59 y
2 2 60−64 y
65−69 y
70−74 y
75−79 y
0 0
80+ y
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
116 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
KM estimates
Figure 3.21 (b)Figure of cumulative
3.21 (b) KM estimates revisionininprimary
revision
of cumulative unicondylar
primary unicondylar knee replacements
knee replacements by by
gender and age
gender and age.
Males Females
30 30
20 20
15 15
10 10
Under 55 y
55−59 y
5 5 60−64 y
65−69 y
70−74 y
75−79 y
0 0
80+ y
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
www.njrcentre.org.uk 117
118
Table 3.26 KM estimates of cumulative revision (95% CI) by gender, age, fixation, constraint and bearing, in primary knee replacements. Blue italics signify that fewer
than 250 cases remained at risk at these time points.
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
1.06 4.39 6.36 11.08 13.19 15.86 0.77 3.89 6.14 10.77 12.74 15.00
All types <55 33,645 47,835
(0.96-1.18) (4.16-4.64) (6.07-6.67) (10.58-11.61) (12.49-13.93) (14.66-17.15) (0.69-0.86) (3.70-4.08) (5.89-6.40) (10.35-11.22) (12.14-13.37) (13.89-16.20)
0.80 3.56 5.02 8.76 10.08 12.32 0.55 2.83 4.47 7.58 9.16 10.67
All cemented <55 22,338 32,098
(0.69-0.93) (3.30-3.84) (4.70-5.37) (8.19-9.36) (9.34-10.88) (10.93-13.88) (0.47-0.64) (2.64-3.04) (4.21-4.75) (7.13-8.06) (8.51-9.87) (9.48-12.00)
0.71 3.19 4.48 8.10 9.37 11.55 0.47 2.41 3.90 6.67 8.21 10.21
unconstrained, fixed <55 14,315 20,810
(0.58-0.86) (2.89-3.53) (4.09-4.90) (7.40-8.87) (8.44-10.39) (9.95-13.38) (0.38-0.58) (2.18-2.65) (3.59-4.23) (6.13-7.26) (7.39-9.12) (8.55-12.16)
1.04 4.30 6.03 9.21 10.55 14.89 0.65 3.06 5.18 8.25 9.63 9.63
unconstrained, mobile <55 1,286 1,582
(0.61-1.79) (3.28-5.62) (4.79-7.58) (7.42-11.41) (8.34-13.31) (9.38-23.19) (0.35-1.21) (2.29-4.10) (4.11-6.51) (6.68-10.16) (7.53-12.28) (7.53-12.28)
0.76 3.99 5.74 10.25 11.88 12.86 0.52 3.41 5.21 9.36 11.46 12.61
posterior-stabilised, fixed <55 5,363 7,877
(0.56-1.05) (3.45-4.61) (5.06-6.52) (9.01-11.65) (10.21-13.81) (10.80-15.27) (0.38-0.71) (2.99-3.89) (4.66-5.83) (8.36-10.49) (9.99-13.12) (10.45-15.18)
posterior-stabilised, 1.37 4.22 5.45 8.23 8.23 1.50 4.86 6.07 8.50 8.50
<55 664 749
mobile (0.71-2.61) (2.91-6.10) (3.92-7.55) (6.02-11.19) (6.02-11.19) (0.83-2.68) (3.51-6.70) (4.53-8.10) (6.37-11.30) (6.37-11.30)
3.06 6.26 8.18 8.18 0.56 1.68 2.40 5.27 5.27
constrained, condylar <55 273 393
(1.54-6.02) (3.72-10.44) (4.90-13.49) (4.90-13.49) (0.14-2.21) (0.70-4.02) (1.03-5.57) (1.71-15.68) (1.71-15.68)
monobloc polyethylene 0.77 4.97 4.97 6.50 6.50 1.04 4.03 4.89 4.89
<55 135 211
tibia (0.11-5.33) (2.26-10.76) (2.26-10.76) (3.06-13.54) (3.06-13.54) (0.26-4.11) (1.94-8.29) (2.45-9.64) (2.45-9.64)
1.68 4.90 7.82 13.14 16.80 1.96 7.03 10.61 13.67 15.62 15.62
bearing type unknown <55 302 476
(0.70-3.99) (2.93-8.14) (5.16-11.78) (9.15-18.66) (11.14-24.90) (1.02-3.72) (5.00-9.86) (8.02-13.97) (10.47-17.76) (11.72-20.66) (11.72-20.66)
0.74 4.37 5.92 8.71 10.47 13.27 0.82 4.15 5.99 8.44 9.27 9.27
All uncemented <55 1,678 1,780
(0.42-1.30) (3.44-5.54) (4.80-7.29) (7.20-10.53) (8.43-12.97) (9.91-17.66) (0.48-1.37) (3.28-5.24) (4.91-7.31) (7.01-10.14) (7.61-11.26) (7.61-11.26)
1.04 4.89 6.48 9.38 11.81 1.08 3.43 4.80 7.48 8.01 8.01
unconstrained, fixed <55 699 674
© National Joint Registry 2018
(0.50-2.18) (3.44-6.93) (4.75-8.82) (7.10-12.34) (8.59-16.12) (0.52-2.26) (2.25-5.22) (3.33-6.89) (5.43-10.27) (5.79-11.02) (5.79-11.02)
0.69 4.18 5.65 8.55 8.55 8.55 0.68 3.99 6.03 8.12 8.68
unconstrained, mobile <55 752 904
(0.29-1.65) (2.90-6.00) (4.10-7.76) (6.35-11.45) (6.35-11.45) (6.35-11.45) (0.31-1.51) (2.85-5.56) (4.55-7.96) (6.22-10.56) (6.57-11.42)
2.80 4.13 6.59 13.87 0.61 7.54 10.49 14.34 18.85
posterior-stabilised, fixed <55 195 0 177
(1.17-6.59) (1.98-8.50) (3.33-12.84) (6.23-29.30) (0.09-4.25) (4.35-12.91) (6.54-16.60) (9.07-22.26) (10.60-32.24)
6.25 12.95
other constraint <55 18 0 4
(0.90-36.77) (3.40-42.67)
8.33 8.33 5.00 5.00
bearing type unknown <55 14 0 21 0
(1.22-46.10) (1.22-46.10) (0.72-30.53) (0.72-30.53)
0.86 3.28 5.90 8.46 8.46 0.48 2.88 4.60 7.51 8.51 11.46
All hybrid <55 349 424
(0.28-2.65) (1.83-5.84) (3.80-9.10) (5.72-12.43) (5.72-12.43) (0.12-1.89) (1.60-5.14) (2.88-7.30) (5.09-10.99) (5.64-12.73) (6.38-20.12)
0.99 3.02 5.69 6.84 6.84 0.79 3.69 5.41 7.54 8.96 12.75
unconstrained, fixed <55 203 256
(0.25-3.90) (1.37-6.60) (3.19-10.04) (4.03-11.51) (4.03-11.51) (0.20-3.11) (1.93-6.97) (3.17-9.14) (4.72-11.93) (5.44-14.57) (6.50-24.18)
1.69 4.22 17.84 5.56
unconstrained, mobile <55 66 0 94 0 0 0
(0.24-11.43) (1.04-16.25) (6.45-43.98) (0.80-33.36)
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
Table 3.26 (continued)
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
3.03 2.38 4.82 7.54
posterior-stabilised, fixed <55 37 0 0 42 0
(0.43-19.63) (0.34-15.72) (1.23-17.94) (2.49-21.65)
9.09 12.12 12.12 4.35 13.46
other constraint <55 33 0 24 0 0
(3.03-25.59) (4.73-29.14) (4.73-29.14) (0.62-27.07) (4.55-36.18)
1.65 5.99 8.96 16.66 21.02 25.21 1.43 6.21 9.90 17.60 21.30 27.47
All unicondylar <55 8,310 9,400
(1.39-1.95) (5.46-6.58) (8.26-9.70) (15.41-18.01) (19.11-23.10) (22.17-28.59) (1.20-1.70) (5.69-6.77) (9.21-10.64) (16.41-18.85) (19.55-23.19) (23.55-31.90)
1.29 5.07 7.99 13.90 19.99 1.00 5.87 9.12 17.01 21.06 25.35
fixed <55 3,512 3,625
(0.96-1.75) (4.30-5.96) (6.95-9.19) (11.91-16.17) (16.29-24.40) (0.71-1.41) (5.06-6.82) (8.01-10.36) (14.88-19.41) (17.78-24.84) (18.97-33.39)
1.89 6.57 9.56 17.98 21.77 27.22 1.70 6.44 10.37 18.02 21.56 28.75
mobile <55 4,712 5,677
(1.53-2.33) (5.85-7.38) (8.65-10.55) (16.41-19.69) (19.54-24.22) (23.47-31.43) (1.39-2.09) (5.78-7.17) (9.49-11.32) (16.60-19.55) (19.49-23.82) (23.85-34.42)
2.33 9.69 12.25 21.38 1.05 5.32 8.71 13.67
bearing type unknown <55 86 98
(0.59-8.98) (4.96-18.46) (6.78-21.60) (12.19-35.91) (0.15-7.24) (2.25-12.32) (4.45-16.67) (7.11-25.39)
2.72 10.21 15.75 23.83 29.36 1.03 6.64 10.59 20.92 23.10 24.85
All patellofemoral <55 967 4,131
(1.85-4.00) (8.31-12.50) (13.28-18.62) (20.17-28.03) (23.81-35.88) (0.75-1.40) (5.85-7.53) (9.53-11.75) (19.02-23.00) (20.72-25.71) (21.60-28.50)
0.68 2.60 3.78 6.19 7.59 8.97 0.48 2.28 3.42 5.70 6.78 7.90
All types 55-64 117,670 141,521
(0.64-0.73) (2.51-2.70) (3.65-3.90) (5.99-6.39) (7.31-7.89) (8.45-9.52) (0.45-0.52) (2.20-2.37) (3.32-3.54) (5.53-5.88) (6.54-7.03) (7.47-8.35)
0.60 2.31 3.28 5.09 6.01 7.23 0.39 1.92 2.80 4.40 5.07 5.93
All cemented 55-64 91,583 115,967
(0.55-0.65) (2.20-2.41) (3.15-3.42) (4.89-5.31) (5.74-6.30) (6.68-7.83) (0.36-0.43) (1.83-2.01) (2.70-2.92) (4.23-4.58) (4.84-5.30) (5.49-6.40)
0.51 2.12 2.99 4.53 5.43 6.62 0.36 1.77 2.50 3.91 4.61 5.43
unconstrained, fixed 55-64 61,924 77,955
(0.45-0.57) (2.00-2.25) (2.83-3.15) (4.29-4.77) (5.11-5.78) (5.98-7.33) (0.32-0.40) (1.67-1.87) (2.37-2.63) (3.71-4.12) (4.33-4.90) (4.90-6.01)
0.76 2.68 3.80 5.94 6.59 6.59 0.52 2.24 3.36 5.14 5.57 8.14
© National Joint Registry 2018
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
119
120
Table 3.26 (continued)
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.92 1.69 2.16 6.67 7.83 0.80 3.30 4.92 8.81 11.18
posterior-stabilised, fixed 55-64 555 385
(0.38-2.19) (0.89-3.23) (1.20-3.87) (4.28-10.30) (4.88-12.46) (0.26-2.44) (1.88-5.74) (3.08-7.81) (6.01-12.83) (7.35-16.81)
2.17 4.62 4.62
other constraint 55-64 47 24 0 0 0
(0.31-14.45) (1.17-17.29) (1.17-17.29)
2.72 4.24 7.41 7.41 7.41 1.72 1.72 1.72
bearing type unknown 55-64 74 68 0 0
(0.69-10.45) (1.38-12.60) (3.14-16.94) (3.14-16.94) (3.14-16.94) (0.24-11.62) (0.24-11.62) (0.24-11.62)
0.49 1.92 2.82 4.57 6.86 6.86 0.58 2.30 3.23 4.41 4.72 4.72
All hybrid 55-64 1,039 1,226
(0.20-1.17) (1.23-2.99) (1.94-4.09) (3.31-6.31) (4.88-9.59) (4.88-9.59) (0.28-1.21) (1.57-3.37) (2.32-4.47) (3.28-5.92) (3.48-6.39) (3.48-6.39)
0.31 1.55 2.53 3.83 5.80 5.80 0.90 2.50 3.34 4.71 5.10 5.10
unconstrained, fixed 55-64 661 783
(0.08-1.22) (0.84-2.86) (1.56-4.10) (2.51-5.81) (3.77-8.87) (3.77-8.87) (0.43-1.89) (1.60-3.89) (2.27-4.90) (3.36-6.60) (3.61-7.19) (3.61-7.19)
0.97 0.97 0.97 0.97 5.47 1.75 2.39 2.39 2.39
unconstrained, mobile 55-64 211 299 0
(0.24-3.81) (0.24-3.81) (0.24-3.81) (0.24-3.81) (1.09-25.15) (0.65-4.64) (0.98-5.75) (0.98-5.75) (0.98-5.75)
2.90 4.47 6.42 2.61 5.40 7.15
posterior-stabilised, fixed 55-64 73 0 79 0
(0.73-11.12) (1.46-13.24) (2.43-16.36) (0.66-10.05) (2.06-13.76) (3.02-16.44)
1.33 5.56 7.13 12.33 2.13 2.13 2.13 2.13
other constraint 55-64 75 54 0
(0.19-9.09) (2.12-14.17) (3.02-16.35) (5.77-25.26) (0.30-14.16) (0.30-14.16) (0.30-14.16) (0.30-14.16)
5.26 5.26
bearing type unknown 55-64 19 0 11
(0.76-31.88) (0.76-31.88)
1.08 4.07 6.10 10.93 14.38 16.61 1.03 4.27 6.64 12.44 15.48 18.16
All unicondylar 55-64 18,377 15,544
(0.94-1.24) (3.77-4.39) (5.71-6.51) (10.28-11.63) (13.37-15.47) (14.82-18.60) (0.88-1.21) (3.93-4.63) (6.20-7.11) (11.70-13.23) (14.42-16.62) (16.46-20.01)
0.54 2.99 5.33 9.66 13.96 14.44 0.74 4.04 6.12 10.99 13.19 15.39
fixed 55-64 6,406 5,171
(0.38-0.77) (2.55-3.51) (4.69-6.07) (8.49-10.98) (11.72-16.57) (12.06-17.25) (0.53-1.02) (3.47-4.69) (5.37-6.97) (9.65-12.50) (11.29-15.38) (12.30-19.16)
1.35 4.61 6.52 11.51 14.79 17.44 1.19 4.39 6.92 13.10 16.46 19.25
mobile 55-64 11,806 10,221
(1.15-1.58) (4.22-5.03) (6.04-7.04) (10.73-12.35) (13.64-16.03) (15.31-19.84) (0.99-1.43) (3.99-4.84) (6.39-7.50) (12.21-14.05) (15.18-17.83) (17.29-21.41)
© National Joint Registry 2018
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
Table 3.26 (continued)
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.65 2.69 4.38 6.14 6.14 0.78 2.12 2.60 2.86 2.86
constrained, condylar 65-74 1,021 1,801
(0.29-1.44) (1.70-4.22) (2.89-6.62) (3.71-10.07) (3.71-10.07) (0.45-1.34) (1.46-3.09) (1.79-3.76) (1.95-4.19) (1.95-4.19)
monobloc polyethylene 0.15 1.49 2.04 2.55 3.51 0.36 1.55 2.12 2.69 2.69
65-74 2,192 3,320
tibia (0.05-0.46) (1.02-2.19) (1.43-2.90) (1.77-3.65) (1.93-6.34) (0.20-0.66) (1.14-2.09) (1.61-2.79) (1.98-3.67) (1.98-3.67)
0.73 2.57 3.12 4.32 4.32 4.32 0.40 1.32 2.33 3.80 4.24 4.24
bearing type unknown 65-74 1,528 1,783
(0.41-1.32) (1.87-3.53) (2.33-4.17) (3.28-5.67) (3.28-5.67) (3.28-5.67) (0.19-0.83) (0.87-2.00) (1.69-3.20) (2.84-5.08) (3.04-5.90) (3.04-5.90)
0.57 1.82 2.43 3.55 4.02 4.83 0.50 2.24 2.94 3.85 4.42 4.93
All uncemented 65-74 8,142 8,186
(0.43-0.77) (1.54-2.15) (2.09-2.82) (3.07-4.09) (3.41-4.74) (3.80-6.13) (0.37-0.68) (1.93-2.61) (2.57-3.36) (3.39-4.38) (3.84-5.09) (4.02-6.05)
0.62 2.36 3.01 4.17 4.17 4.17 0.56 2.81 3.32 4.14 4.46 4.79
unconstrained, fixed 65-74 3,196 2,944
(0.40-0.97) (1.87-2.99) (2.44-3.72) (3.41-5.10) (3.41-5.10) (3.41-5.10) (0.34-0.91) (2.25-3.51) (2.70-4.08) (3.39-5.04) (3.62-5.49) (3.78-6.05)
0.49 1.43 1.88 2.92 3.71 4.64 0.50 1.91 2.73 3.63 4.25 5.00
unconstrained, mobile 65-74 4,234 4,687
(0.31-0.75) (1.10-1.86) (1.48-2.38) (2.33-3.65) (2.81-4.89) (3.23-6.63) (0.33-0.75) (1.54-2.36) (2.27-3.28) (3.02-4.34) (3.46-5.21) (3.55-7.02)
0.96 1.80 2.63 3.38 4.26 0.22 2.27 3.16 4.54 5.50 5.50
posterior-stabilised, fixed 65-74 552 454
(0.40-2.29) (0.94-3.43) (1.49-4.62) (1.97-5.75) (2.37-7.59) (0.03-1.56) (1.19-4.33) (1.80-5.52) (2.70-7.58) (3.18-9.42) (3.18-9.42)
1.27 2.72 2.72
other constraint 65-74 79 27 0 0 0
(0.18-8.65) (0.68-10.47) (0.68-10.47)
5.45 9.66 9.66 1.59 1.59 3.93 7.63
bearing type unknown 65-74 81 0 0 74 0
(2.08-13.87) (4.24-21.20) (4.24-21.20) (0.23-10.74) (0.23-10.74) (0.97-15.20) (2.35-23.24)
0.57 2.01 2.25 3.22 3.49 3.49 0.55 1.73 1.92 2.59 3.01 3.72
All hybrid 65-74 1,597 1,832
(0.30-1.10) (1.41-2.87) (1.60-3.16) (2.37-4.37) (2.53-4.79) (2.53-4.79) (0.30-1.02) (1.21-2.46) (1.37-2.70) (1.90-3.52) (2.17-4.18) (2.36-5.84)
0.27 1.83 2.03 2.87 3.21 3.21 0.32 1.22 1.22 2.05 2.56 3.39
unconstrained, fixed 65-74 1,143 1,257
