National Joint Replacement Registry. Demographics and Outcome of Ankle Arthroplasty SUPPLEMENTARY

National Joint Replacement Registry Demographics and Outcome of Ankle Arthroplasty SUPPLEMENTARY Report 2017

AOAnjrr 2016 supplementary report AOAnjrr 2016 supplementary report AUSTRALIAN ORTHOPAEDIC ASSOCIATION NATIONAL JOINT REPLACEMENT REGISTRY Director: Professor Stephen Graves AOANJRR E: segraves@aoanjrr.org.au SAHMRI, North Terrace Manager: Ms Cindy Turner ADELAIDE SA 5000 E: cturner@aoanjrr.org.au T: +61 8 8128 4280 The AOANJRR is funded by the Australian Government Department of Health Photographer: John Gollings Suggested citation: Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). Hip, Knee & Shoulder Arthroplasty: 2017 Annual Report. Adelaide: AOA, 2017. www.aoa.org.au Australian Orthopaedic Association National Joint Replacement Registry 2017 ISSN 1445-3657 DEMOGRAPHICS AND OUTCOME OF ANKLE ARTHROPLASTY Data Period 1 September 1999 31 December 2015 Data Period 1 September 1999 31 December 2015 2 aoa.org.au aoa.org.au 3

Acknowledgements The Registry continues to receive support and invaluable assistance from the Commonwealth Government, state and territory health departments and orthopaedic companies. Table of Contents The Registry acknowledges the cooperation and support provided by those undertaking the surgery and completing the data forms, in particular, all orthopaedic surgeons, registrars and nursing staff. The Registry would also like to acknowledge the ongoing support of all hospitals, both public and private, that undertake arthroplasty surgery nationally. The support provided by each hospital through their nominated coordinator(s) is appreciated. A complete list of participating hospitals and coordinators is presented at the end of this report (Appendix 1). The Registry greatly appreciates the participation of all joint replacement patients throughout Australia. It is their contribution that allows ongoing improvements in arthroplasty outcomes to be achieved. DEPUTY DIRECTORS DIRECTOR AOANJRR STAFF Professor Richard de Steiger Professor Stephen Graves Ms Cindy Turner (Manager) Mr Peter Lewis Dr Sophia Rainbird Professor Ian Harris Ms Rychelle Brittain Ms Grace O Donohue Assistant Deputy Directors Mr James Stoney Mr Bill Donnelly Clinical Advisors Professor Richard Page (Upper Limb Joint Replacement) Mr Andrew Beischer (Ankle Replacement) SOUTH AUSTRALIAN HEALTH AND MEDICAL RESEARCH INSTITUTE STAFF Ms Liddy Griffith Dr Mandy Anderson Ms Michelle Lorimer Ms Alana Cuthbert Ms Andrea Peng Ms Alesha Hatton Ms Janey Barrow Mr Robert Armitage Many thanks to the Data Entry Team INTRODUCTION... 6 Background... 6 Purpose... 7 Aims... 7 Benefits... 7 Governance... 8 DATA QUALITY... 9 Data Collection... 9 Data Validation... 9 Outcome Assessment... 9 Report Review Prior to Publication... 10 ANKLE REPLACEMENT... 11 Categories of Ankle Replacement... 11 PRIMARY TOTAL ANKLE REPLACEMENT... 13 Demographics of Ankle Replacement... 13 Outcome for All Diagnoses... 15 Primary Diagnosis... 15 Reason for Revision... 15 Type of Revision... 15 Reoperation... 15 Prosthesis Types... 15 Outcome for Osteoarthritis... 19 Age and Gender... 19 Fixation... 19 LIST OF TABLES... 24 LIST OF FIGURES... 25 David Hale Stephen Graves Richard de Steiger Peter Lewis Ian Harris Roger Brighton John Radovanovic Alexander Burns Sean Williams Russell Furzer David Campbell Neil Bergman Peter Myers Richard Page Andrew Beischer Matthew Scott-Young James Stoney Bill Donnelly AOANJRR COMMITTEE Chairperson Director Deputy Director (Victoria) Deputy Director (South Australia) Deputy Director (New South Wales) New South Wales Queensland Australian Capital Territory Western Australia Tasmania President, Arthroplasty Society of Australia AOA Representative - AOANJRR Consultative Committee Australian Knee Society Shoulder & Elbow Society Representative Foot & Ankle Society Representative Spine Society of Australia Representative Assistant Deputy Director (observer status) Assistant Deputy Director (observer status) 4 aoa.org.au aoa.org.au 5

The 2017 Demographics and Outcome of Ankle Arthroplasty Report is based on the analysis of 2,235 ankle procedures recorded by the Registry with a procedure date up to and including 31 December 2016. The Ankle Arthroplasty Report is one of 12 supplementary reports to complete the AOANJRR Annual Report for 2017. The supplementary reports include: 1. Lay Summary Hip & Knee Replacement 2. Demographics of Hip, Knee & Shoulder Arthroplasty 3. Cement in Hip and Knee Arthroplasty 4. Mortality of Hip and Knee Arthroplasty 5. Revision of Hip and Knee Arthroplasty 6. Metal/Metal Bearing Surface in Total Conventional Hip Arthroplasty 7. Metal and Ceramic Bearing Surface in Total Conventional Hip Arthroplasty 8. The Outcome of Classes of Hip and Knee Prostheses No Longer Used 9. Demographics and Outcome of Elbow and Wrist Arthroplasty 10. Demographics and Outcome of Ankle Arthroplasty 11. Demographics of Spinal Disc Arthroplasty 12. Analysis of State and Territory Health Data All Arthroplasty 1993/1994 2015/2016 In addition to the 12 supplementary reports, Investigations of Prostheses with Higher than Anticipated Rates of Revision are published on: https://aoanjrr.sahmri.com/annual-reports- 2017. All hospitals, public