(0.09-0.82) (1.19-2.83) (1.34-3.07) (1.97-4.16) (2.17-4.75) (2.17-4.75) (0.12-0.85) (0.74-2.02) (0.74-2.02) (1.36-3.08) (1.67-3.93) (1.90-5.99)
1.50 1.50 2.17 3.03 3.03 1.45 3.24 4.21 4.21 4.21
unconstrained, mobile 65-74 279 417
(0.56-3.94) (0.56-3.94) (0.88-5.30) (1.29-6.99) (1.29-6.99) (0.65-3.19) (1.84-5.67) (2.45-7.16) (2.45-7.16) (2.45-7.16)
1.19 3.72 3.72 6.19 1.19 2.47 2.47 2.47
posterior-stabilised, fixed 65-74 87 92 0
© National Joint Registry 2018
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
121
122
Table 3.26 (continued)
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.36 1.00 1.32 1.91 2.08 2.08 0.32 0.91 1.22 1.75 1.95 2.04
All cemented 75+ 117,987 181,083
(0.33-0.40) (0.94-1.07) (1.25-1.40) (1.79-2.03) (1.92-2.24) (1.92-2.24) (0.30-0.35) (0.86-0.96) (1.16-1.28) (1.66-1.84) (1.83-2.08) (1.89-2.19)
0.35 0.96 1.24 1.79 1.95 1.95 0.29 0.87 1.13 1.64 1.77 1.83
unconstrained, fixed 75+ 79,738 118,345
(0.31-0.39) (0.89-1.04) (1.15-1.33) (1.65-1.94) (1.77-2.15) (1.77-2.15) (0.26-0.32) (0.81-0.93) (1.06-1.20) (1.53-1.75) (1.64-1.91) (1.68-2.00)
0.37 0.97 1.61 2.03 2.03 2.03 0.40 1.00 1.41 1.95 2.12 2.12
unconstrained, mobile 75+ 3,874 6,634
(0.22-0.63) (0.70-1.36) (1.22-2.12) (1.55-2.65) (1.55-2.65) (1.55-2.65) (0.27-0.59) (0.77-1.28) (1.13-1.76) (1.57-2.42) (1.69-2.65) (1.69-2.65)
0.38 1.07 1.46 2.16 2.43 2.43 0.35 0.97 1.37 1.97 2.28 2.28
posterior-stabilised, fixed 75+ 28,441 46,243
(0.32-0.47) (0.95-1.21) (1.30-1.63) (1.91-2.45) (2.07-2.84) (2.07-2.84) (0.30-0.41) (0.88-1.07) (1.25-1.50) (1.79-2.16) (2.01-2.58) (2.01-2.58)
posterior-stabilised, 0.59 1.56 1.73 2.18 2.18 0.56 0.98 1.29 1.70 2.87
75+ 1,061 1,657
mobile (0.27-1.31) (0.94-2.58) (1.06-2.83) (1.35-3.53) (1.35-3.53) (0.29-1.07) (0.59-1.63) (0.80-2.05) (1.08-2.65) (1.24-6.58)
1.06 2.27 2.27 2.27 0.87 1.46 1.88 2.23 2.23
constrained, condylar 75+ 841 1,974
(0.53-2.12) (1.33-3.86) (1.33-3.86) (1.33-3.86) (0.53-1.41) (0.96-2.21) (1.24-2.83) (1.40-3.55) (1.40-3.55)
monobloc polyethylene 0.30 1.15 1.42 1.77 0.38 0.74 0.99 1.45 1.45
75+ 2,893 4,201
tibia (0.15-0.59) (0.79-1.68) (1.00-2.04) (1.23-2.54) (0.23-0.64) (0.51-1.08) (0.69-1.41) (0.92-2.27) (0.92-2.27)
0.09 0.59 0.99 1.93 1.93 1.93 0.65 1.33 1.81 2.55 2.95 5.84
bearing type unknown 75+ 1,139 2,029
(0.01-0.64) (0.27-1.32) (0.51-1.90) (1.05-3.55) (1.05-3.55) (1.05-3.55) (0.38-1.12) (0.90-1.97) (1.28-2.56) (1.84-3.54) (2.00-4.32) (2.85-11.78)
0.50 1.30 1.67 2.21 2.21 2.21 0.56 1.31 1.62 1.91 2.45 2.97
All uncemented 75+ 5,193 6,754
(0.34-0.74) (1.01-1.67) (1.32-2.10) (1.77-2.75) (1.77-2.75) (1.77-2.75) (0.41-0.78) (1.06-1.63) (1.32-1.97) (1.58-2.31) (1.87-3.21) (1.98-4.45)
0.57 1.09 1.58 2.01 2.01 2.01 0.75 1.52 1.82 1.88 3.00 3.00
unconstrained, fixed 75+ 2,022 2,597
(0.32-1.03) (0.70-1.68) (1.08-2.30) (1.40-2.90) (1.40-2.90) (1.40-2.90) (0.48-1.18) (1.10-2.09) (1.35-2.46) (1.40-2.52) (1.85-4.83) (1.85-4.83)
0.53 1.34 1.66 2.14 2.14 2.14 0.42 1.20 1.42 1.79 2.03 2.03
unconstrained, mobile 75+ 2,762 3,634
(0.31-0.89) (0.95-1.88) (1.21-2.28) (1.58-2.89) (1.58-2.89) (1.58-2.89) (0.25-0.70) (0.88-1.63) (1.06-1.89) (1.36-2.36) (1.45-2.84) (1.45-2.84)
2.42 2.42 3.83 3.83 0.49 1.04 2.33 4.00 4.00 8.80
posterior-stabilised, fixed 75+ 314 0 430
(1.09-5.31) (1.09-5.31) (1.88-7.71) (1.88-7.71) (0.12-1.93) (0.39-2.75) (1.09-4.94) (2.10-7.55) (2.10-7.55) (2.93-24.80)
© National Joint Registry 2018
2.63 2.63
other constraint 75+ 41 0 9
(0.37-17.25) (0.37-17.25)
3.33 1.20 1.20 1.20 1.20 1.20
bearing type unknown 75+ 54 0 0 0 84
(0.48-21.39) (0.17-8.25) (0.17-8.25) (0.17-8.25) (0.17-8.25) (0.17-8.25)
0.47 0.99 1.42 2.12 2.12 2.12 0.64 1.34 1.59 2.17 2.17 2.17
All hybrid 75+ 1,093 1,597
(0.20-1.13) (0.53-1.83) (0.82-2.45) (1.25-3.59) (1.25-3.59) (1.25-3.59) (0.35-1.19) (0.86-2.06) (1.06-2.39) (1.48-3.16) (1.48-3.16) (1.48-3.16)
0.40 0.97 1.52 2.38 2.38 2.38 0.56 1.14 1.47 2.04 2.04 2.04
unconstrained, fixed 75+ 775 1,090
(0.13-1.25) (0.46-2.03) (0.81-2.81) (1.34-4.21) (1.34-4.21) (1.34-4.21) (0.25-1.24) (0.65-2.00) (0.89-2.43) (1.30-3.22) (1.30-3.22) (1.30-3.22)
1.04 1.04 1.04 1.04 1.14 1.52 1.52 2.58
unconstrained, mobile 75+ 192 375
(0.26-4.11) (0.26-4.11) (0.26-4.11) (0.26-4.11) (0.43-3.00) (0.63-3.63) (0.63-3.63) (0.99-6.61)
2.04 2.04 2.04 4.87 4.87 4.87
posterior-stabilised, fixed 75+ 56 0 73 0
(0.29-13.62) (0.29-13.62) (0.29-13.62) (1.59-14.38) (1.59-14.38) (1.59-14.38)
other constraint 75+ 56 0 0 0 0 40 0 0 0 0
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
Table 3.26 (continued)
Males Females
Age at
Time since primary Time since primary
Fixation/constraint/ primary
bearing type (years) n 1 year 3 years 5 years 10 years 12 years 14 years n 1 year 3 years 5 years 10 years 12 years 14 years
0.90 2.41 3.33 5.88 6.42 10.06 1.11 3.06 4.60 8.15 8.95 12.31
All unicondylar 75+ 7,251 6,507
(0.70-1.16) (2.05-2.84) (2.87-3.86) (5.05-6.85) (5.42-7.60) (5.92-16.83) (0.88-1.41) (2.63-3.55) (4.04-5.23) (7.22-9.19) (7.85-10.20) (8.68-17.31)
0.52 1.57 2.59 4.56 5.62 0.83 2.62 3.80 5.82 6.39
fixed 75+ 2,283 2,002
(0.29-0.94) (1.08-2.28) (1.86-3.61) (3.17-6.54) (3.51-8.95) (0.50-1.38) (1.93-3.56) (2.89-4.98) (4.44-7.62) (4.73-8.61)
1.08 2.76 3.67 6.14 6.57 10.66 1.23 3.24 4.86 8.67 9.56 13.75
mobile 75+ 4,918 4,447
(0.82-1.43) (2.30-3.32) (3.10-4.32) (5.19-7.27) (5.47-7.87) (6.16-18.13) (0.94-1.61) (2.72-3.85) (4.19-5.63) (7.56-9.93) (8.25-11.08) (9.35-19.96)
2.08 2.08 17.28 1.75 3.64 7.71
bearing type unknown 75+ 50 0 58
(0.30-13.88) (0.30-13.88) (7.26-37.97) (0.25-11.81) (0.92-13.81) (2.95-19.31)
© National Joint Registry 2018
Others/unknown 14 19
Note: Total sample on which results are based is 1,087,611 primary knee replacements. Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
www.njrcentre.org.uk
123
Table 3.26 shows gender and age stratified Kaplan- 3.4.3 Revisions after primary knee
Meier estimates of the cumulative percentage
probability of first revision, for any cause, firstly for
replacement surgery by main brands
all cases combined, then by knee fixation/constraint for TKR and UKR
sub-divisions. Estimates are shown, along with 95%
As in previous reports, we have only included those
CI, at 1, 3, 5, 10, 12 and 14 years after the primary
brands that have been used in a primary knee
operation. These refine results in our 2016 report,
replacement in 1,000 or more operations. The figures
but now with larger numbers of cases and therefore
in blue italics are at time points where fewer than 250
generally narrower confidence intervals.
primary knee replacements remain at risk. No results
Unicompartmental knee replacements seem to fare are shown at all where the number had fallen below
worse compared to TKR with the chance of revision at ten cases. No attempt has been made to adjust for
each estimated time point being approximately double other factors that may influence the chance of revision
or more than that of a TKR. The revision rate for so the figures are unadjusted probabilities. In addition,
unicondylar (medial or lateral UKR) knee replacements simple indicators of the age profile and proportion of
is 2.8 times higher than the observed rate for all types male patients who typically receive that implant brand
of knee replacement at 14 years and the revision are shown.
rate for patellofemoral replacement is over four times
higher at 12 and 14 years although less than 250
remain at risk at 14 years. First revision of an implant
is slightly less likely in females than males overall for
the most commonly used fixation method (cemented)
but, broadly, a patient from a younger age group
is more likely to be revised irrespective of gender,
with the youngest group having the worst predicted
outcome in terms of the risk of subsequent revision.
Conversely, female patients are more likely to have a
unicondylar implant revised compared to their male,
age equivalent, counterpart. The reverse pattern is
seen in patellofemoral implant survivorship.
124 www.njrcentre.org.uk
Table 3.27 KM estimates of cumulative revision (95% CI) by total knee replacement brands. Blue italics indicate
that fewer than 250 cases remained at risk at these time points.
Median Time since primary
Number of (IQR) age Percentage
Brand 1
knee joints at primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
All total knee 70 0.41 1.55 2.20 2.70 3.43 4.54
977,488 43
replacements (63-76) (0.40-0.42) (1.52-1.57) (2.17-2.24) (2.66-2.74) (3.37-3.48) (4.40-4.67)
68 0.76 2.37 2.82 2.82
ACS PC 1,070 49
(61-74) (0.38-1.51) (1.56-3.60) (1.90-4.17) (1.90-4.17)
70 0.54 2.11 2.90 3.55 4.22 4.58
Advance MP 8,481 47
(64-76) (0.40-0.72) (1.81-2.46) (2.54-3.33) (3.12-4.05) (3.67-4.85) (3.84-5.45)
Advance MP 69 0.07 1.82 2.98 2.98
1,391 14
Stature (62-75) (0.01-0.52) (1.20-2.75) (2.10-4.22) (2.10-4.22)
72 0.60 2.36 3.08 4.23 5.93 6.60
Advance PS 1,223 45
(66-77) (0.29-1.25) (1.60-3.48) (2.17-4.36) (3.06-5.83) (4.19-8.36) (4.56-9.50)
71 0.29 1.58 2.14 2.70 3.58 5.14
AGC 28,190 42
(64-77) (0.24-0.36) (1.43-1.74) (1.96-2.33) (2.49-2.92) (3.29-3.89) (4.46-5.92)
71 0.31 1.49 2.15 2.70 3.56 5.03
AGC V2 38,843 43
(65-77) (0.25-0.37) (1.38-1.62) (2.00-2.31) (2.54-2.88) (3.34-3.79) (4.57-5.53)
AS Columbus 64 0.19 1.46 1.74
1,203 38
Cemented (58-70) (0.05-0.76) (0.77-2.77) (0.94-3.23)
69 0.36 1.52 3.08
Attune 17,212 43
(62-75) (0.27-0.47) (1.22-1.90) (1.78-5.29)
Columbus 71 0.45 1.71 2.48 2.81 3.73
12,164 44
Cemented (65-77) (0.34-0.59) (1.46-2.00) (2.15-2.86) (2.43-3.24) (3.04-4.56)
E-Motion 67 0.70 2.60 3.46 4.23 4.62
3,215 45
Bicondylar Knee (61-74) (0.46-1.05) (2.08-3.24) (2.83-4.22) (3.49-5.12) (3.78-5.64)
Endo Rotating 76 1.36 3.76 5.23 6.11 9.25
1,293 27
Hinge (68-83) (0.84-2.21) (2.75-5.12) (3.95-6.92) (4.63-8.05) (6.47-13.13)
71 0.40 1.51 2.05 2.51 3.04 3.48
Genesis 2 66,839 42
(65-77) (0.36-0.46) (1.41-1.62) (1.93-2.18) (2.36-2.68) (2.83-3.27) (3.00-4.05)
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 7,376 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
www.njrcentre.org.uk 125
Table 3.27 (continued)
Median Time since primary
Number of (IQR) age Percentage
Brand1 knee joints at primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
© National Joint Registry 2018
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 7,376 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
Table 3.28 KM estimates of cumulative revision (95% CI) by unicompartmental knee replacement brands. Blue
italics indicate that fewer than 250 cases remained at risk at these time points.
Number Median Time since primary
of knee (IQR) age Percentage
Brand1 joints at primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
All
63 1.08 4.27 6.59 8.76 12.23 17.85
unicompartmental 110,078 49
(56-70) (1.02-1.15) (4.14-4.40) (6.42-6.76) (8.55-8.98) (11.93-12.53) (17.07-18.66)
knee replacements
Unicondylar
64 2.17 5.90 7.59 9.96 12.96 16.26
AMC/Uniglide 2,943 50
(57-71) (1.70-2.77) (5.09-6.84) (6.65-8.66) (8.84-11.22) (11.48-14.61) (13.81-19.09)
63 0.92 3.92 5.96 7.54 10.23 13.49
†MG Uni 2,389 54
(56-70) (0.61-1.40) (3.21-4.79) (5.07-7.00) (6.53-8.69) (9.02-11.60) (11.01-16.47)
64 1.14 4.00 6.08 8.05 11.44 17.19
© National Joint Registry 2018
Tables 3.27 and 3.28 show the Kaplan-Meier a primary TKR or primary UKR by implant brand and
estimates of the cumulative percentage probability of bearing/constraint type for those brands/bearing types
first revision, for any reason, of a primary TKR (Table which were implanted on at least 1,000 occasions.
3.27) and primary UKR (Table 3.28) by implant brand. Again, patient summaries of age and gender by brand
Table 3.29 shows Kaplan-Meier estimates of the are also given.
cumulative percentage probability of first revision of
126 www.njrcentre.org.uk
Table 3.29 KM estimates of cumulative revision (95% CI) by fixation, constraint and brand. Blue italics signify
that fewer than 250 cases remained at risk at these time points.
Median
Time since primary
Number (IQR)
of knee age at Percentage
Brand1 joints primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
Total knee replacements
AGC
Cement, unconstrained 71 0.29 1.57 2.12 2.64 3.48 5.10
26,906 42
fixed (64-77) (0.23-0.36) (1.42-1.73) (1.94-2.31) (2.43-2.87) (3.19-3.79) (4.38-5.93)
AGC V2
Cement, unconstrained 71 0.25 1.40 2.04 2.58 3.39 4.84
36,662 43
fixed (65-77) (0.21-0.31) (1.28-1.52) (1.90-2.20) (2.41-2.76) (3.18-3.63) (4.37-5.37)
Uncemented hybrid, 69 1.26 3.39 4.23 4.96 6.49 8.66
1,925 50
unconstrained fixed (63-76) (0.85-1.87) (2.66-4.31) (3.41-5.25) (4.05-6.06) (5.34-7.89) (6.63-11.28)
AS Columbus Cemented
Cement, unconstrained 65 0.12 1.37 1.37
1,067 39
fixed (59-71) (0.02-0.84) (0.67-2.77) (0.67-2.77)
Advance MP
Cement, unconstrained 70 0.53 2.06 2.77 3.45 4.14 4.51
8,239 47
fixed (64-76) (0.39-0.72) (1.76-2.41) (2.41-3.19) (3.01-3.94) (3.58-4.78) (3.76-5.40)
Advance MP Stature
Cement, unconstrained 69 0.07 1.82 2.99 2.99
1,385 14
fixed (62-75) (0.01-0.53) (1.20-2.76) (2.11-4.24) (2.11-4.24)
Advance PS
Cement, posterior- 72 0.60 2.36 3.08 4.23 5.93 6.60
1,223 45
stabilised fixed (66-77) (0.29-1.25) (1.60-3.48) (2.17-4.36) (3.06-5.83) (4.19-8.36) (4.56-9.50)
Attune
Cement, unconstrained 69 0.35 1.84
9,598 44
fixed (61-75) (0.24-0.51) (1.35-2.50)
Cement, unconstrained 69 0.16 0.65 3.29
1,505 41
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
* denotes that this brand is now marketed by Lima.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 6,213 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
127
Table 3.29 (continued)
Median
Time since primary
Number (IQR)
of knee age at Percentage
Brand1 joints primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
†LCS
Uncemented hybrid, 70 0.74 1.86 2.40 2.48 2.67 3.32
1,369 42
unconstrained mobile (62-76) (0.40-1.36) (1.26-2.73) (1.70-3.37) (1.77-3.47) (1.92-3.70) (2.41-4.56)
LCS Complete
Cement, unconstrained 70 0.44 1.62 2.65 3.35 4.23
11,286 42
mobile (64-76) (0.33-0.58) (1.39-1.88) (2.34-2.99) (3.00-3.76) (3.77-4.75)
Uncemented hybrid, 69 0.46 1.72 2.43 2.82 3.30
15,224 46
unconstrained mobile (62-75) (0.36-0.58) (1.52-1.95) (2.18-2.71) (2.54-3.14) (2.96-3.68)
MRK
Cement, unconstrained 70 0.31 1.25 1.67 2.23 2.73 3.16
11,745 43
fixed (64-77) (0.23-0.44) (1.05-1.50) (1.42-1.97) (1.90-2.61) (2.31-3.22) (2.48-4.03)
Maxim
Cement, unconstrained 69 0.23 1.70 2.35 3.20 5.11 9.35
1,325 43
fixed (63-76) (0.07-0.71) (1.12-2.58) (1.65-3.35) (2.35-4.37) (3.90-6.67) (5.64-15.30)
NRG
Cement, unconstrained 70 0.35 1.46 2.46 2.93 3.97
8,306 42
fixed (64-76) (0.24-0.50) (1.21-1.76) (2.11-2.86) (2.53-3.40) (3.29-4.79)
Cement, posterior- 70 0.46 1.74 2.45 2.95 3.73
4,787 44
stabilised fixed (63-77) (0.31-0.70) (1.40-2.16) (2.03-2.95) (2.47-3.51) (3.08-4.51)
Natural Knee II
Cement, unconstrained 70 0.33 1.36 2.17 3.21 3.93 5.32
2,710 41
fixed (64-76) (0.17-0.64) (0.98-1.87) (1.68-2.81) (2.57-4.02) (3.16-4.89) (3.97-7.10)
Nexgen
Cement, unconstrained 70 0.30 1.11 1.61 2.11 2.64 2.99
69,110 43
fixed (63-76) (0.26-0.34) (1.03-1.20) (1.50-1.73) (1.97-2.27) (2.45-2.86) (2.68-3.34)
Cement, posterior- 70 0.44 1.64 2.61 3.39 4.60 5.80
69,265 41
© National Joint Registry 2018
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
* denotes that this brand is now marketed by Lima.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 6,213 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
128
Table 3.29 (continued)
Median
Time since primary
Number (IQR)
of knee age at Percentage
Brand1 joints primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
†Rotaglide +
Cement, unconstrained 71 0.47 2.82 3.63 4.26 5.67 6.70
1,716 43
mobile (64-77) (0.24-0.94) (2.12-3.73) (2.83-4.66) (3.38-5.37) (4.59-6.99) (5.42-8.26)
Scorpio
Cement, unconstrained 71 0.46 1.89 2.65 3.20 4.00 5.58
10,825 41
fixed (64-77) (0.34-0.60) (1.64-2.16) (2.36-2.98) (2.87-3.56) (3.62-4.43) (4.39-7.08)
Cement, unconstrained 69 0.34 2.62 3.80 4.65 5.70
1,176 43
mobile (63-75) (0.13-0.91) (1.84-3.73) (2.83-5.09) (3.56-6.06) (4.42-7.35)
Cement, posterior- 71 0.23 1.58 2.39 3.15 3.98 5.07
6,146 41
stabilised fixed (65-77) (0.14-0.39) (1.29-1.92) (2.03-2.81) (2.73-3.64) (3.49-4.55) (4.40-5.84)
68 0.37 1.71 2.33 2.66 3.43 3.95
Cement, PS mobile 1,374 44
(61-75) (0.15-0.88) (1.14-2.56) (1.64-3.29) (1.92-3.69) (2.53-4.63) (2.92-5.32)
Uncemented hybrid, 71 0.62 1.85 2.53 3.15 4.07 5.15
4,850 45
unconstrained fixed (64-77) (0.44-0.89) (1.50-2.27) (2.12-3.02) (2.68-3.70) (3.49-4.76) (4.31-6.16)
TC Plus
Cement, unconstrained 70 0.74 1.92 2.57 3.04 3.69 6.63
7,976 46
fixed (64-76) (0.58-0.96) (1.64-2.25) (2.24-2.94) (2.68-3.45) (3.25-4.18) (3.97-10.95)
Cement, unconstrained 70 0.54 1.51 2.05 2.48 3.21
5,048 44
mobile (64-76) (0.37-0.79) (1.21-1.90) (1.69-2.50) (2.07-2.97) (2.72-3.80)
Uncemented hybrid, 71 0.49 1.43 2.13 2.46 3.41
2,327 41
unconstrained mobile (64-77) (0.27-0.89) (0.99-2.05) (1.56-2.90) (1.82-3.31) (2.51-4.61)
Triathlon
Cement, unconstrained 70 0.43 1.45 1.96 2.39 2.98
74,134 43
fixed (63-76) (0.38-0.48) (1.35-1.55) (1.83-2.09) (2.22-2.56) (2.71-3.28)
Cement, posterior- 70 0.63 1.78 2.57 2.92 4.17
17,296 41
stabilised fixed (63-76) (0.52-0.77) (1.57-2.02) (2.29-2.88) (2.60-3.28) (3.17-5.47)
Uncemented hybrid, 69 0.70 2.53 3.19 3.45
2,122 50
unconstrained fixed (62-75) (0.40-1.20) (1.79-3.57) (2.30-4.41) (2.46-4.83)
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
* denotes that this brand is now marketed by Lima.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 6,213 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
129
Table 3.29 (continued)
Median
Time since primary
Number (IQR)
of knee age at Percentage
Brand1 joints primary (%) male 1 year 3 years 5 years 7 years 10 years 14 years
Patellofemoral knee replacements
Avon
58 0.77 4.26 7.52 10.34 15.21 20.51
© National Joint Registry 2018
Patello-femoral 5,704 22
(50-68) (0.57-1.05) (3.73-4.87) (6.78-8.33) (9.44-11.33) (13.97-16.55) (18.37-22.85)
FPV
59 0.87 7.00 10.10 13.89 18.00
Patello-femoral 1,619 23
(52-68) (0.52-1.47) (5.83-8.39) (8.66-11.76) (12.09-15.94) (15.35-21.06)
Journey PFJ Oxinium
58 2.15 7.75 12.89 17.88 21.73
Patello-femoral 1,743 23
(50-67) (1.55-2.98) (6.50-9.24) (11.20-14.81) (15.76-20.24) (19.16-24.60)
Sigma HP (PF)
58 2.59 8.74 13.30 17.37
Patello-femoral 1,310 23
(50-66) (1.84-3.64) (7.20-10.59) (11.24-15.69) (14.70-20.47)
Zimmer PFJ
56 0.73 4.82 7.39 10.89
Patello-femoral 2,138 22
(50-65) (0.43-1.23) (3.85-6.04) (6.02-9.06) (8.74-13.54)
† denotes a brand that has been discontinued/withdrawn/not implanted in the last three years.
* denotes that this brand is now marketed by Lima.
1
Brands shown have been used in at least 1,000 primary total knee replacement operations. Excludes 6,213 primary operations where the knee brand was not recorded.
Note: Blank cells indicate the number at risk is below ten and therefore estimates are omitted as they are unreliable.
3.4.4 Revisions for different clinical replacements (medial and lateral UKR), the highest
three incidence rates for reasons for revising the
causes after primary knee replacement implant were ‘other’ indication, aseptic loosening and
Table 3.30 shows the revision incidence rates for each pain, respectively.
reason recorded on the MDS forms for knee revision
In Table 3.31 (page 133), the PTIRs for each indication
surgery, for all cases and then sub-divided by fixation
are shown separately for different time periods from
type and whether the primary procedure was a TKR or
the primary knee replacement, within the first year
an UKR.
from primary operation, and between 1-3, 3-5, 5-7,
For TKRs, the highest PTIRs in descending order, 7-10 and 10+ years after surgery (Note: the maximum
were for revision due to aseptic loosening, ‘other’ follow-up for any implant is now 14.75 years). It is
indications, infection, pain and instability. Revision clear that most of the PTIRs for a particular indication
incidences for pain and aseptic loosening were do vary, especially for infection, aseptic loosening
slightly higher for implants which were uncemented, and pain for different time intervals after surgery.
compared to prosthesis implanted using a hybrid or Infection is most likely to be the reason that a joint is
cemented fixation, but revision for infection was lower. revised in the first year but after seven years or more,
is less likely than other reasons. Conversely, revision
For patellofemoral unicompartmental replacements, between one and three years after surgery is more
the top three reasons for revision were for ‘other’ likely for aseptic loosening and pain, with incidence
indications (including progressive arthritis), pain and rates dropping off for pain later on. Aseptic loosening
aseptic loosening. The first two reasons had the PTIRs continue to remain relatively higher than other
highest incidence rates across all reasons by fixation indicated reasons for revision for implants surviving for
method breakdowns. Similarly, for unicondylar knee longer periods after surgery.
130
Table 3.30 PTIR estimates of indications for revision (95% CI) by fixation, constraint and bearing type.
Pros- Number of revisions per 1,000 prosthesis-years for: Pros- Revision
By fixation, thesis- thesis- per 1,000
constraint and years Peri- years patient-
bearing sub- at risk Dislocation/ Aseptic prosthetic Implant Implant Malalign- Other indi- at risk years for
groups (x1,000) All causes Pain sublaxtion Infection loosening Lysis fracture fracture1 wear1 Instability ment cation2 (x1,000) stiffness3
4.79 0.86 0.18 0.93 1.25 0.24 0.16 0.03 0.28 0.69 0.38 1.15 0.33
All cases 5,854.0 5,398.6
(4.74-4.85) (0.84-0.88) (0.17-0.19) (0.90-0.95) (1.22-1.28) (0.22-0.25) (0.15-0.17) (0.02-0.03) (0.27-0.30) (0.67-0.71) (0.36-0.40) (1.13-1.18) (0.31-0.34)
Total knee replacement
3.78 0.60 0.11 0.99 1.00 0.21 0.14 0.02 0.18 0.64 0.33 0.63 0.34
All cemented 4,898.7 4,531.2
(3.73-3.83) (0.58-0.62) (0.10-0.12) (0.96-1.02) (0.97-1.03) (0.20-0.23) (0.13-0.16) (0.02-0.02) (0.16-0.19) (0.62-0.67) (0.32-0.35) (0.60-0.65) (0.32-0.35)
unconstrained, 3.43 0.58 0.10 0.90 0.83 0.19 0.11 0.02 0.15 0.59 0.32 0.61 0.32
3,203.5 2,962.6
fixed (3.37-3.50) (0.56-0.61) (0.09-0.11) (0.87-0.93) (0.80-0.86) (0.17-0.20) (0.10-0.12) (0.01-0.02) (0.14-0.17) (0.57-0.62) (0.30-0.34) (0.59-0.64) (0.30-0.35)
unconstrained, 4.52 0.82 0.21 1.00 1.38 0.32 0.18 0.04 0.30 0.93 0.44 0.50 0.48
250.6 232.0
mobile (4.27-4.80) (0.72-0.94) (0.16-0.28) (0.88-1.13) (1.24-1.53) (0.26-0.40) (0.13-0.24) (0.02-0.07) (0.24-0.37) (0.81-1.05) (0.37-0.53) (0.42-0.59) (0.40-0.58)
posterior- 4.37 0.56 0.11 1.18 1.35 0.27 0.21 0.02 0.19 0.68 0.34 0.64 0.32
1,208.9 1,115.4
stabilised, fixed (4.26-4.49) (0.52-0.61) (0.09-0.13) (1.12-1.25) (1.29-1.42) (0.24-0.30) (0.18-0.24) (0.01-0.03) (0.17-0.22) (0.64-0.73) (0.31-0.38) (0.60-0.69) (0.29-0.35)
posterior-
4.75 0.87 0.19 0.91 1.14 0.28 0.24 0.05 0.29 0.95 0.21 1.02 0.65
stabilised, 79.2 74.3
(4.29-5.25) (0.69-1.10) (0.11-0.31) (0.72-1.15) (0.92-1.40) (0.18-0.42) (0.15-0.38) (0.02-0.13) (0.19-0.44) (0.76-1.19) (0.13-0.35) (0.82-1.27) (0.49-0.86)
mobile
constrained, 5.64 0.37 0.44 2.69 1.00 0.15 0.37 0.04 0.26 0.77 0.26 0.66 0.35
27.1 25.5
condylar (4.81-6.61) (0.20-0.69) (0.25-0.78) (2.14-3.38) (0.68-1.45) (0.06-0.39) (0.20-0.69) (0.01-0.26) (0.12-0.54) (0.50-1.19) (0.12-0.54) (0.42-1.05) (0.18-0.68)
monobloc
3.28 0.60 0.13 0.85 0.71 0.14 0.24 0.05 0.09 0.57 0.39 0.46 0.35
polyethylene 63.5 62.2
(2.86-3.75) (0.44-0.82) (0.06-0.25) (0.65-1.11) (0.53-0.95) (0.07-0.27) (0.14-0.39) (0.02-0.15) (0.04-0.21) (0.41-0.79) (0.27-0.58) (0.32-0.66) (0.23-0.54)
tibia
bearing type 5.59 0.96 0.21 1.12 1.43 0.26 0.20 0.11 0.39 0.85 0.47 1.18 0.25
65.8 59.2
unknown (5.05-6.19) (0.75-1.22) (0.13-0.36) (0.90-1.41) (1.17-1.75) (0.16-0.42) (0.11-0.34) (0.05-0.22) (0.27-0.58) (0.65-1.11) (0.33-0.67) (0.95-1.48) (0.15-0.42)
All 4.68 0.98 0.18 0.67 1.60 0.29 0.15 0.06 0.30 0.80 0.42 0.80 0.38
© National Joint Registry 2018
292.4 261.9
uncemented (4.44-4.94) (0.87-1.10) (0.14-0.24) (0.58-0.77) (1.46-1.75) (0.23-0.36) (0.11-0.20) (0.04-0.10) (0.24-0.37) (0.70-0.91) (0.36-0.51) (0.70-0.91) (0.31-0.46)
unconstrained, 4.95 0.88 0.10 0.63 1.80 0.27 0.16 0.06 0.27 0.81 0.44 0.97 0.36
116.9 105.5
fixed (4.57-5.38) (0.73-1.07) (0.06-0.18) (0.50-0.80) (1.57-2.06) (0.19-0.39) (0.10-0.25) (0.03-0.13) (0.19-0.39) (0.66-0.99) (0.33-0.57) (0.80-1.16) (0.26-0.49)
unconstrained, 4.22 0.96 0.21 0.69 1.38 0.23 0.12 0.07 0.27 0.71 0.35 0.63 0.38
150.2 134.8
mobile (3.90-4.56) (0.81-1.13) (0.15-0.30) (0.57-0.83) (1.21-1.59) (0.16-0.32) (0.08-0.19) (0.04-0.12) (0.20-0.36) (0.58-0.85) (0.27-0.46) (0.51-0.77) (0.29-0.50)
posterior- 6.60 1.55 0.40 0.80 1.85 0.80 0.30 0.10 0.65 1.20 0.90 1.20 0.46
20.0 17.4
stabilised, fixed (5.57-7.83) (1.09-2.20) (0.20-0.80) (0.49-1.31) (1.34-2.55) (0.49-1.31) (0.13-0.67) (0.03-0.40) (0.38-1.12) (0.80-1.79) (0.57-1.43) (0.80-1.79) (0.23-0.92)
4.99 3.56 0.71 1.43 0.71 0.71 0.71 1.45
other constraint 1.4 0 0 0 0 0 1.4
(2.38-10.46) (1.48-8.56) (0.10-5.06) (0.36-5.70) (0.10-5.06) (0.10-5.06) (0.10-5.06) (0.36-5.81)
bearing type 4.52 0.75 0.25 2.76 0.75 0.25 0.50 1.76 0.50 0.50 0.35
4.0 0 0 2.9
unknown (2.85-7.17) (0.24-2.33) (0.04-1.78) (1.53-4.98) (0.24-2.33) (0.04-1.78) (0.13-2.01) (0.84-3.68) (0.13-2.01) (0.13-2.01) (0.05-2.48)
1
The reason implant failure, as reported on in annual reports up to 2013, has been renamed implant wear as this reflects the wearing down of the implant but distinguishes from the implant itself breaking. The latter cause for revision is now indicated
separately as implant fracture.
2
Other indication now includes arthritis and incorrect sizing. Both these reasons were only asked in MDSv1 and so are associated with primaries which took place in the first few years of the registry with little potential for long term follow-up of the
incidence of revision for these specific clinical reasons.
3
This reason was asked in versions MDSv2, v3 and v6 of the clinical assessment forms for joint replacement/revision surgery and hence, for these reasons, there are fewer prosthesis-years at risk.
131
132
Table 3.30 (continued)
Pros- Number of revisions per 1,000 prosthesis-years for: Pros- Revision
By fixation, thesis- thesis- per 1,000
constraint and years Peri- years patient-
bearing sub- at risk Dislocation/ Aseptic prosthetic Implant Implant Malalign- Other indi- at risk years for
groups (x1,000) All causes Pain sublaxtion Infection loosening Lysis fracture fracture1 wear1 Instability ment cation2 (x1,000) stiffness3
3.84 0.67 0.16 0.92 1.12 0.19 0.11 0.04 0.33 0.69 0.34 0.46 0.21
All hybrid 69.8 56.9
(3.41-4.33) (0.51-0.90) (0.09-0.28) (0.72-1.17) (0.90-1.40) (0.11-0.32) (0.06-0.23) (0.01-0.13) (0.22-0.50) (0.52-0.91) (0.23-0.51) (0.32-0.65) (0.12-0.37)
unconstrained, 3.43 0.59 0.15 0.86 0.98 0.19 0.08 0.04 0.36 0.56 0.31 0.36 0.17
52.1 41.4
fixed (2.97-3.98) (0.42-0.85) (0.08-0.31) (0.64-1.16) (0.74-1.29) (0.10-0.36) (0.03-0.20) (0.01-0.15) (0.23-0.57) (0.39-0.80) (0.19-0.50) (0.23-0.57) (0.08-0.35)
unconstrained, 4.23 0.52 0.21 0.83 1.44 0.21 0.10 0.31 0.93 0.72 0.62 0.24
9.7 0 8.3
mobile (3.11-5.74) (0.21-1.24) (0.05-0.83) (0.41-1.65) (0.86-2.44) (0.05-0.83) (0.01-0.73) (0.10-0.96) (0.48-1.78) (0.34-1.51) (0.28-1.38) (0.06-0.96)
posterior- 5.97 1.00 1.74 1.99 0.25 0.50 1.24 0.25 0.29
4.0 0 0 0 0 3.5
stabilised, fixed (4.00-8.91) (0.37-2.65) (0.83-3.65) (1.00-3.98) (0.04-1.77) (0.12-1.99) (0.52-2.99) (0.04-1.77) (0.04-2.04)
other 4.32 1.66 0.33 0.67 0.33 0.33 0.33 1.33 1.00 0.69
3.0 0 0 0 2.9
constraint (2.51-7.45) (0.69-4.00) (0.05-2.36) (0.17-2.66) (0.05-2.36) (0.05-2.36) (0.05-2.36) (0.50-3.54) (0.32-3.09) (0.17-2.74)
bearing
12.10 2.20 2.20 4.40 1.10 1.10 1.10 3.30
type 0.9 0 0 0 0 0.8 0
(6.70-21.86) (0.55-8.80) (0.55-8.80) (1.65-11.73) (0.16-7.81) (0.16-7.81) (0.16-7.81) (1.06-10.24)
unknown
Unicompartmental knee replacement
12.34 2.66 0.76 0.55 3.29 0.43 0.27 0.04 1.09 1.01 0.68 4.97 0.21
All unicondylar 518.9 479.5
© National Joint Registry 2018
(12.04-12.65) (2.53-2.81) (0.69-0.84) (0.49-0.62) (3.14-3.45) (0.38-0.49) (0.23-0.32) (0.03-0.06) (1.01-1.19) (0.93-1.10) (0.61-0.75) (4.79-5.17) (0.17-0.25)
11.01 2.71 0.11 0.64 3.06 0.38 0.25 0.05 0.97 0.77 0.62 4.33 0.25
fixed 150.5 142.7
(10.50-11.56) (2.45-2.98) (0.07-0.18) (0.52-0.78) (2.80-3.36) (0.29-0.49) (0.18-0.35) (0.03-0.11) (0.83-1.14) (0.64-0.92) (0.50-0.76) (4.01-4.67) (0.18-0.35)
12.90 2.63 1.04 0.52 3.39 0.46 0.28 0.04 1.15 1.10 0.70 5.23 0.19
mobile 361.8 330.8
(12.54-13.28) (2.47-2.81) (0.94-1.15) (0.45-0.60) (3.21-3.59) (0.40-0.54) (0.23-0.35) (0.02-0.06) (1.04-1.27) (1.00-1.21) (0.62-0.79) (5.00-5.48) (0.15-0.25)
bearing type 11.82 3.33 0.15 0.30 2.73 0.15 0.15 0.76 1.36 0.61 5.46
6.6 0 6.0 0
unknown (9.47-14.76) (2.20-5.06) (0.02-1.08) (0.08-1.21) (1.72-4.33) (0.02-1.08) (0.02-1.08) (0.32-1.82) (0.71-2.62) (0.23-1.61) (3.93-7.56)
All 20.21 5.26 0.83 0.42 2.36 0.16 0.18 0.14 1.72 1.00 1.41 11.45 0.54
73.7 69.0
patellofemoral (19.21-21.26) (4.76-5.81) (0.64-1.06) (0.30-0.60) (2.03-2.74) (0.09-0.29) (0.10-0.30) (0.07-0.25) (1.45-2.05) (0.80-1.26) (1.16-1.71) (10.70-12.25) (0.39-0.74)
Others/ 2.03 2.03 2.03
0.5 0 0 0 0 0 0 0 0 0 0 0
unknown (0.29-14.38) (0.29-14.38) (0.29-14.38)
1
The reason implant failure, as reported on in annual reports up to 2013, has been renamed implant wear as this reflects the wearing down of the implant but distinguishes from the implant itself breaking. The latter cause for revision is now indicated
separately as implant fracture.