and private, undertaking joint replacement submit their data to the Registry. Currently there are 310 participating hospitals. However, this may vary from time to time due to hospital closures, new hospitals, or changes to services within hospitals. Introduction BACKGROUND Joint replacement is a commonly performed major surgical procedure that has considerable success in alleviating pain and disability. The Australian Orthopaedic Association (AOA) recognised the need to establish a national joint replacement registry in 1993. At that time, the outcome of joint replacement in Australia was unknown. Patient demographics were not available and the types of prostheses and techniques used to implant them were unknown. The need to establish a Registry was, in part, based on the documented success of a number of arthroplasty registries in other countries. In particular, the Swedish arthroplasty registries. In Sweden, the ability to identify factors important in achieving successful outcomes has resulted in both improved standards and significant cost savings. In 1998, the Commonwealth Department of Health (DoH) funded the AOA to establish the Registry. The Department of Health continues to provide funding to maintain the Registry. In June 2009, Federal Parliament passed legislation to enable the government to cost recover this funding from the orthopaedic industry. This legislation was updated in 2015. Most hospitals began providing data on ankle arthroplasty procedures on or after the official commencement date of the 15 November 2007. The Registry had approval to collect these additional arthroplasty procedures from a number of hospitals prior to the commencement date. These data have also been included. The reporting period for ankle replacement in this report is 28 July 2006 to 31 December 2016. PURPOSE The purpose of the Registry is to define, improve and maintain the quality of care for individuals receiving joint replacement surgery. This is achieved by collecting a defined minimum data set that enables outcomes to be determined based on patient characteristics, prosthesis type and features, method of prosthesis fixation and surgical technique used. The principal outcome measure is time to first revision surgery. This is an unambiguous measure of the need for further intervention. Combined with a careful analysis of potential confounding factors, this can be used as an accurate measure of the success, or otherwise, of a procedure. The Registry also monitors mortality of patients, which is critical when determining the rate of revision. AIMS 1. Establish demographic data related to joint replacement surgery in Australia. 2. Provide accurate information on the use of different types of prostheses. 3. Determine regional variation in the practice of joint surgery. 4. Identify the demographic and diagnostic characteristics of patients that affect outcomes. 5. Analyse the effectiveness of different prostheses and treatment for specific diagnoses. 6. Evaluate the effectiveness of the large variety of prostheses currently on the market by analysing their survival rates. 7. Educate orthopaedic surgeons on the most effective prostheses and techniques to improve patient outcomes. 8. Provide surgeons with an auditing facility. 9. Provide information that can instigate tracking of patients if necessary. 10. Provide information for the comparison of the practice of joint replacement in Australia and other countries. BENEFITS Since its inception, the Registry has enhanced the outcome of joint replacement surgery in Australia. There are many factors known to influence the outcome of joint replacement surgery. Some of these include age, gender, diagnosis, ASA score and BMI of patients, as well as the type of prosthesis and surgical technique used. Another coexisting influence is the rapid rate of change in medical technology. There is continual development and use of new types of prostheses and surgical techniques, for many of which the outcome remains uncertain. Information obtained by the analysis of Registry data is used to benefit the community. The Registry releases this information through publicly available annual and supplementary reports, journal publications and ad hoc reports (256 in 2016). These ad hoc reports are specific analyses requested by surgeons, hospitals, academic institutions, government and government agencies as well as orthopaedic companies. The Registry provides surgeons with access to their individual data and downloadable reports through a secure online portal. Separate online facilities are available for orthopaedic companies to monitor their own prostheses, and for Australian and regulatory bodies in other countries to monitor prostheses used in Australia. The data obtained through the online facilities are updated daily and are over 90% complete within six weeks of the procedure date. 6 aoa.org.au aoa.org.au 7