2
Other indication now includes arthritis and incorrect sizing. Both these reasons were only asked in MDSv1 and so are associated with primaries which took place in the first few years of the registry with little potential for long term follow-up of the
incidence of revision for these specific clinical reasons.
3
This reason was asked in versions MDSv2, v3 and v6 of the clinical assessment forms for joint replacement/revision surgery and hence, for these reasons, there are fewer prosthesis-years at risk.
Table 3.31 PTIR estimates of indications for revision (95% CI) by years following primary knee replacement.
Pros- Pros- Revision
thesis- Number of revisions per 1,000 prosthesis-years for: thesis- per 1,000
Time years Peri- years patient-
since at risk Dislocation/ Aseptic prosthetic Implant Implant Other at risk years for
primary (x1,000) All causes Pain sublaxtion Infection loosening Lysis fracture fracture1 wear1 Instability Malalignment indication2 (x1,000) stiffness3
4.79 0.86 0.18 0.93 1.25 0.24 0.16 0.03 0.28 0.69 0.38 1.15 0.32
All cases 5,854.0 5623.5
(4.74-4.85) (0.84-0.88) (0.17-0.19) (0.90-0.95) (1.22-1.28) (0.22-0.25) (0.15-0.17) (0.02-0.03) (0.27-0.30) (0.67-0.71) (0.36-0.40) (1.13-1.18) (0.31-0.34)
4.61 0.57 0.38 1.71 0.60 0.11 0.27 0.01 0.18 0.55 0.33 0.73 0.31
<1 year 1,030.7 1010.2
(4.49-4.75) (0.52-0.61) (0.34-0.42) (1.63-1.79) (0.56-0.65) (0.09-0.13) (0.24-0.30) (0.01-0.02) (0.16-0.21) (0.51-0.60) (0.30-0.37) (0.68-0.78) (0.28-0.35)
6.63 1.52 0.21 1.25 1.59 0.25 0.12 0.03 0.22 0.99 0.60 1.50 0.56
1-3 years 1,706.0 1666.3
(6.51-6.75) (1.46-1.58) (0.18-0.23) (1.19-1.30) (1.53-1.65) (0.23-0.27) (0.10-0.14) (0.02-0.04) (0.20-0.24) (0.94-1.04) (0.56-0.64) (1.44-1.56) (0.53-0.60)
4.19 0.86 0.10 0.61 1.28 0.24 0.12 0.02 0.21 0.62 0.35 1.06 0.28
3-5 years 1,264.9 1227.5
(4.08-4.31) (0.81-0.91) (0.09-0.12) (0.57-0.66) (1.22-1.35) (0.22-0.27) (0.10-0.14) (0.01-0.03) (0.19-0.24) (0.58-0.67) (0.32-0.38) (1.00-1.11) (0.25-0.31)
3.49 0.50 0.10 0.48 1.17 0.25 0.12 0.03 0.30 0.49 0.26 1.04 0.16
5-7 years 885.2 850.5
(3.37-3.61) (0.45-0.55) (0.08-0.12) (0.43-0.53) (1.10-1.25) (0.22-0.29) (0.10-0.14) (0.02-0.04) (0.27-0.34) (0.45-0.54) (0.23-0.30) (0.98-1.11) (0.13-0.19)
© National Joint Registry 2018
3.53 0.36 0.11 0.35 1.29 0.32 0.18 0.03 0.47 0.54 0.20 1.18 0.11
7-10 years 730.1 683.7
(3.40-3.67) (0.32-0.41) (0.09-0.14) (0.31-0.39) (1.21-1.37) (0.28-0.36) (0.15-0.21) (0.02-0.05) (0.42-0.52) (0.49-0.59) (0.17-0.23) (1.11-1.26) (0.09-0.14)
4.27 0.26 0.12 0.33 1.66 0.37 0.23 0.05 0.92 0.67 0.21 1.38 0.09
10+ years 237.3 185.4
(4.01-4.54) (0.20-0.34) (0.08-0.18) (0.26-0.41) (1.50-1.83) (0.30-0.45) (0.17-0.30) (0.03-0.09) (0.81-1.05) (0.57-0.78) (0.16-0.28) (1.24-1.54) (0.05-0.14)
1
The reason implant failure, as reported on in annual reports up to 2013, has been renamed implant wear as this reflects the wearing down of the implant but distinguishes from the implant itself breaking. The latter cause for revision is now
indicated separately as implant fracture.
2
Other indication now includes arthritis and incorrect sizing. Both these reasons were only asked in MDSv1 and so are associated with primaries which took place in the first few years of the registry with little potential for long term follow-up
of the incidence of revision for these specific clinical reasons.
3
This reason was asked in versions MDSv2, v3 and v6 of the clinical assessment forms for joint replacement/revision surgery and hence, for these reasons, there are fewer prosthesis-years at risk.
www.njrcentre.org.uk
133
3.4.5 Mortality after primary after the primary operation when calculating the
cumulative probability of death (see Survival analysis
knee surgery method note in section 3.2). Of the 1,087,611 records
This section describes the mortality of the cohort up to of a primary knee replacement there were 11,466
14 years from primary operation, according to gender bilateral operations in which the patient had both
and age group. Deaths were updated on 16 February knees replaced on the same day; here the second
2018 using data from the NHS Personal Demographic of the two has been excluded, leaving 1,076,145
Service. For simplicity, we do not take into account procedures of whom 143,099 had died before the end
whether the patient had a first (or further) joint revision of 2017.
Table 3.32 KM estimates of cumulative mortality (95% CI) by age and gender, in primary knee replacement. Blue
italics indicate that fewer than 250 cases remained at risk at these time points
134 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Table 3.32 shows Kaplan-Meier estimates of In total there were 68,148 revisions recorded on 57,207
cumulative percentage mortality at 30 days, 90 days individual patient-sides5 (54,633 actual patients). In
and at 1, 5, 10 and 14 years from the primary knee addition to the 28,717 revised primaries described
replacement, for all cases and by age and gender. previously in this section, there were 28,490 additional
Fewer men than women have had a primary knee revisions for a patient-side for which we have no
replacement and, proportionally, more women than associated primary operation recorded in the NJR.
men undergo surgery above the age of 75. Males,
particularly in the older age groups, had a higher Revisions are classified as single-stage, stage one of
cumulative percentage probability of dying in the short two or stage two of two-stage revisions. Information
or longer term after their primary knee replacement on stage one and stage two of two-stage revisions
operation than females in the equivalent age group. are entered into the database separately, whereas
stage one and stage two revisions in practice have
Note: These cases were not censored when further to be linked. Although not all patients who undergo
revision surgery was undertaken. Whilst such surgery a stage one of two revision will undergo a stage two
may have contributed to the overall mortality, the of two revision. In some cases, stage one revisions
impact of this is not investigated in this section. have been entered without stage two, and vice versa,
making identification of individual revision episodes
3.4.6 Overview of difficult. An attempt has been made to do this later in
knee revisions this section.
5 For 86 patient sides, multiple procedures had been entered on the same operation date, 65 had two procedures on the same date. Details of the
components that had been entered for these cases were reviewed. As a result of this, 161 of the duplicated patient side records with the same operation
date have been dropped and 11 have been reclassified.
www.njrcentre.org.uk 135
Table 3.33 Number and percentage of failures by procedure type and year.
Table 3.33 gives an overview of all knee revision patient-side (discussed later in this section). The
procedures carried out each year since April 2003. increase in the number of operations over time reflects
There were up to a maximum of ten documented the increasing number of at-risk implants prevailing in
revision procedures associated with any individual the database.
136 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Table 3.34 Number and percentage of knee revision by indication and procedure type.
*This reason was not recorded in the earliest phase of the registry; only in MDSvs 2, 3 & 6. The number of joints on which the percentage is based is stated below
the percentage figure.
Table 3.34 shows the stated reasons for the revision documented revision procedure (of any type) to the time
knee surgery. Please note that, as several reasons can at which a second revision procedure was undertaken.
be selected, the reasons are not mutually exclusive For this purpose, we regarded an initial stage one
and therefore column percentages do not add up to followed by either a stage one or a stage two as being
100%. Aseptic loosening is the most common reason the same revision episode and these were disregarded,
for revision, accounting for approximately one third of looking instead for the start of a second revision
single stage revision operations, while other indication, episode. (We counted the maximum number of distinct
pain and instability account for almost a fifth each. Of revision episodes for any patient-side to be eight).
the two-stage revision operations, infection is the main
reason recorded for revision surgery in approximately Kaplan-Meier estimates of the cumulative percentage
four-fifths of either stage one or stage two procedures. probability of having a subsequent revision (re-
revision) were calculated. There were 5,236 re-
3.4.7 Rates of knee re-revision revisions and, for 8,842 cases, the patient died
without having been re-revised. The censoring date for
For a given patient-side, we have looked at the survival the remainder was the end of 2017.
following the first documented revision procedure in the
NJR (n=57,207). In most instances (85.2%), the first
revision procedure was a single stage revision, however
in the remaining 14.8% it was part of a two-stage
procedure. We have looked at the time from the first
www.njrcentre.org.uk 137
Figure 3.22 (a) KM estimates of cumulative re−revision, in linked primary knee replacements
Figure 3.22 (a) KM estimates of cumulative re−revision, in linked revised primary knee replacements.
20
Cumulative re−revision (%)
15
© National Joint Registry 2018
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Years since first revision
Number at risk
57,207 49,684 42,779 36,253 30,358 25,075 19,840 15,467 11,467 8,100 5,267 3,201 1,800 821 257
138 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
35
Cumulative re−revision (%)
30
25
Figure 3.22 (b) shows estimates of re-revision by replacements, after which the rates converge apart
type of primary knee replacement. Patellofemoral from the hybrid TKRs and patellofemoral knee
knee replacements have the lowest risk of re-revision replacement re-revisions, but the numbers at risk are
until seven years, followed by unicondylar knee low and should therefore be interpreted with caution.
www.njrcentre.org.uk 139
Figure 3.22 (c) KM estimates of cumulative re−revision by years since first revision, in linked primary
knee replacements. Figure 3.22 (c) KM estimates of cumulative re−revision by years since first revision, in linked
primary knee replacements
20
Cumulative re−revision (%)
16
© National Joint Registry 2018
12
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 4,918 4,067 3,440 2,873 2,398 1,982 1,586 1,275
First rev. 1−3y 11,663 10,196 8,634 7,258 6,029 4,937 3,773 2,829
First rev. 3−5y 5,399 4,602 3,869 3,201 2,548 1,953 1,328 848
First rev. 5+y 6,737 5,080 3,704 2,533 1,621 1,003 508 223
140 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Figure 3.23 (a) KM estimates of cumulative re−revision in primary cemented TKRs by years since
first revision. Figure 3.23 (a) KM estimates of cumulative re−revision in primary cemented TKRs by
years since first failure
20
Cumulative re−revision (%)
15
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 3,507 2,856 2,371 1,947 1,580 1,270 988 776
First rev. 1−3y 8,182 7,013 5,849 4,822 3,925 3,135 2,348 1,763
First rev. 3−5y 3,550 2,959 2,433 1,965 1,540 1,156 779 480
First rev. 5+y 3,883 2,880 2,061 1,378 869 530 268 111
For those with documented primary knee those who had their first revision within one year of
replacements within the NJR, Figures 3.23 (a) to (e) the initial primary knee replacement, experienced the
show cumulative re-revision rates following the first worst re-revision rates. However, for hybrid TKRs,
revision, according to the main type of primary knee the worst re-revision rates were experienced by
replacement. Each sub-group has been further sub- those who had their first revision within 3 to 5 years
divided according to the time interval from the primary of the initial primary knee replacement; however, the
to the first revision, i.e. less than 1 year, 1 to 3, 3 to 5 numbers were small and therefore the results should
and more than 5 years. For cemented, uncemented, be interpreted with caution.
patellofemoral and unicondylar knee replacements,
www.njrcentre.org.uk 141
Figure 3.23 (b) KM estimates of cumulative re−revision in primary uncemented TKRs by
years since first failure
Figure 3.23 (b) KM estimates of cumulative re−revision in primary uncemented TKRs by years since
first revision.
20
Cumulative re−revision (%)
15
© National Joint Registry 2018
10
0
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 239 211 191 176 160 146 123 103
First rev. 1−3y 581 527 477 436 386 337 267 189
First rev. 3−5y 249 233 208 177 145 108 68 40
First rev. 5+y 324 251 181 135 82 55 26 14
142 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
35
Cumulative re−revision (%)
30
20
15
10
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 51 44 39 32 30 29 27 25
First rev. 1−3y 106 99 95 90 80 72 64 48
First rev. 3−5y 43 38 31 28 21 16 11 6
First rev. 5+y 73 63 48 35 27 18 12 6
www.njrcentre.org.uk 143
Figure 3.23 (d) KM estimates of3.23
Figure cumulative re−revision
(d) KM estimates in re−revision
of cumulative primary patellofemoral knee replacements by
in primary patellofemoral
knee replacements by years since first failure
years since first revision.
20
Cumulative re−revision (%)
15
© National Joint Registry 2018
10
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 146 129 118 97 82 68 52 42
First rev. 1−3y 531 487 411 345 289 242 181 132
First rev. 3−5y 336 302 254 211 174 134 84 54
First rev. 5+y 486 375 273 175 115 66 29 12
144 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Figure 3.23 (e) KM estimates of cumulative re−revision in primary unicondylar knee replacements
by years since first revision.
25
Cumulative re−revision (%)
20
10
0
0 1 2 3 4 5 6 7
Years since first revision
Number at risk
First rev. <1y 975 827 721 621 546 469 396 329
First rev. 1−3y 2,263 2,070 1,802 1,565 1,349 1,151 913 697
First rev. 3−5y 1,221 1,070 943 820 668 539 386 268
First rev. 5+y 1,970 1,510 1,140 809 527 333 172 79
www.njrcentre.org.uk 145
Table 3.35 (a) KM estimates of cumulative re-revision (95% CI).
Number
© National Joint Registry 2018
Table 3.35 (b) KM estimates of cumulative re-revision (95% CI) by years since first revision.
Number of
Time since first revision
Primary in the NJR where first revised
© National Joint Registry 2018
Table 3.35 (b) shows that primary knee replacements needing re-revision at each time point compared with
that fail within the first year after surgery have primaries that last more than five years.
approximately two to three times the chance of
146 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
Table 3.35 (c) KM estimates of cumulative re-revision (95% CI) by fixation and constraint.
Table 3.35 (c) shows cumulative re-revision rates at 1, and constraint. Overall, the worst re-revision rates
3, 5 and 7 years following the first revision for those were demonstrated in those where the initial primary
with documented primary knee replacements within had been a cemented TKR in the short term and an
the NJR, broken down by type of knee replacement uncemented TKR in the longer term.
www.njrcentre.org.uk 147
3.4.8 Reason for knee re-revision
Table 3.36 shows a breakdown of the stated indications reports the indications for the second recorded revision.
for the first revision and for any second revision (note It is interesting to note that infection, dislocation/
the indications are not mutually exclusive). Column (i) subluxation, instability and stiffness are more common
shows the indications for all knee revisions recorded in indications for second revision than first revision. This
the NJR, (ii) reports the indications for the first recorded reflects the complexity and soft tissue elements that
revision, (iii) records the number and percentage of first contribute to the outcome of revision knee replacement.
recorded revisions that were revised, and column (iv)
148 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
www.njrcentre.org.uk 149
Table 3.37 (b) Number of re-revisions by year, stage, and whether or not primary is in NJR.
Tables 3.37 (a) and (b) show that the numbers of 3.4.10 Conclusions
revisions and the relative proportion of revisions with
an associated primary in the NJR increased with time. There are now over 1.08 million primary knee
Approximately 70% of revisions performed in 2017 had a replacements with a maximum follow-up of 14.75
linked primary in the NJR. This is likely to reflect improved years recorded in the NJR making this the largest
data capture over time, improved linkability of records dataset of its kind in the world. Of these, 96% of the
and the longevity of knee replacements with a proportion procedures are performed for osteoarthritis as the
of primaries being revised being performed before NJR only indication. Approximately 90% of the procedures
data capture began or outside the coverage of the NJR. are TKRs, 9% medial or lateral unicondylar knee
replacements and 1% patellofemoral replacements.
3.4.9 90-day mortality after These proportions have remained relatively constant
knee revision over time but the proportion of unicondylar knee
replacements has risen slightly, hitting 10% for the
The overall cumulative percentage probability of first time in 2017. Cemented, unconstrained, fixed
mortality at 90 days after knee revision was lower in bearing TKR remains by far the most common type
the cases with their primaries documented in the NJR of knee replacement followed by cemented, posterior
compared with the remainder (Kaplan-Meier estimates stabilized, fixed bearing TKR. Patients who received
0.56 (95% CI 0.48-0.66) versus 0.89 (0.79-1.01)), which unicondylar or patellofemoral knee replacement were
may reflect the fact that this patient group was younger typically younger than those receiving a TKR. TKR
at the time of their first revision, median age of 67 (IQR and patellofemoral replacement are more likely to
60-74) years, compared to the group without primaries be conducted in females whereas unicondylar knee
documented in the NJR who had a median age of 72 replacement is more likely to be performed in males.
(IQR 65-78) years. The percentage of males was similar
in both groups (45.4% versus 47.2% respectively).
150 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Knees
TKRs with a monobloc polyethylene tibia consistently Infection accounts for the majority of the two-stage
show some of the lowest crude revision rates although revision procedures performed. Only approximately
the numbers at risk in later years is small so the results 6% of revisions for infection that have been carried out
must be interpreted with caution. Cemented TKRs in the NJR to date have been single stage procedures
that are unconstrained with a fixed bearing, as well indicating low usage and take up of this technique.
as being the most common type of TKR, consistently The soft tissue envelope makes single stage revision
show low revision rates in comparison to alternatives; surgery potentially more challenging than in the hip
crude revision rates are approximately 1% lower in which may explain the differences in utilization of a
comparison to cemented unconstrained TKRs with single stage approach.
a mobile bearing, cemented TKRs that are posterior
stabilized with either a fixed or mobile bearing at 10 The risk of re-revision following a revision procedure
years. Age and gender influence the risk of revision is higher than for the risk of revision of a primary TKR
surgery with younger patients and males being more across all types of knee replacement. The risk of re-
likely to undergo revision, it has previously been revision of a revised patellofemoral replacement is
felt that this may explain the higher revision rates slightly lower than the other types of knee with the
observed in UKR. We have continued presentation of rest being broadly similar. This suggests that caution
results divided by gender and age band from the 14th should be used when suggesting that partial knee
annual report and these show the risk of revision of replacement may be considered an interim procedure
a unicondylar knee replacement is at least 1.9 times or a lesser intervention than a TKR as the crude re-
higher in males and 2.3 times higher in females at revision rates are worse than the revision rates for
ten years than a cemented TKR. The risk of revision primary TKR and are broadly similar regardless of
of a patellofemoral replacement is at least 2.7 times the type of the knee replacement implanted at the
higher in males and 2.8 times higher in females than a primary procedure. This area requires further research
cemented TKR across all age groups. The difference to explore the risk of revision in light of the different
in revision rates rises from age less than 55 up to the demographics in these groups. The risk of re-revision
65-74 age group and declines again in the over 75s. is higher for those revised after a shorter period of
time following the primary and is associated with the
The most common causes of revision across all indication for revision. This suggests that not all of the
primary knee replacements were aseptic loosening, processes that lead to revision are the same and some
‘other’ indications, infection, pain and instability. For are more aggressive than others with consequences
uncemented TKRs, the incidence of revision for pain beyond the initial revision.
and aseptic loosening were higher but the risk of
revision for infection lower than for cemented TKR. Knee replacement remains a safe procedure with low
For unicondylar knee replacements, the highest risk rates of perioperative mortality. The rates of mortality
of revision was for ‘other’ indication, a proportion are, unsurprisingly, higher for males than for females.
of which are accounted for by progression of The average age of a patient undergoing TKR is 70
osteoarthritis elsewhere in the knee but as this is also years, just over 50% of males and 38% of females
the second most common indication selected by in the 70-74 age bracket will have died 14 years
surgeons for TKR, this clearly does not account for all after their knee replacement. This means that for the
of these cases. Aseptic loosening, pain and implant average patient undergoing a knee replacement, their
wear are the next three most common causes and knee replacement should last them for the rest of their
are all higher than for TKR but the risk of revision for life without the need for revision surgery.
infection is lower.
www.njrcentre.org.uk 151
Part 3
3.5 Outcomes after
ankle replacement
National Joint Registry | 15th Annual Report | Ankles
Table 3.38 Descriptive statistics of ankle procedures performed by consultant and unit by year of surgery.
Year of primary
Number of primary replacements during
each year ≤20101 2011 2012 2013 2014 2015 2016 2017
Operations (n) 417 523 583 552 546 613 719 734
Units (n) 111 128 145 133 137 143 140 139
Mean number of primary replacements
3.8 4.1 4 4.2 4 4.3 5.1 5.3
per unit
1
Includes 14 operation dates prior to 2010.
Table 3.38 shows an increasing number of cases eight-year period of data capture. The maximum
reported annually over the eight year period. This could number of procedures for any consultant was 264.
represent improved compliance or the reporting of a Similarly, the total number of units involved was 247;
true increase in caseload. A total of 244 consultants 63 (25.5%) of which carried out twenty or more over
carried out these primary procedures; 75 (30.7%) of the eight-year time period. The maximum number of
them entered twenty or more procedures over the procedures carried out by any unit was 331.
www.njrcentre.org.uk 153
Table 3.39 Numbers (%) of primary ankle replacements by ankle brand.
Zenith 922 (19.7) 78 (18.7) 109 (20.8) 126 (21.6) 133 (24.1) 152 (27.8) 158 (25.8) 108 (15.0) 58 (7.9)
Box 595 (12.7) 23 (5.5) 29 (5.5) 45 (7.7) 51 (9.2) 84 (15.4) 133 (21.7) 124 (17.2) 106 (14.4)
Salto 303 (6.5) 23 (5.5) 29 (5.5) 40 (6.9) 45 (8.2) 56 (10.3) 55 (9.0) 47 (6.5) 8 (1.1)
Hintegra 275 (5.9) 15 (3.6) 18 (3.4) 35 (6.0) 65 (11.8) 46 (8.4) 55 (9.0) 32 (4.5) 9 (1.2)
Star 428 (9.1) 16 (3.8) 29 (5.5) 31 (5.3) 35 (6.3) 60 (11.0) 82 (13.4) 84 (11.7) 91 (12.4)
Rebalance 60 (1.3) 0 (0) 4 (0.8) 13 (2.2) 13 (2.4) 6 (1.1) 4 (0.7) 13 (1.8) 7 (1.0)
Inbone2 155 (3.3) 0 (0) 0 (0) 2 (0.3) 4 (0.7) 22 (4.0) 20 (3.3) 58 (8.1) 49 (6.7)
Infinity2 710 (15.1) 0 (0) 0 (0) 0 (0) 0 (0) 28 (5.1) 96 (15.7) 212 (29.5) 374 (51.0)
AKILE 22 (0.5) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 4 (0.7) 8 (1.1) 10 (1.4)
TARIC 1 (0) 0 (0) 0 (0) 1 (0.2) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)
Not known 85 (1.8) 3 (0.7) 8 (1.5) 4 (0.7) 4 (0.7) 5 (0.9) 6 (1.0) 33 (4.6) 22 (3.0)
Total 4,687 (100) 417 (100) 523 (100) 583 (100) 552 (100) 546 (100) 613 (100) 719 (100) 734 (100)
1
Includes 14 operation dates prior to 2010.
2
In 2016 and earlier years, 49 Inbone and 330 Infinity implants were classified as cemented by the manufacturer and NJR. In 2017 this was changed to uncemented.
Table 3.39 shows an overall breakdown of brands used modular exchange. Therefore, from now on, any
and further breakdowns by year of primary operation. subsequent procedure in which an implant (including
Please note that 14 procedures had dates of operation the polyethylene liner in a mobile bearing implant)
before 2010 (one in 2006, four in 2008 and nine in is added, removed or exchanged is considered
2009) and these have been combined with those a revision procedure and should be recorded on
performed in 2010 for the purposes of reporting. The an NJR A2 MDS form. A DAIR without a modular
most common brand overall was Mobility, which was exchange should also be recorded as a revision on
used in just under a quarter of the procedures overall. an NJR A2 MDS form. Only 211 (4.5%) of the 4,687
Use of the Mobility began to decline from 2012 and in primary procedures had a linkable NJR A2 MDS form
June 2014, it was withdrawn from the market. In 2017, completed to indicate revision before the end of 2017.
the most common brand used was the Infinity (51.0%), The first revisions shown here include 29 conversions
followed by the Box (14.4%) and the Star (12.4%). to arthrodesis but no amputations have been
recorded. These small numbers likely reflect a failure to
3.5.2 Revisions after primary record removal of the prosthesis during a conversion
ankle surgery to fusion or an amputation procedure as a revision in
line with the accepted definition and mandated by the
From June 2018 the NJR’s minimum dataset Department of Health.
(version 7) for ankle revisions includes Debridement
and Implant Retention (DAIR), with or without
154 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Ankles
Table 3.40 KM estimates of revision (95% CI) after primary ankle replacement, by gender and age. Figures in blue
italics signify time points where fewer than 250 patients remain at risk.
www.njrcentre.org.uk 155
Table 3.41 Indications for the 211 (first) revisions following primary ankle replacement.
Note: these are not mutually exclusive.
Indication Number
Infection High suspicion (e.g. pus or confirmed micro) 11
Low suspicion (awaiting micro/histology) 36
Aseptic loosening 1
Tibial component 70
Talar component 75
Lysis2
Tibia 18
© National Joint Registry 2018
Talus 26
Malalignment 31
Implant fracture 3
Tibial component 4
2
Talar component 4
Implant fracture Meniscal component 3
Wear of polyethylene component 13
Meniscal insert dislocation 4
Component migration/dissociation 15
Pain (undiagnosed) 54
Stiffness 23
Soft tissue impingement 14
Other indication for revision 33
1
46 patients had aseptic loosening of both tibial and talar component.
2
Eleven patients had lysis of both tibial and talar component.
3
Two patients had implant fracture of both tibial and talar component.
4
In the 14th Annual Report 2017, three cases of tibial component fracture were reported, the submitting unit has edited one of the records leading to a reduction to
two tibial complete fractures in this report.
Table 3.41 shows the reasons for revision of 3.5.3 Mortality after primary
ankle replacements, with loosening and pain the
commonest reasons.
ankle replacement
Our analysis excluded one procedure where the NHS
The British Orthopaedic Foot and Ankle Society
number was untraceable (and hence the age could not
(BOFAS) concurs that the small number of revisions
be validated) and also excludes the second of each
reported may indicate under-reporting of the revision
of the five bilateral procedures. Among the remaining
procedures as these figures are lower than published
4,681, a total of 218 patients had died before the end
data in the literature. BOFAS and the NJR encourage
of 2017.
surgeons to complete A2 MDS forms where relevant
and wishes to remind surgeons and hospitals that
this is a mandated requirement and that all revisions,
conversion of an ankle replacement to an arthrodesis,
and amputations require the completion of an NJR A2
MDS form.
156 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Ankles
Table 3.42 KM estimates of mortality (95% CI) after primary ankle replacement, by gender and age. Figures in
blue italics signify time points where fewer than 250 patients remain at risk.
1
Some patients had operations on the left and right side on the same day. The second of bilateral operations performed on the same day were excluded.
Table 3.42 shows the estimated cumulative first revisions even smaller, although we believe
percentage probability of death at different times after that there is under-reporting of revision procedures,
surgery by gender and age at primary. Unsurprisingly, making outcome analysis difficult.
earlier death was associated with male gender
and older age. Overall the cumulative percentage A total of 69% of consultant surgeons and 74%
probability of death was 0.13 (95% CI 0.06-0.29) at of units have submitted less than twenty primary
90-days; 0.68 (95% CI 0.47-0.98) at 1 year; 3.27 procedures in the eight years the NJR has been
(95% CI 2.71-3.94) at 3 years; and 6.35 (95% CI capturing data. Since the withdrawal of the Mobility
5.45-7.39) at 5 years. implant in 2014 the fixed bearing Infinity implant has
gained rapid popularity to become the market leader.
3.5.4 Conclusions The cumulative percentage probability of 90-day
mortality following primary ankle surgery is very low.
The collection of data relating to ankle primary
operations only began in 2010 and hence total
number of primaries remain small and numbers of
www.njrcentre.org.uk 157
Part 3
3.6 Outcomes
after shoulder
replacement
National Joint Registry | 15th Annual Report | Shoulders
Table 3.43 Numbers of primary shoulder replacements (elective and acute trauma), by year with percentages of
each type.
Year of primary
2012 2013 2014 2015 2016 2017
All years N(%) N(%) N(%) N(%) N(%) N(%)
All cases 30,720 (100) 2,563 (100) 4,373 (100) 5,261 (100) 5,641 (100) 6,356 (100) 6,526 (100)
Humeral
6,207 (20.2) 862 (33.6) 1,258 (28.8) 1,256 (23.9) 1,043 (18.5) 991 (15.6) 797 (12.2)
hemiarthroplasty
Resurfacing 2,479 (8.1) 457 (17.8) 562 (12.9) 523 (9.9) 366 (6.5) 356 (5.6) 215 (3.3)
Table 3.43 demonstrates that the number of primary performed on women (women 70.7%; men 29.3%).
shoulder replacements has continued to increase The median age at the primary operation was 73 years
year by year and gives a breakdown by the type of (IQR 67-79 years) overall, with a range of 17-99 years.
replacement. The majority of the replacements were
www.njrcentre.org.uk 159
For humeral components, a stemmed component shoulder replacements in 2017. The use of humeral
is defined as a humeral component in which hemiarthroplasty continues to decrease while the use
any part enters the diaphysis, while a stemless of total shoulder replacement remains stable. Stemless
humeral component is defined as being completely humeral components for conventional replacements
confined to the metaphysis with no part entering have only increased a small amount in the last three
the diaphysis. The number of reverse polarity years, while stemless components for reverse polarity
total shoulder arthroplasties continues to increase replacements have not gained popularity.
annually, accounting for 53% of all primary
Table 3.44 Numbers of units and consultant surgeons providing primary shoulder replacements over the last
five years, 2013-2017.
Number of Number of
© National Joint Registry 2018
Primary shoulder replacements over the last five The reasons given for the elective cases are
years were undertaken by 712 consultant surgeons documented in Table 3.45. The reasons entered were
working across 390 units. A breakdown of the not all mutually exclusive, in some cases more than
numbers of units and consultants for each year, one indication was recorded on the MDS. Amongst
together with their number of primaries, is shown in these 27,990 cases, 1,636 (5.8%) had two or more
Table 3.44. reasons stated, the most common combinations
included osteoarthritis together with cuff tear
Table 3.45 details the indications for the primary arthropathy (670).
operation, for the cases overall and with further sub-
division by type of procedure. Hemiarthroplasty is being used across all indications
including trauma, while total shoulder replacement
Acute trauma accounted for 2,730 cases, these have is used mainly for osteoarthritis. Interestingly reverse
been separated from the remaining 27,990 elective polarity shoulder replacement is in common use
cases. Please note that 91 of the 2,730 acute trauma across all indications, not just cuff tear arthropathy
cases had another reason(s) stated in addition to and acute fracture.
acute trauma; the most common reasons being
osteoarthritis (33) and trauma sequelae (30).
160 www.njrcentre.org.uk
Table 3.45 Reasons for main types of primary shoulder replacements.
Stemmed 9 5,856 5,443 (93.0) 20 (0.3) 116 (2.0) 215 (3.7) 154 (2.6) 60 (1.0)
Reverse polarity total shoulder
1,438 11,363 3,631 (32.0) 6,748 (59.4) 1,097 (9.7) 439 (3.9) 223 (2.0) 241 (2.1)
replacement
Stemless 0 109 38 (34.9) 67 (61.5) 3 (2.8) 2 (1.8) 4 (3.7) 0 (0)
Stemmed 1,438 11,254 3,593 (31.9) 6,681 (59.4) 1,094 (9.7) 437 (3.9) 219 (2.0) 241 (2.1)
Pyrocarbon Ball 0 15 12 (80.0) 0 (0) 1 (6.7) 0 (0) 2 (13.3) 0 (0)
Unclassifiable 144 2,720 1,621 (59.6) 714 (26.3) 200 (7.4) 135 (5.0) 76 (2.8) 132 (4.9)
*Percentages based on the total numbers of elective cases; note the listed reasons are not mutually exclusive in the sense that more than one reason could have been stated but this was only 5.8% for elective cases.
**Includes 41 metastatic cancer/malignancies that have only been documented separately since November 2014, when MDSv6 was introduced.
www.njrcentre.org.uk
161
Table 3.46 Gender and age at primary for the main types of primary shoulder replacements. These are shown
separately for acute trauma and elective cases.
*Percentages not shown where n<10; **IQR=Inter-quartile range, i.e. 25th to 75th centile – not given where number is small; ***Range is lowest – highest.
162 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
Figure 3.31 (a) gender and age distribution of elective shoulder primaries for
humeral
Figure 3.24 (a) Gender and agehemiarthroplasty
distribution of elective shoulder primaries for humeral hemiarthroplasty.
Male
800
600
200
Frequency
0
Female
800
600
400
200
0
10 20 30 40 50 60 70 80 90 100
Age at primary
Figure 3.31 (b) gender and age distribution of elective shoulder primaries for
total conventional shoulder arthroplasty
Figure 3.24 (b) Gender and age distribution of elective shoulder primaries for total conventional
shoulder replacement.
Male
1500
1000
© National Joint Registry 2018
500
Frequency
0
Female
1500
1000
500
0
10 20 30 40 50 60 70 80 90 100
Age at primary
www.njrcentre.org.uk 163
Figure 3.31 (c) gender and age distribution of elective shoulder primaries for
reverse polarity total shoulder arthroplasty
Figure 3.24 (c) Gender and age distribution of elective shoulder primaries for reverse polarity
total shoulder replacement.
Male
3000
© National Joint Registry 2018
2000
1000
Frequency
0
Female
3000
2000
1000
0
10 20 30 40 50 60 70 80 90 100
Age at primary
Figures 3.24 (a) to (c) illustrate the distributions by Table 3.47 lists the main stemmed brands used in
gender and age groups of the elective patients, primary shoulder procedures. The table shows the
according to the primary patient procedure. Over the total numbers recorded in the registry since April 2012
last three years, the percentage of elective patients as well as the numbers within the last twelve months
under 55 years having shoulder replacements was (i.e. in 2017). The latter are further sub-divided into
6.5%, 6.1% and 5.9% respectively. As some younger acute trauma and elective cases. The numbers of
patients of both genders are undergoing these types elective cases are further divided into the types of
of procedures, the age group revision rates later in this implant. Finally, Tables 3.48 (a) and (b) show similar
section have been modified to help provide information tables for stemless brands and the resurfacing brands
on younger age related revision rates. used in primary shoulder replacements.
164 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
www.njrcentre.org.uk 165
Table 3.48 Stemless brands and resurfacing brands used in primary shoulder replacements, shown separately.
(a) Stemless brands
Resurfacing number of Total number Total number Total number of humeral hemi- total shoulder
brands primaries 2017 of primaries primaries arthroplasty replacement
Aequalis Resurfacing 270 20 0 20 17 3
Arrow 43 3 0 3 3 0
Arthrosurface 1 0 - - - -
Copeland 1,419 122 0 122 122 0
Epoca 434 34 0 34 6 28
Equinoxe 27 14 0 14 7 7
Global CAP 505 28 0 28 28 0
SMR 128 14 0 14 14 0
Vaios 56 18 0 18 18 0
Total 2,883 253 0 253 215 38
166 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
Glenoid components used in total conventional sub-brands are fully captured so that the performance
shoulder replacement of these different glenoid components and their
fixation methods can be analysed.