GOVERNANCE The AOANJRR is an initiative of the AOA funded by the Commonwealth. In 2009, the Commonwealth established the AOANJRR Consultative Committee, which is administered and chaired by the Department of Health. The purpose is to provide advice on the overall strategic direction of the Registry. Committee Members 1. Chair, Department of Health 2. AOANJRR Director 3. A representative of a. Department of Health b. Australian Orthopaedic Association c. Consumer s Health Forum d. Therapeutic Goods Administration e. Prostheses List Advisory Committee f. Private Healthcare Australia g. Australian Private Hospitals Association h. Orthopaedic Industry (2) i. Medical Technology Association of Australia ii. Non Medical Technology Association of Australia The National Board of the AOA established the AOANJRR Committee to develop and manage AOANJRR policies. The Committee reports to the AOA Board. Members include the Chairperson, AOANJRR Director, three AOANJRR Deputy Directors and two Assistant Deputy Directors. In addition, an orthopaedic surgeon from each state, the ACT, and a representative from each of the AOA specialty arthroplasty groups are included. A complete list of the current AOANJRR Committee is provided in the acknowledgements section of this report. The Director, Deputy Directors and Assistant Deputy Directors are appointed by the AOA Board and are responsible for providing strategic and clinical guidance. Additionally, the Directors are responsible for ensuring the cooperation of hospitals, surgeons and government, maintaining the profile and reputation of the Registry, continued collaboration with other arthroplasty registries internationally, and sustaining the current level of excellence. The AOANJRR staff include the Registry Manager, Administration Officer, Research Coordinator and Prosthesis Library Coordinator. The AOANJRR team are responsible for the day-to-day operations, implementing new strategies, provision of data reports, research and publications activity, and coordinating the preparation of the Annual Report. DATA COLLECTION Hospitals provide data on specific Registry forms, which are completed in theatre at the time of surgery and submitted to the Registry each month. Examples of Registry data forms are available on the website. Hard copy forms are sent to the Registry where a small team of expert data entry staff enter the data directly into the database. Onsite Data Managers are available to resolve queries at the time of data entry to reduce any potential data entry errors. The Registry data entry system uses a predictive text function which greatly reduces the possibility of transcription errors and enables the experienced data entry staff to enter the data rapidly and accurately. The Registry has also established mechanisms to collect data electronically when it becomes feasible for contributing hospitals to do so. To date, there are no hospitals providing data electronically. DATA VALIDATION The Registry validates data collected from both public and private hospitals by comparing it to data provided by state and territory health departments. Validation of Registry data is a sequential multi-level matching process against health department unit record data. The validation process identifies: 1. Registry procedure records for procedures notified to state/territory health departments by hospitals. 2. State/territory records for procedures not submitted to the Registry by hospitals. 3. Exact match procedures, that is, records held by the Registry and state/territory health departments. 4. Procedures that match on some parameters, but which require additional checking with hospitals to enable verification. Initial validation is performed using hospital and patient identity numbers with subsequent verification undertaken on relevant procedure codes and appropriate admission periods. Data errors can occur within Government or Registry data at any of these levels; that is, Data Quality errors in patient identification, coding or admission period attribution by either the hospital, state/territory health department or the Registry. Data mismatches are managed depending on the nature of the error. For example, a health department record for a primary knee may match a Registry held record for a hip on all parameters except procedure type. The Registry would regard the Registry data to be correct in this instance as the Registry record contains details of the prostheses implanted. Other errors may be resolved by contacting hospitals for clarification. Most commonly, this may include a reassessment of procedure codes or admission period. The validation process identifies procedures not submitted to the Registry. As in previous years, the majority of these procedures have an ICD10 code for hemiarthroplasty of the femur. Sufficient information is provided in the state unit record data to enable the Registry to request hospitals to provide forms for unreported procedures. The Registry is able to obtain over 98% of joint replacement procedures undertaken in Australia. On initial submission of forms from participating hospitals, the Registry s capture rate is 96.8%. Following verification against health department data, checking of unmatched data and subsequent retrieval of unreported procedures, the Registry is able to obtain an almost complete dataset relating to hip, knee and shoulder replacement in Australia. OUTCOME ASSESSMENT The Registry describes the time to first revision using the Kaplan-Meier estimates of survivorship. The cumulative percent revision at a certain time, for example five years, is the complement (in probability) of the Kaplan-Meier survivorship function at that time, multiplied by 100. The cumulative percent revision accounts for right censoring due to death and closure of the database at the time of analysis. Mortality information is obtained by matching all procedures with the National Death Index (NDI) biannually. The NDI is the national mortality database maintained by the Australian Institute of Health and Welfare (AIHW). The AIHW requires ethics approval for access to the NDI data. 8 aoa.org.au aoa.org.au 9