Many manufacturers continue to have more than
one glenoid type as an option for a conventional total Table 3.49 highlights the current glenoid brands
shoulder replacement and these also now include recorded in the NJR and those used in the last year.
augmented implants. More work is required between
the NJR and implant manufacturers to ensure these
www.njrcentre.org.uk 167
3.6.2 Revisions after primary shoulder Kaplan-Meier estimates of the cumulative percentage
revision at 1, 2, 3, 4 and 5 years after the primary
replacement surgery operation, together with 95% Confidence Intervals (CI),
A total of 860 linked shoulders were subsequently for all cases are shown in Table 3.50, together with a
revised, 71 of these have had a further re-revision. separation into acute trauma and elective cases.
Table 3.50 KM estimates of cumulative revision (95% CI) for primary shoulder replacement for acute trauma and
© National Joint Registry 2018
elective cases. Figures in blue italics signify time points where fewer than 250 patients remain at risk.
5
Cumulative revision (%)
4
© National Joint Registry 2018
0
0 1 2 3 4 5
Years since primary
Number at risk
Acute trauma 2,730 1,979 1,348 841 431 118
Elective 27,990 21,515 15,376 10,167 5,548 1,979
Figure 3.25 further compares the acute trauma and after which time point there were too few cases for a
elective cases for all time points up to five years, meaningful summary to be presented.
168 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
Table 3.51 KM estimates of cumulative revision (95% CI) for elective shoulder primaries by gender and age.
Figures in blue italics signify time points where fewer than 250 patients remain at risk.
Males Females
Age at
A further breakdown by gender and age of the compared to females. While this remains the case at
cumulative percentage of revisions in the elective three years in the 56-64 year group, the results in the
cases is shown in Table 3.51. It demonstrates a worse new category of even younger patients under 55 years
outcome up to four years for men and a trend to indicate equal gender revision rates of over 7% at
worse outcome in younger patients of either gender. three years. The acute trauma group remains too small
We have indicated in previous reports that results for a similar analysis to be conducted.
were worse at three years in males under 65 years
www.njrcentre.org.uk 169
Figure 3.24 KM estimates of cumulative revision for primary shoulder replacement surgery,
by type of procedure in elective cases only
Figure 3.26 KM estimates of cumulative revision for primary shoulder replacement, by type of
procedure in elective cases only.
12
10
Cumulative revision (%)
6
© National Joint Registry 2018
0 1 2 3 4 5
Years since primary
Number at risk
Resurfacing HA 2,474 2,235 1,804 1,382 878 368
Stemless HA 797 630 452 316 168 54
Stemmed HA 1,797 1,506 1,162 816 486 192
Resurfacing TA 404 362 284 199 126 40
Stemless TA 2,564 1,852 1,232 739 359 124
Stemmed TA 5,856 4,604 3,377 2,207 1,186 413
Stemless RTA 109 82 56 32 17 6
Stemmed RTA 11,254 7,932 5,239 3,229 1,586 508
170 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
Table 3.52 KM estimates of cumulative revision (95% CI) for elective shoulder primary by main type of procedure.
Figures in blue italics signify time points where fewer than 250 patients remain at risk.
In Figure 3.26 and Table 3.52, the elective cases have the performance of those replacements not revised in
been sub-divided by the type of procedure. each sub-group (see comments in section 3.6.3).
The initially worse cumulative revision rate observed With the evolution of stemless humeral implants, these
and reported in previous annual reports for the reverse have been included as separate sub-groups in this
geometry shoulder replacements during the first two year’s report. Stemless reverse shoulder replacement
years after the primary replacement has stabilised, and makes up a small sub-group and while the numbers
demonstrates a revision rate similar to total shoulder are small and make statistical conclusions less robust,
replacement at years four and five. Hemiarthroplasty there are much higher revision rates at all time points.
operations (including resurfacing) have higher revision While this is the first time this has been observed in
rates by year three and this continues in years four the NJR annual report, it is encouraging that this type
and five compared to total shoulder and reverse of procedure has not become mainstream and seems
shoulder replacement. Stemmed hemiarthroplasty has to have diminished in numbers, possibly due to such
the lower revision rate at four years and five years of early revision problems being observed by surgeons
the hemiarthroplasty group but the numbers at risk at and implant manufacturers. Conversely, stemless
year five in this group is low and therefore estimates conventional total shoulder replacement seems to
may be unreliable. Clinically, this may be in relation to display the lowest revision rates at years 3, 4 and 5.
the greater ease with which resurfacing and stemless
hemiarthroplasty can be revised but data on this is not Finally, this report currently does not yet look at revision
collected in the NJR. NJR PROMs data would provide rates by combined age group and implant types.
more detailed information regards these revisions and
www.njrcentre.org.uk 171
172
Table 3.53 Numbers of first revisions for each type of primary shoulder replacement and indications for revision. Acute trauma and elective cases are shown separately.
(a) Acute trauma cases only
www.njrcentre.org.uk
Total shoulder
9 0 - - - - - - - -
replacement
Reverse polarity
total shoulder 1,438 18 2 (11) 10 (56) 0 (0) 2 (11) 1 (6) 0 (0) 1 (6) 3 (17)
© National Joint Registry 2018
replacement
Pyrocarbon Ball 0 0 - - - - - - - -
Unclassifiable 144 1 0 0 1 0 0 1 0 0
Table 3.53 Numbers of first revisions for each type of primary shoulder replacement and indications for revision. Acute trauma and elective cases are shown separately.
(b) Elective cases only
Stemmed 5,856 143 9 (6) 52 (36) 74 (52) 16 (11) 4 (3) N/A 2 (1) 2 (1) 22 (15)
Reverse polarity
total shoulder 11,363 261 48 (18) 86 (33) 7 (3) 32 (12) 31 (12) N/A 15 (6) 2 (1) 63 (24)
replacement
Stemless 109 10 1 (10) 3 (30) 1 (10) 2 (20) 0 (0) N/A 0 (0) 0 (0) 4 (40)
Stemmed 11,254 251 47 (19) 83 (33) 6 (2) 30 (12) 31 (12) N/A 15 (6) 2 (1) 59 (24)
Pyrocarbon Ball 15 0 - - - - - - - - -
Unclassifiable 2,720 125 19 (15) 27 (22) 22 (18) 14 (11) 8 (6) 24 (19) 0 (0) 0 (0) 21 (17)
www.njrcentre.org.uk
173
Table 3.53 gives a breakdown of the number of Oxford Shoulder Score (OSS) after their surgery. We
primaries that were subsequently revised together with can now provide more detail on this 8%. This sub-set
the indications for the first revision procedure. Please contained slightly more males than those whose OSS
note, the indications for revision were not mutually had improved (35% versus 28%) and also tended to
exclusive and, for 120 of the 860 first revisions, more be younger (median age 71 (IQR 63-76) versus 73
than one reason was recorded. (IQR 68-78). We have tracked 274 of these patients in
this year’s report; 33 had been revised by the end of
Different proportions of reasons for revision exist for 2017 and 32 died without further revision. PROMs that
the different implant types. For hemiarthroplasty, there were worse than before primary surgery were seen
is a mixture of indications, with cuff insufficiency and across all implant types.
conversion to conventional total replacement being
the most common reasons. The latter presumably Further analysis of these 274 cases revealed ten of
for ongoing pain attributed to not replacing the the 107 hemiarthroplasties in this group had been
glenoid. For conventional total shoulder replacement, revised, 11 of the 54 total shoulder arthroplasties in
51% were revised for cuff insufficiency and 37% for this group had been revised, and five of 91 reverse
dislocation. For reverse total shoulder replacement, total shoulder arthroplasties in this group had been
33% were for instability and 18% for infection. 24% revised. For those with a PROMs score worse than
were for ‘other’ reasons. These reasons for revision do prior to their primary shoulder replacement, 42 were
suggest a substantially higher infection rate for reverse in cases where the implant could not be verified.
total shoulder replacement. As some implant types are easier to revise than
others, it follows that they may be more likely to
3.6.3 PROMs Oxford Shoulder Scores undergo revision. Other implants such as reverse total
(OSS) associated with primary shoulder replacement are more difficult to revise and
perhaps less likely to undergo revision. This further
shoulder replacement surgery
highlights the importance of PROMs collection in
Having completed a three year PROMs pilot in 2015, providing information on outcome following shoulder
we published the results in the NJR Annual Report replacement. Using PROMs or revision surgery as
2016. Besides reporting the feasibility of PROMs endpoints in isolation thus may give very different
collection, their additional value was demonstrated impressions of outcome. This further highlights the
in the identification of unrevised cases with importance of longer term PROMs follow-up and data
poor outcomes. capture to determine the trends over time.
In our 2016 report we reported on a cohort of 3,411 Last year the NJR began collecting three and five year
patients with complete Q1 (pre-surgery) and Q2 (six PROMs and in the future we will be able to publish
months post-surgery) PROMs questionnaires. A total some of these longer term PROMs results.
of 275 (8%) of the 3,331 elective patients had a worse
174 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Shoulders
3.6.4 Mortality after primary shoulder 30,697 implants, 1,793 of the recipients had died by
the end of December 2017.
replacement surgery
It is important to separate mortality rates following
For this analysis, the second procedure or side of the
acute trauma from mortality rates after elective surgery
23 pairs of bilateral operations performed on the same
due to the different populations and risks involved.
day (see Table 3.2) were excluded. Of the remaining
Table 3.54 KM estimates of cumulative mortality (95% CI) for acute trauma and elective cases. Blue italics signify
that fewer than 250 cases remained at risk at these time points.
Table 3.54 shows the overall cumulative percentage operative period, we would expect higher rates in older
probability of mortality shown separately for acute age groups, and also in men. In the subsequent table,
trauma and the elective cases and shows higher rates Table 3.55, the larger elective group has been sub-
in the acute trauma group. divided by gender and age; the number remains too
small for further breakdown in the acute trauma cases.
However, this shows all-cause mortality and in
extended follow-up beyond the immediate post-
www.njrcentre.org.uk 175
176
Table 3.55 KM estimates of cumulative mortality (95% CI) for elective cases by age and gender. Blue italics signify that fewer than 250 cases remained at risk at these
time points.
Males Females
Age at Time since primary Time since primary
primary
(years)* n 90 days 1 year 2 years 3 years 4 years 5 years n 90 days 1 year 2 years 3 years 4 years 5 years
0.6 1.7 2.8 4.2 6.4 0.1 0.6 1.3 1.6 4.3 6.1
<55 943 0 784
(0.3-1.4) (1.0-3.0) (1.7-4.5) (2.6-6.6) (4.1-9.9) (0.02-0.9) (0.2-1.5) (0.7-2.6) (0.8-3.0) (2.6-7.3) (3.5-10.5)
0.3 1.0 2.1 3.4 4.5 5.3 0.1 0.5 1.4 2.3 4.1 4.9
55-64 1,567 2,017
(0.1-0.8) (0.6-1.6) (1.4-3.1) (2.4-4.8) (3.2-6.2) (3.7-7.6) (0.03-0.4) (0.3-0.9) (0.9-2.1) (1.6-3.3) (3.0-5.6) (3.5-6.7)
www.njrcentre.org.uk
0.3 1.1 2.8 4.3 6.2 9.1 0.2 0.8 2.1 3.5 5.6 7.9
65-74 3,200 7,288
(0.1-0.5) (0.7-1.5) (2.2-3.5) (3.5-5.3) (5.1-7.6) (7.3-11.2) (0.1-0.3) (0.6-1.0) (1.7-2.5) (3.0-4.1) (4.8-6.4) (6.8-9.3)
© National Joint Registry 2018
0.7 3.1 6.4 11.5 17.8 25.0 0.3 1.6 4.2 8.1 12.7 16.6
75+ 2,642 9,522
(0.4-1.1) (2.5-3.9) (5.5-7.6) (10.0-13.2) (15.7-20.2) (21.8-28.6) (0.2-0.5) (1.4-1.9) (3.8-4.7) (7.4-8.8) (11.7-13.8) (16.2-19.2)
*Excludes 12 cases where the NHS number was not traced therefore the age could not be validated.
National Joint Registry | 15th Annual Report | Shoulders
www.njrcentre.org.uk 177
Part 3
3.7 Outcomes
after elbow
replacement
National Joint Registry | 15th Annual Report | Elbows
3.7.1 Overview of primary elbow A total of 2,872 primary replacements were available
for analysis for a total of 2,776 patients. Of these
replacement surgery patients, 96 had documented replacements on both
This section contains an overview of the primary left and right sides, and in one patient these were
elbow replacements with data linked revision both performed on the same day (bilateral), see Table
and mortality data entered into the registry since 3.2 in section 3.2.
recording began (1 April 2012) up to the end of 31
The majority of replacements were performed on
December 2017, and documents the first revision
women (70.5%) and the median age at the primary
and mortality for these primaries. Primary elbow
operation was 68 years (IQR 57-76), with an overall
replacement in this section refers to total prosthetic
range of 14 to 98 years.
replacements, humeral hemiarthroplasty, lateral
resurfacing and radial head replacement.
Table 3.56 Number of primary elbow replacements by year and percentages of each type of procedure.
Year of primary
2012* 2013 2014 2015 2016 2017
Table 3.56 shows that the number of primary elbow been included in the minimum dataset (MDS) on
replacements entered into the NJR has continued the primary elbow (E1) form and this has resulted in
to increase year on year. This is likely to reflect an implants being entered with an incorrect procedure.
increase in data capture as well as an increase in the From 2018, humeral hemiarthroplasty will be included
volume of procedures. in the MDS (v7). There may be further anomalies and
a full independent review will be needed.
This table also gives a breakdown by the stated type
of replacement. Five were reclassified on the basis A number of primaries entered as total replacements
of obvious component anomalies (i.e. ‘radial head only had humeral components entered (n=106).
replacements’ with humeral components entered Given a large proportion of these (n=92) were
(n=2) and ‘lateral resurfacings’ with either an ulnar branded Latitude Humerals, which can also be
component or a ‘linked’ humeral component entered used in humeral hemiarthroplasties, we classified
(n=3). Humeral hemiarthroplasty has not previously them, together with one further IBP brand, to
www.njrcentre.org.uk 179
humeral hemiarthroplasty if we found the associated suggested a definite category (eight were retained as
components included an anatomical spool (n=86). total prosthetic replacements; three were changed to
Six implants entered as total elbow replacements had hemiarthroplasties and the remainder left as Uncertain).
ulnar parts entered as well as anatomical spools and
were classified as ‘Uncertain’. Table 3.57 details the type of primary operation
in each year. A total of 934 (32.5%) elbow
Finally, 49 of the total elbow replacements had only replacements were carried out for acute trauma.
accessories entered. These were considered as These have been separated from the remaining 1,938
‘Uncertain’ unless the accessories/accessory brands elective cases in the rest of this section.
Table 3.57 Types of primary elbow procedures used in acute trauma and elective cases by year.
2012 39 16 0 8 3 66*
2013 76 29 0 10 4 119
Acute
2014 63 51 0 7 1 122
trauma
2015 109 77 0 11 4 201
2016 87 95 0 19 1 202
2017 83 114 0 25 2 224
All years 1,772 106 19 11 30 1,938
2012 171 8 9 1 7 196
2013 306 7 5 3 10 331
Elective 2014 319 6 2 2 1 330
2015 315 18 0 2 6 341
2016 314 29 1 3 5 352
2017 347 38 2 0 1 388
*Includes one primary operation with date entered as 2010.
Table 3.58 Reasons for main types of primary elbow replacements, by year of primary.
(a) Total prosthetic replacements
Elective
Acute
trauma Number (%)* for each reason (amongst elective cases only):
Other
© National Joint Registry 2018
*Percentages based on the total numbers of elective cases; note the listed reasons are not mutually exclusive, more than one reason could have been stated.
180 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Elbows
Elective
Acute
trauma Number* for each reason (amongst elective cases only):
Elective
Acute
trauma Number* for each reason (amongst elective cases only):
Tables 3.58 (a) to (c) detail the indications for stated. In 26 of the 934 acute trauma cases a second
the primary operation shown separately for reason for surgery was given. In 95 of 1,938 elective
total replacements, lateral resurfacing/humeral cases (4.9%) more than one indication was given.
hemiarthroplasty and radial head replacement.
www.njrcentre.org.uk 181
182
Table 3.59 Number of units and consultant surgeons providing primary elbow replacements during each year from 2015 to 2017 (includes total prosthetic replacement,
lateral resurfacing, humeral hemiarthroplasty and radial head replacements).
Year of primary
2015 2016 2017
Total number of primary replacements during each year 542 554 612
Number of units providing any primary replacement
162 164 161
types in the year
Mean number of any primary replacements per unit 3.3 3.4 3.8
Median (IQR) number of any primary replacements per
2 (1-5) 3 (1-4) 2 (1-5)
www.njrcentre.org.uk
All primary unit
replacements Number of units who entered:
(i) only acute trauma cases 30 26 21
(ii) only elective cases 74 75 78
Units (iii) both acute trauma and elective cases 58 63 62
Number of units providing primary total elbow
141 144 142
replacements in the year
Number of units providing only total elbow
107 102 94
Primary total elbow replacements in the year (and no other types)
replacements only Mean number of primary total elbow replacements per
3 2.8 3
unit
Median (IQR) number of primary total elbow
2 (1-4) 2 (1-3) 2 (1-3)
replacements per unit
Number of consultants providing any primary
214 218 222
replacements in the year
© National Joint Registry 2018
Over the last three years (from 2015), 1,708 of all acute trauma) over the whole three-year period; the
types of primaries have been entered into the registry maximum number entered over this three-year period
(see Table 3.57). These procedures were performed by any one unit was 48, with five units entering 30 or
by 339 consultants, working across 228 units. more. However, 121 units (58%) had entered fewer
than five elective cases over this same period.
Table 3.59 shows a breakdown of unit and
consultant caseload for each year for all primary In 2017, taking elective and trauma cases together, the
elbow replacements performed, together with the numbers of units and surgeons doing only one primary
number of units and consultants entering only acute total prosthetic replacement in that year were 56 and 78
trauma cases, only elective cases, and both types respectively. The numbers of units and surgeons doing
within that year. Also shown is the number of units fewer than five total prosthetic replacements in that year
and consultants who did any primary total elbow (2017) were 113 and 155 respectively.
replacements and the number of these who only did
total elbow replacements. Table 3.60 lists the brands used in total prosthetic
replacements, with sub-division by acute trauma and
A total of 207 units had entered at least one primary elective cases.
total prosthetic replacement (either elective or
K Elbow 4 0 4
IBP 11 1 10
NES 2 0 2
Linked brands:
Latitude (+ ulnar cap) 160 22 138
Discovery 648 122 526
Coonrad Morray 1,266 308 958
GSB III 41 4 37
Mutars 2 0 2
Uncertain 4 0 4
Total 2,229 457 1,772
www.njrcentre.org.uk 183
Table 3.61 Radial head brands used in radial head replacements.
RHS 19 9 10
rHead (Recon) 6 3 3
Mono brands:
Corin Radial Head 25 21 4
Evolve Proline 84 67 17
ExploR 44 35 9
Anatomic Radial Head 236 195 41
MoPyC 8 6 2
Ascension 41 26 15
Liverpool 4 3 1
Uni Radial (Standard) 6 4 2
Uncertain: 13 12 1
Total 488 382 106
Table 3.61 lists the radial head brands used for radial 3.7.2 Revisions after primary elbow
head replacement, with sub-division by acute trauma
and elective procedures.
replacement surgery
A total of 80 elbow primaries (11 acute trauma cases
The lateral resurfacings and humeral
and 69 elective) had been revised up to the end of
hemiarthroplasties are not tabulated. Of the 19
2017, including three excision arthroplasties. Revision
lateral resurfacings, 15 were using the LRE brand,
procedures in the registry have been entered by 184
two used the Uni elbow capitellum system and the
consultant surgeons working across 135 units.
remaining two are Uncertain. Of the 91 humeral
hemiarthroplasties, 90 were Latitude prostheses and During 2017 there were 132 revision procedures
the remaining one was an IBP. entered in to the NJR by 68 consultants working
across 52 units.
184 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Elbows
Table 3.62 KM estimates of cumulative revision (95% CI) by primary elbow procedures for acute trauma and
elective cases. Figures in blue italics signify time points at which fewer than 250 patients remain at risk.
Table 3.62 shows Kaplan-Meier estimates of the Amongst the 106 elective radial head replacements,
cumulative percentage probability of revision up to two were revised by the end of 2017. The total
four years after the primary operation, together with prosthetic replacements performed for acute trauma
95% Confidence Intervals for all cases and for acute cases had similar cumulative revision rates up to 2.5
trauma and elective cases separately. Generally the years to those for elective cases, as further illustrated
group sizes were too small for meaningful sub-division in Figure 3.27, after which the rates are unreliable. At
by type of procedure. Amongst the 382 radial head the current time, there are too few cases for further
replacements carried out for acute trauma, only two sub-division into age/gender sub-groups, but we hope
revisions had been reported up to the end of 2017. to do this in future reports as the numbers increase.
www.njrcentre.org.uk 185
Figure 3.25 KM estimates of cumulative revision after primary total prosthetic
elbow replacement by acute trauma and elective cases
Figure 3.27 KM estimates of cumulative revision after primary total prosthetic elbow replacement
by acute trauma and elective cases.
6
Cumulative revision (%)
© National Joint Registry 2018
0
0 1 2 3 4 5
Years since primary
Numbers at risk
Acute trauma 457 349 252 143 94 25
Elective 1,772 1,384 1,036 699 391 131
Table 3.63 Indications for first data linked revision after any primary elbow replacement.
Acute trauma and elective cases are shown separately, for (i) total replacements, lateral resurfacings and humeral
hemiarthroplasties and (ii) radial head replacements.
186 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report | Elbows
Table 3.63 gives a breakdown of the indications for 3.7.3 Mortality after primary elbow
the first data linked revision procedure, the most
common reasons being for infection and for aseptic
replacement surgery
loosening. Please note, the indications for revision For this analysis, the second procedure of the pair
were not mutually exclusive; in eight of the 80 revisions of bilateral operations performed on the same day
more than one reason was stated. A few cases (n=8) (see Table 3.2) were excluded. Among the remaining
once revised had gone on to have more revision 2,871 implants, 232 of the recipients had died by the
procedures (other than planned two-stage revisions end of December 2017. Estimates of the cumulative
for infection). percentage probability of mortality in this cohort were
0.53 (95% CI 0.32-0.88) at 90 days and 2.56 (95% CI
2.01-3.25), 5.23 (95% CI 4.38-6.25), 9.09 (95% CI 7.84-
10.53) and 12.96 (95% CI 11.27-14.87) respectively at
1, 2, 3 and 4 years after the primary operation.
Table 3.64 KM estimates of cumulative mortality (95% CIs) by time from primary elbow replacement, for acute
trauma and elective cases. Figures in blue italics denote time points where fewer than 250 cases remained at risk,
hence the 95% CIs are not reliable.
Table 3.64 shows the overall cumulative percentage However this is all-cause mortality and in extended
probability of mortality shown separately for acute follow-up beyond the immediate post-operative
trauma and the elective cases, and shows higher rates period, we would expect higher rates in older age
in the acute trauma group. Radial head replacements groups; and also in men. As the size of the dataset
are shown separately from the remainder for the increases, we will be able to present mortality for
acute cases but not for the 106 elective radial head elective cases in age/gender sub-groups.
replacements, as there were no subsequent deaths.
www.njrcentre.org.uk 187
3.7.4 Conclusions The distribution of indications for total elbow
replacement have been consistent over the five
The annual number of primary elbow replacement years of data entry with inflammatory replacement
procedures entered into the registry continues to accounting for half of cases. In 2017 there were 430
increase. It is not known how accurate or complete primary total elbow replacements performed in 142
the dataset is as an independent audit of elbow units by 177 consultants. Three implant types account
replacement data has yet to be undertaken. for 97% of total elbow replacements performed.
An attempt has been made to separate out different The cumulative probability of revision of elbow
procedure types based on the description of the replacement at two years was 2.19 (95% CI 1.65-
procedure entered and the types of prosthesis used. 2.90) with equivalent rates of revision for trauma and
This has identified a number of anomalies and further elective total elbow replacement. The main indications
work is required to address these. for revision were infection and aseptic loosening.
Elbow hemiarthroplasty has been reported for The one-year mortality rate following elbow
the first time this year. This was not listed in the replacement (excluding radial head replacement) is
minimum dataset until 2018 and there may be significantly higher in the trauma population than in
cases not submitted to the NJR – despite this there those having elective surgery, however this may well
appears to be an annual increase in the number of represent a difference in the demographics of these
hemiarthroplasty procedures performed. In the future two groups.
it should be possible to compare the revision rates
for this relatively new procedure compared to total
elbow replacement. Most humeral hemiarthroplasty
procedures are performed for acute trauma and
trauma sequelae as expected.
188 www.njrcentre.org.uk
Part 3
3.8 In-depth
studies
The NJR Research Committee encourages use of the Study cohorts
NJR dataset to maximise its value to patients and the
Type of hip arthroplasty was defined as the exposure
wider health community; providing datasets to both
of interest. The MoM group included both resurfacing
internal NJR studies and external researchers.
and stemmed devices while patients with both a MoM
Here we present summaries of three in-depth studies and a Non-MoM hip were excluded.
carried out using NJR data. These examine the risk
Outcome measures
of cardiac failure following total hip arthroplasty, the
probability of revision for periprosthetic fracture, and the Heart failure was defined using ICD-10 codes specified
use of trabecular metal-coated acetabular components. by the National Heart Failure Audit (NHFA). Primary
outcome: Time to incident heart failure event following
3.8.1 Heart failure after metal-on- THR. Secondary outcome: All-cause mortality.
metal hip replacement Statistical analysis
In response to a number of case series suggesting We constructed Cox proportional hazards models
an increased incidence of heart failure after metal-on- to estimate Hazard Ratios (HR) for heart failure and
metal hip replacement, we examined NJR records up all-cause mortality for each cohort. We analysed
to the end of December 2014. (a) all eligible hip replacements and (b) propensity
matched cohorts. Propensity matching was one-to-
We analysed time to an admission to hospital for one for MoM with Non-MoM recipients. The cohort
cardiac failure or death following hip replacement. The was stratified by gender before propensity matching
aim was to compare these outcomes for metal-on- and matched by age, prior history of diabetes,
metal (MoM) versus other types of hip replacement heart disease and hypertension, Charlson index,
(Non-MoM), after controlling for patient factors. and ASA class. A caliper of 0.2 standard deviations
was used for propensity score. Sensitivity analyses
Methods
were performed for (i) modular MoM hips, (ii) hip
Data sources and linkage resurfacings, (iii) ASR XL alone versus Non-MOM hips
We performed data linkage between the National Joint and (iv) patients with a prior history of heart failure.
Registry (NJR) for England, Wales, Northern Ireland
Results
and the Isle of Man, Hospital Episodes Statistics
(HES), and Office for National Statistics (ONS) records A total of 535,776 (97.3%) patients were included in
on deaths. Patient-level linkage was performed using the main analysis after 14,813 (2.7%) patients were
NHS number, date of birth, gender and postcode for excluded due to prior heart failure. In total, 53,529
patients that had provided NJR consent. (10.0%) patients received a MoM hip implant. MoM
patients were younger (mean age 58.6 years versus
Patients 69.2), a greater proportion were male (61.1% versus
A total of 550,589 patients with an elective primary 38.1%), and had fewer co-morbidities. By design,
hip replacement recorded on the NJR between 1 propensity matching reduced these differences (see
January 2003 and 31 December 2014 were linkable Table 3.65).
to HES after data cleaning. HES data was available
a) Analysis of all eligible hip replacements
from 1997, providing a minimum period of six years
to establish a diagnosis of pre-existing heart failure at There were 1,431 heart failure events in the MoM
the time of total hip replacement (THR). Patients with cohort and 21,245 in the Non-MoM cohort. The crude
pre-existing heart failure or a new diagnosis in the event rates were 3.8 and 9.7 per 1,000 person-years’
six months after THR were excluded from the main exposure, respectively, with a crude relative rate of
analysis, but included in sensitivity analyses. 0.389 (95% CI 0.368–0.410). Following adjustment,
the hazard ratio (aHR) for MoM patients relative to
Non-MoM was 0.901 (95% CI 0.853–0.953). The
190 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Table 3.65 Baseline characteristics of patients with implanted metal-on-metal hip prostheses and non metal-on-
metal controls: all patients and those matched by propensity score.
www.njrcentre.org.uk 191
aHR for all-cause mortality was 0.892 (95% CI 0.862– Conclusion
0.924). In the sub-group analyses the aHR was below
unity for all analyses (Table 3.66), although this did This study found a lower incidence of heart failure
not achieve significance for all groups. In age-specific and mortality in patients with metal-on-metal hip
sub-group analysis there was a trend for an increasing replacements, compared to other hip types in the
hazard ratio with increasing age, ranging from 0.600 first seven years after surgery. Whilst there may be
(0.353–1.019) in the youngest cohort (≤44 years) to residual confounding by indication, these results
1.115 (0.908–1.370) in the oldest (≥85 years) group. should provide reassurance to clinicians and patients
alike regarding cardiac sequelae associated with these
b) Propensity matched analysis devices. We recommend epidemiological analysis at
In the propensity matched analysis there were 1,431 five yearly intervals to investigate for any latent effects.
heart failure events in the MoM cohort and 1,004 in
the Non-MoM cohort. Due to the longer follow-up,
Full paper details:
the crude event rates were lower in the MoM cohort The risk of cardiac failure following metal-on-metal
(3.8 events per 1,000 person years, versus 4.1) with hip arthroplasty
an unadjusted relative rate of 0.917 (95% CI 0.846–
0.994). In adjusted analyses, the aHR was 0.909 (95% S. A. Sabah, J. C. Moon, S. Jenkins-Jones, C. LI.
CI 0.838–0.987). For all-cause mortality, the aHR was Morgan, C. J. Currie, J. M. Wilkinson, M. Porter,
0.877 (95% CI 0.835–0.922). There were no sub- G. Captur, J. Henckel, N. Chaturvedi, P. Kay, J. A.
groups with a significantly increased aHR. Skinner, A. Hart, C. Manisty
192 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Table 3.66 Adjusted hazard ratio of heart failure for patients with implanted metal-on-metal hip prostheses and non
metal-on-metal controls: all patients by sub-group.
www.njrcentre.org.uk 193
3.8.2 Revision for Periprosthetic unchanged, revision burden in relation to PFF
specifically has gradually increased from 0.43 in 2004
Femoral Fracture in total hip to 1.07 in 2014. Incidence of PFF presented in terms
arthroplasty of implant-years at risk also showed an increase
between 2006 and 2014 (trend, P=0.022). Figure 3.28
Temporal Trends demonstrates the change in incidence of revision for
An analysis of NJR data demonstrates an increasing PFF compared with aseptic loosening between 2006
number of revisions for periprosthetic femoral fractures and 2014.
(PFF) being performed since 2003. Although the
overall all-indication revision burden was largely
Figure 3.28 Temporal changes in incidence of revision for PFF and aseptic loosening.
12
© National Joint Registry 2018
10
2006 2008 2010 2012 2014 2006 2008 2010 2012 2014
Year
194 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Figure 3.29 (a) to (d) Probability of revision for PFF using Flexible Parametric Competing Risks models in gender
and age-based groups.
0.5 0.5
0 0
0 2.5 5 7.5 10 12.5 0 2.5 5 7.5 10 12.5
Years since primary Years since primary
1.0 1.0
0.5 0.5
0 0
0 2.5 5 7.5 10 12.5 0 2.5 5 7.5 10 12.5
Years since primary Years since primary
www.njrcentre.org.uk 195
Males <70 Cemented Stems Males <70 Cementless Stems
0.5 0.5
0 0
0 2.5 5 7.5 10 12.5 0 2.5 5 7.5 10 12.5
Years since primary Years since primary
1.0 1.0
0.5 0.5
0 0
0 2.5 5 7.5 10 12.5 0 2.5 5 7.5 10 12.5
Years since primary Years since primary
196 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
A comparison of cementless collared and collarless Using a flexible parametric model accounting for
stems demonstrated an unadjusted hazard ratio the competing risks of death and revision for other
for PFF revision comparing collared to collarless indications, at ten years, the probability of PFF revision
cementless stems was 0.51 (95% CI 0.41-0.63). After for the most commonly used stem brands in England
adjustment for significant covariates (age, gender, and Wales showed variation particularly in older age
patient ASA grade), the HR was 0.45 (95% CI 0.36- groups (Table 3.67).
0.55), suggesting a lower risk of PFF revision for
collared cementless stems.
Table 3.67 Cumulative probability of revision at ten years (expressed as percentages) for periprosthetic femoral
www.njrcentre.org.uk 197
95% CI 0.43-0.76; P<0.001) (see Figure 3.30). The This work demonstrated that in patients undergoing
5-year cumulative implant survival rate free from primary THA, TM coated implants had a reduced
aseptic acetabular loosening was 99.9% (95% CI risk of both aseptic and septic revision compared
99.8%-99.9%) in the TM group compared with with non-TM implants. However these differences in
99.8% (95% CI 99.6%-99.9%) in the non-TM group revision risk between the groups were small, and may
(SHR 0.35, 95% CI 0.14-0.90; P=0.029). The 5-year only be clinically significant if the TM designs were
cumulative implant survival rate free from infection after implanted in the most complex cases.
primary THA was 99.5% (95% CI 99.3%-99.7%) in
the TM group compared with 99.1% (95% CI 98.8%-
99.3%) in the non-TM group (SHR 0.51, 95% CI 0.34-
0.76; P=0.001).
Figure 3.30 Cumulative acetabular component survival rate following primary THA in TM and non-
TM implants.
100
99
© National Joint Registry 2018
98
97
96
95
0 1 2 3 4 5
Follow−up time (years)
Trabecular metal in revision total hip implants and non-TM implants. The 6-year cumulative
arthroplasty all-cause acetabular survival rate was 97.2% (95%
CI 96.2%-97.9%) in revision THAs with TM coatings
The matched cohort included 3,862 revision THAs compared with 96.9% (95% CI 95.8%-97.6%) in
(mean age 71.7 years and 59% female), with 1,931 non-TM coatings (See Figure 3.31) (P=0.636). The
hips in both the TM (1,707 TM Modular and 224 6-year cumulative implant survival rate free from
Continuum) and non-TM groups (1,717 Trilogy and aseptic loosening was 98.8% (95% CI 98.1%-99.3%)
214 Trilogy IT). in the TM group and 99.1% (95% CI 98.5%-99.5%) in
the non-TM group (P=0.410). The 6-year cumulative
The re-revisions rates for all outcomes of interest
implant survival rate free from infection after revision
were comparable between revision THAs with TM
198 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Figure 3.31 Cumulative acetabular component survival rate following revision THA in TM and non-
TM implants.
100
98
96
94
92
90
0 1 2 3 4 5 6
Follow−up time (years)
THA was 98.7% (95% CI 98.0%-99.2%) in the TM Trabecular metal acetabular components reduce the
group compared with 98.2% (95% CI 97.4%-98.8%) risk of revision following primary total hip arthroplasty:
in the non-TM group (P=0.165). A propensity score matched study from the National
Joint Registry for England and Wales.
Overall, 247 revision THAs were initially performed
for infection (116 TM and 131 non-TM). There was Gulraj S. Matharu, BSc (Hons), MRCS, MRes, DPhil,
no difference in the risk of re-revision for all-causes Andrew Judge, BSc, MSc, PhD, David W. Murray, MD,
(P=0.225), aseptic acetabular loosening (P=0.608), and FRCS (Orth) and, Hemant G. Pandit, DPhil, FRCS (Orth)
infection (P=0.706) between TM and non-TM coatings.
The Journal of Arthroplasty 2018;33(2):447-452. DOI:
This work demonstrated that in patients undergoing https://doi.org/10.1016/j.arth.2017.08.036
revision THA, both TM and non-TM coated acetabular
components were associated with comparable and Trabecular metal versus non-trabecular metal acetabular
relatively low re-revision rates. Extended follow-up of components and the risk of re-revision following revision
large revision THA cohorts will establish whether TM total hip arthroplasty. A propensity score-matched study
components have any clinical benefit over non-TM from the National Joint Registry for England and Wales.
designs when used in patients with similar acetabular
Matharu GS, Judge A, Murray DW, Pandit HG.
bone stock.
The Journal of Bone & Joint Surgery America
Full paper details: 2018;100:1132-40. DOI: http://dx.doi.org/10.2106/
This study was supported by a grant from Zimmer JBJS.17.00718
Biomet, the manufacturer of the implants investigated
in this study. The funder was not involved in the
performance of the study.
www.njrcentre.org.uk 199
Part 4
Implant and unit-
level activity and
outcomes
National Joint Registry | 15th Annual Report
Cup name Numbers implanted Latest PTIR Notified as outlier Last implanted
© National Joint Registry 2018
www.njrcentre.org.uk 201
Table 3 Level 1 outlier stem/cup combinations.
Numbers
Combination implanted Latest PTIR Notified as outlier Last implanted
ASR Resurfacing Head/ASR Resurfacing Cup 2,914 2.89 2010 July 2010
Metafix Stem/Cormet 2000 Resurfacing Cup 173 2.76 2010 February 2011
CPT/Adept Resurfacing Cup 268 3.46 2011 May 2010
Corail/ASR Resurfacing Cup 2,729 5.44 2011 June 2010
CPT/BHR Resurfacing Cup 116 2.5 2011 September 2010
Accolade/Mitch TRH Cup 274 2.64 2011 January 2011
Summit Cementless Stem/ASR Resurfacing Cup 128 4.79 2012 August 2009
CPT/Durom Resurfacing Cup 184 2.48 2012 September 2009
© National Joint Registry 2018
S-Rom Cementless Stem/ASR Resurfacing Cup 147 4.25 2012 February 2010
CPCS/BHR Resurfacing Cup 255 1.47 2012 May 2010
Anthology/BHR Resurfacing Cup 510 3.1 2012 August 2011
SL-Plus Cementless Stem/Cormet 2000 Resurfacing Cup 627 2.19 2013 April 2010
Profemur L Modular/Conserve Plus Resurfacing Cup 159 2.68 2013 June 2010
Bimetric Cementless Stem/M2A 38 1,302 1.82 2014 June 2011
Corin Proxima/Cormet 2000 Resurfacing Cup 102 2.23 2015 September 2009
CPT/Novae Stick 400 1.31 2015 February 2018
Synergy Cementless Stem/BHR Resurfacing Cup 1,584 1.88 2016 May 2011
Adept Cementless Stem/Adept Resurfacing Cup 200 2.02 2017 November 2010
Taperloc Cementless Stem/Apollo 147 2.65 2017 February 2018
Exeter V40/Trabecular Metal Revision Shell 172 1.88 2017 December 2017
CLS Cementless Stem/Adept Resurfacing Cup 218 2.4 2017 March 2011
CPT/ADES Cemented 397 1.41 2018 February 2018
Spectron/Opera 216 1.04 2018 February 2014
202 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
www.njrcentre.org.uk 203
Independent hospitals However, revisions taken only from the last five years
of the registry showed only ten hospitals reporting
• 81% achieved a proportion of patients with a linkable
higher than expected rates for knees, and seven
NHS number greater than 95%
for hips.
• 14% achieved a proportion of 80% to 95%
• 5% recorded a proportion of linkable records of less The 90-day mortality for hip and knee replacement
than 80% was calculated using the last five years of data for all
hospitals by plotting standardised mortality ratios for
Note: Independent hospitals might be expected to each hospital against the expected number of deaths.
have lower linkability rates than NHS hospitals, as a No hospitals had higher than expected mortality rates
proportion of their patients may come from abroad for either hip or knee replacement.
and not have an NHS number. Linkability figures are
not currently available for Northern Ireland. Note: The case mix adjustment for mortality includes
age, gender and ASA grade. Trauma cases have
4.3 Outlier units for been excluded from both the hip and knee mortality
analyses together with hips implanted for failed
revision rates for the reason began). Also, where both left and right side
joints were implanted on the same day, only one side
204 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Outliers for Hip mortality rates since 20131 Outliers for Hip revision rates, all linked primaries from
None identified 20031
Royal Cornwall Hospital (Treliske)
Outliers for Knee mortality rates since 20131 Salisbury District Hospital
None identified Shepton Mallet Treatment Centre (Somerset)
Spire Alexandra Hospital (Kent)
Outliers for Hip revision rates, all linked primaries from Spire Cardiff Hospital (Glamorgan)
20031 Spire Dunedin Hospital (Berkshire)
Ashtead Hospital (Surrey)
Spire Tunbridge Wells Hospital (Kent)
Basingstoke and North Hampshire Hospital
St Albans City Hospital
BMI Gisburne Park Hospital (Lancashire)
St Michael's Hospital
BMI Sarum Road Hospital (Hampshire)
Sussex Orthopaedic NHS Treatment Centre
BMI The Somerfield Hospital (Kent)
The Berkshire Independent Hospital (Berkshire)
Cheltenham General Hospital
The Royal London Hospital
Clifton Park Hospital (North Yorkshire)
University Hospital (Coventry)
Conquest Hospital
University Hospital of Hartlepool
Homerton University Hospital
University Hospital of North Tees
KIMS Hospital (Kent)
Wansbeck Hospital
Llandough Hospital
Watford General Hospital
Maidstone District General Hospital
York Hospital
Medway Maritime Hospital
Musgrove Park Hospital Note: 1 Date range 1 April 2003 to 1 March 2018 inclusive. 2 Date range 1
March 2013 to 1 March 2018 inclusive.
Nevill Hall Hospital
New Hall Hospital (Wiltshire)
North Tyneside General Hospital
Northampton General Hospital (Acute)
Nuffield Health Brighton Hospital (East Sussex)
Nuffield Health Cheltenham Hospital (Gloucestershire)
Nuffield Health Haywards Heath Hospital (West Sussex)
Nuffield Health Tees Hospital (County Durham)
Nuffield Health Wessex Hospital (Hampshire)
Nuffield Health York Hospital (North Yorkshire)
Pilgrim Hospital
Prince Charles Hospital
Rotherham District General Hospital
www.njrcentre.org.uk 205
Outliers for Hip revision rates, all linked primaries from Outliers for Knee revision rates, all linked primaries
20132 from 20132
Homerton University Hospital BMI The London Independent Hospital (Greater London)
KIMS Hospital (Kent) BMI The Meriden Hospital (West Midlands)
Salisbury District Hospital Broadgreen Hospital
Southampton General Hospital Ealing Hospital
St Richard's Hospital Guy's Hospital
Wansbeck Hospital Heatherwood Hospital
Weston General Hospital King Edward VII Hospital Sister Agnes (Greater London)
Leighton Hospital
Outliers for Knee revision rates, all linked primaries Nevill Hall Hospital
from 20031 St Richard's Hospital
Bradford Royal Infirmary
BMI Bishops Wood Hospital (Middlesex) Note: 1 Date range 1 April 2003 to 1 March 2018 inclusive. 2 Date range 1
March 2013 to 1 March 2018 inclusive.
BMI Goring Hall Hospital (West Sussex)
BMI The London Independent Hospital (Greater London)
BMI The Meriden Hospital (West Midlands)
Broadgreen Hospital
Castle Hill Hospital
Charing Cross Hospital
Conquest Hospital
County Hospital Louth
Good Hope Hospital
Guy's Hospital
Hinchingbrooke Hospital
Homerton University Hospital
Horton NHS Treatment Centre (Oxfordshire)
Hospital of St Cross
King Edward VII Hospital Sister Agnes (Greater London)
Kings Mill Hospital
Llandough Hospital
New Hall Hospital (Wiltshire)
Southampton General Hospital
Southmead Hospital
Spire Alexandra Hospital (Kent)
Spire Clare Park Hospital (Surrey)
Spire Southampton Hospital (Hampshire)
St Albans City Hospital
St Richard's Hospital
University College Hospital
Withybush General Hospital
206 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
4.4 Better than expected Better than expected for Hip revision rates, all linked
primaries from 20132
www.njrcentre.org.uk 207
Summary of key facts about joint replacement during the 2017
Hips
NJR Patien
Consent
t
105,306
replacement
60%
average ages:
NJR Patien
t
procedures
Consent
Knees
NJR Patien
Consent
t
112,836
replacement
56%
average ages:
procedures
NJR Patien
t
Consent
3.8%
recorded on the NJR
since April 2003 (108,713 in 2016) 69.2 69.4
Shoulders
NJR Patien
t
7,525
replacement
70%
Consent
average ages:
procedures
8%
recorded on the NJR
since April 2012 (6,967 in 2016) 69.3 74.1
208 www.njrcentre.org.uk
/nationaljointregistry @jointregistry
calendar year
Elbows
average
NJR PatBMI
Consen
ient
t
813
28.8
NJR Patien
t
90% Consent
replacement
osteoarthritis
procedures
= recorded on the NJR 12.6%
‘overweight’ since April 2012
Diagnosis (722 in 2016)
71% 27%
inflammatory
arthropathy
average ages:
average BMI 18%
osteoarthritis
98%
osteoarthritis 30.9 61.6 66.4 Diagnosis
= Ankles
Diagnosis
‘obese’
NJR Patient
Consent
886
replacement
procedures
5.6%
54% recorded on the NJR
since April 2010
(839 in 2016)
osteoarthritis
visit www.njrreports.org.uk.209
For more data on clinical activity during the 2017 calendar yearwww.njrcentre.org.uk
Glossary
National Joint Registry | 15th 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.
ALVAL Aseptic Lymphocyte-dominated Vasculitis-Associated Lesion. This term is used in the Annual Report
to describe the generality of adverse responses to metal debris, but in its strict sense refers to the
delayed type-IV hypersensitivity response.
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.
Beyond Compliance A system of post market surveillance initiated in 2013. Under this system a scrutiny committee closely
monitors the usage and performance of implants which are new to the market in order that any
problems may be quickly identified and that the necessary corrective actions are undertaken in order
to protect patient safety.
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 Zenith 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, Wales, Northern
Ireland and the Isle of Man 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 is based on the number of procedures submitted to HES and PEDW.
www.njrcentre.org.uk 211
Compliance Confidence Interval (CI) A ‘Confidence Interval’ (CI) is calculated to accompany anything being estimated from just a random
sample of cases, for example the cumulative probability of revision; a CI tells us something about the
range of values that the ‘true’ (population) value can take. Whilst calculated Confidence Intervals by
their very nature will vary from sample to sample, calculation of a ‘95% Confidence Interval’ (95% CI)
means that 95% of all such calculated intervals should actually contain the ‘true’ value.
Confounding Can occur when an attempt to quantify how a particular variable of interest affects outcome is
hampered by another variable(s) being related to both the variable of interest and the outcome. For
example a comparison of the revision rates between two distinct types of implant may be hampered
by the fact that one implant has been used on an older group of patients than the other; age here
is a ‘confounder’ for the relationship between implant type and outcome because revision rate also
depends on age. Statistical methods may help to ‘adjust’ for such confounding variables.
Cox ‘proportional hazards’ model A type of multivariable regression model used in survival analysis to look at the simultaneous effects of
a number of variables (‘predictors’) on outcome (first revision or death). The effect of each variable is
adjusted for the effects of all the other ‘predictor’ variables in the model so the Cox model can be
used to adjust for ‘confounders’ (see above). Some regression models used in survival modelling make
assumptions about the way the hazard rate changes with time (see ‘hazard rate’). The Cox model
doesn’t make any assumptions about how the hazard rate changes however it does assume that
the predictor variables affect the hazard rates in a ‘proportional’ way; the latter requiring some careful
model checking when this method is used.
Cross-linked polyethylene See modified polyethylene.
Cumulative incidence function (CIF) A different way of estimating failure compared to Kaplan-Meier, see Kaplan-Meier. Also known as
observed or crude failure, as the estimate reflects what is seen in practice.
Cup See Acetabular component.
Data collection periods for annual The NJR Annual Report Part One reports on data collected between 1 April 2017 and 31 March
report analysis 2018 – the 2017/18 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 2017
inclusive – the 2017 calendar year. The NJR Annual Report Part Three reports on hip, knee, ankle and
shoulder and elbow joint replacement revision rates for procedures that took place between 1 April
2003 and 31 December 2017.
DAIR Debridement And Implant Retention. In cases of infection, the surgeon may debride (surgically clean)
the surgical site and retain the joint replacement implants.
DAIR with Modular Exchange Debridement And Implant Retention with Modular Exchange. In cases of infection where the implants
are modular, the surgeon may debride (surgically clean) the surgical site, exchange the modular
components (e.g. head, acetabular liner) and retain the non-modular joint replacement implants.
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.
212 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
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.
Funnel plot A graphical device to compare unit or surgeon performance. Measures of performance (e.g. a ratio
of number of observed events to the expected number based on case-mix) are plotted against an
interpretable measure of precision. Control limits are shown to indicate acceptable performance. Points
outside of the control limits suggest ‘special cause’ as opposed to ‘common cause’ variation (see for
example D Spiegelhalter, Stats in Medicine, 2005).
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.
Hazard rate Rate at which ‘failures’ occur at a given point in time after the operation conditional on ‘survival’ up
to that point. In the case of first revision, for example, this is the rate at which new revisions occur in
those previously unrevised.
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 NHS England.
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.
www.njrcentre.org.uk 213
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 Used to estimate the cumulative probability of ‘failure’ at various times from the primary operation, also
known as Net Failure. ‘Failure’ may be either a first revision or a death, depending on the context. The
method properly takes into account ‘censored’ data. Censorings arise from incomplete follow-up; for
revision, for example, a patient may have died or reached the end of the analysis period (end of 2017)
without having been revised.
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.
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.
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.
MDSv1 Minimum dataset version one, used to collect data from 1 April 2003. MDS version one closed to new
data entry on 1 April 2005.
MDSv2 Minimum dataset version two, introduced on 1 April 2004. MDS version two replaced MDS version
one as the official dataset on 1 June 2004.
MDSv3 Minimum dataset version three, introduced on 1 November 2007 replacing MDSv2 as the new
official dataset.
MDSv4 Minimum dataset version four, introduced on 1 April 2010 replacing MDSv3 as the new official dataset.
This dataset has the same hip and knee MDSv3 dataset but includes the data collection for total ankle
replacement procedures.
MDSv5 Minimum dataset version five, introduced on 1 April 2012 replacing MDSv4 as the new official dataset.
This dataset has the same hip, knee and ankle MDSv4 dataset but includes the data collection for total
elbow and total shoulder replacement procedures.
MDSv6 Minimum dataset version six, introduced on 14 November 2014 replacing MDSv5 as the new official
dataset. This dataset includes the data collection for hip, knee ankle, elbow and 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 10cm). This may require the use of
special instruments.
214 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
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.
ODEP Orthopaedic Data Evaluation Panel of the NHS Supply Chain. www.odep.org.uk.
ODEP ratings ODEP ratings are the criteria for product categorisation of prostheses for primary total hip and knee
replacement against benchmarks. An ODEP rating consists of a number and a letter and a star. The
number represents the number of years for which the product’s performance has been evidenced.
The letter represents the strength of evidence (data) presented by the manufacturer. The star has
been added to the rating system following revised guidelines from NICE in February 2014, in which a
benchmark revision rate of less than 5% at 10 years was defined. The star is awarded where products
are evidenced to comply with this benchmark. A* represents evidence above A and B. Ratings without
a star signify compliance with the prior NICE guidance of a replacement rate of less than 10% at 10
years. The same benchmark has been adopted by ODEP for knees. All implants that are used without
a 10-year benchmark should be followed up closely. See www.odep.org.uk.
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 acceptable control limits. See also
‘Funnel plot’.
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.
Patellofemoral knee Procedure involving replacement of the trochlear and replacement resurfacing of the patella.
Patellofemoral prosthesis Two-piece knee prosthesis that provides a prosthetic (knee) articulation surface between the patella
and trochlear.
www.njrcentre.org.uk 215
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 total of the lengths of time a cohort of patients were ‘at risk’. In the calculation of PTIRs for
revision, for example, each individual patient’s time is measured from the date of the primary operation
to the date of first revision or, if there has been no revision, the date of patient’s death or the last
observation date. The individual time intervals are then added together.
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.
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 The total of the lengths of time a cohort of prostheses were ‘at risk’. In the calculation of PTIRs
for revision, for example, each individual prosthesis time is measured from the date of the primary
operation to the date of first revision or, if there has been no revision, the date of patient’s death or the
last observation date. The individual time intervals are then added together.
Prosthesis-time The total of the lengths of time a cohort of prostheses were ‘at risk’. In the calculation of PTIRs for revision,
for example, each individual prosthesis time is measured from the date of the primary operation to the date
of first revision or, if there has been no revision, the date of patient’s death or the last observation date. The
individual time intervals are then added together.
PROMs Patient Reported Outcome Measures.
PTIR PTIR Prosthesis-Time Incidence Rate. The total number of events (e.g. first revisions) divided by the
total of the lengths of times the prosthesis was at risk (see ‘Prosthesis-time’).
Pulmonary Embolism A pulmonary embolism is a blockage in the pulmonary artery, which is the blood vessel that carries
blood from the heart to the lungs.
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
shoulder replacement prosthesis for whatever reason.
216 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Shoulder hemi-arthroplasty 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.
SOAL Lower Layer Super Output Areas. Geographical areas for the collection and publication of small area
statistics. These are designed to contain a minimum population of 1,000 and a mean population size
of 1,500. Please also see Office for National Statistics at www.ons.gov.uk.
Subtalar The joints between the talus and the calcaneum, also known as the talocalcaneal joints.
Surgical approach Method used by a surgeon to gain access to, and expose, the joint.
Survival (or failure) analysis Statistical methods to look at time to a defined failure ‘event’ (for example either first revision or death);
see Kaplan-Meier estimates and Cox ‘proportional hazards’ models. These methods can take into
account cases with incomplete follow-up (‘censored’ observations).
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),
patellofemoral joint (knee), talar component (ankle), reverse shoulder (shoulder) and radial head
replacement (elbow).
www.njrcentre.org.uk 217
U
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.
218 www.njrcentre.org.uk
National Joint Registry | 15th Annual Report
Data collection For research and analysis purposes, NJR data is annually
The National Joint Registry (NJR) produces this report linked to data from other healthcare systems using
using data collected, collated and provided by third patient identifiers, principally a patient’s NHS number.
parties. As a result of this the NJR takes no responsibility These other datasets include the Hospital Episodes
for the accuracy, currency, reliability and correctness Statistics (HES) service, the Patient Episode Database
of any data used or referred to in this report, nor for the Wales (PEDW), data from the NHS England Patient
accuracy, currency, reliability and correctness of links or Reported Outcomes Measures (PROMs) programme,
references to other information sources and disclaims all and data from the Office of National Statistics. The
warranties in relation to such data, links and references to purpose of linking to these data sets is to expand and
the maximum extent permitted by legislation. broaden the type of analyses that the NJR can undertake
without having to collect additional data. This linkage
The NJR shall have no liability (including but not limited has been approved by the Health Research Authority
to liability by reason of negligence) for any loss, damage, under Section 251 of the NHS Act 2006 on the basis
cost or expense incurred or arising by reason of any of improving patient safety and patient outcomes: the
person using or relying on the data within this report support provides the legal basis for undertaking the
and whether caused by reason of any error, omission or linkage of NJR data to the health data sets listed above.
misrepresentation in the report or otherwise. This report
is not to be taken as advice. Third parties using or relying Once the datasets have been linked, patient identifiable
on the data in this report do so at their own risk and will data are removed from the new dataset so that it is not
be responsible for making their own assessment and possible to identify any patient. This data is then made
should verify all relevant representations, statements and available to the NJR’s statistics and analysis team at
information with their own professional advisers. the University of Bristol whose processing of the data is
also subject to strict guidelines set out in an approved
Information governance and patient confidentiality System Level Security Policy. The work undertaken by
The NJR ensures that all patient data is processed and the University of Bristol is directed by the NJR’s Steering
handled in line with international and UK standards Committee and the NJR’s Editorial Board and the results
and within UK and European legislation: protecting and of the analyses are published in the NJR’s Annual Report
applying strict controls on the use of patient data is of the and in professional journals. All published work is based
highest importance. on aggregated data, rather than individual record level
data. This means that no patient could be identified.
NJR data is collected via a web-based data entry
application and stored and processed in Northgate
Public Services’ (NPS) data centre. In addition to being
accredited to ISO 27001 and ISO 9001, NPS is also
compliant with the NHS’ Information Governance Toolkit.
Contact:
Email: enquiries@njrcentre.org.uk
www.njrcentre.org.uk 219
Website: www.njrcentre.org.uk
www.njrcentre.org.uk
www.njrreports.org.uk
HIPS
KNEES
ANKLES
ELBOWS
SHOULDERS
PROMs
Every effort has been made at the time of
publication to ensure that the information
contained in this report is accurate. If
amendments or corrections are required
after publication, they will be published on
the NJR website at www.njrcentre.org.
uk and on the dedicated NJR Reports
website at www.njrreports.org.uk.
/nationaljointregistry
@jointregistry