Prior to 2013, the Registry reported the revisions per 100 observed component years. This statistic provides a good estimate of the overall rate of revision. However, it does not allow for changes in the rate of revision over time. A more informative estimate of the rate of revision over time is the cumulative percent revision. Confidence intervals for the cumulative percent revision are unadjusted point-wise Greenwood estimates and should not be used to infer significant differences in revision between groups. Reported hazard ratios should be used when judging statistical significance. Hazard ratios (HR) from Cox proportional hazards models, adjusting for age and gender where appropriate, are used to compare rates of revision. For each model, the assumption of proportional hazards is checked analytically. If the interaction between the predictor and the log of time is statistically significant in the standard Cox model, then a time varying model is estimated. Time points are iteratively chosen until the assumption of proportionality is met, then the hazard ratios are calculated for each selected time period. If no time period is specified, then the hazard ratio is over the entire follow up period. All tests are two-tailed at the 5% level of significance. The cumulative percent revision (CPR) is displayed until the number at risk for the group reaches 40, unless the initial number for the group is less than 100, in which case the cumulative percent revision is reported until 10% of the initial number at risk remains. This avoids uninformative, imprecise estimates at the right tail of the distribution where the number at risk is low. Analytical comparisons of revision rates using the proportional hazards model are based on all available data 1. In the presence of a competing risk for revision, the Kaplan-Meier method is known to overestimate the true probability of revision. Death of the patient before revision presents such a competing risk. In circumstances where the risk of death is high, e.g. in elderly patients with fractured neck of femur, the bias in the Kaplan-Meier estimates may be substantial and the reported cumulative percent revision should be interpreted with caution. The Registry is currently investigating the introduction of different analytical methods to cope with competing risks. Cumulative incidence is one method of estimating the probability of revision in the presence of competing risks. Cumulative incidence revision diagnosis graphs deal with the competing risks of reasons for revision, highlighting the differences between groups in the pattern of revision over time. They also provide important insight into different mechanisms of failure. This year the Registry has provided cumulative incidence of revision for primary total conventional hip replacement compared to other types of primary hip arthroplasty used for the management of fractured neck of femur. More detailed information on the statistical methods used in this report is presented in Appendix 2. An important Registry focus has been the continued development of a standardised algorithm to identify prostheses or combination of prostheses not performing to the level of others in the same class. The Registry refers to this group as prostheses with a higher than anticipated rate of revision. A three-stage approach has been developed and is outlined in detail in the relevant chapter of the report. REPORT REVIEW PRIOR TO PUBLICATION Prior to publication there are two workshops held to review, comment, and provide advice on the report. Members of the AOA and Arthroplasty Society are invited to attend a two-day workshop to review all sections of the report other than the shoulder procedures section. This workshop was held in Adelaide on the weekend of 5 and 6 August 2017. Members of the AOA with expertise in shoulder surgery are invited to attend a separate workshop to review this section of the report. This second workshop was held in Adelaide on 12 August 2017. Following these workshops, the report was provided to the AOA Board for consideration and final approval prior to publication. CATEGORIES OF ANKLE REPLACEMENT The Registry groups ankle replacement into two broad categories: primary total and revision ankle replacement. A primary replacement is the initial replacement procedure undertaken on a joint and involves replacing all (total) of the articular surface. Revision procedures are re-operations of previous ankle replacements where one or more of the prosthetic components are replaced, removed, or another component is added. Revisions include re-operations of primary total or previous revision procedures. Ankle revisions are sub-categorised into three classes: major total, major partial and partial minor revisions. These are defined in the chapter on revisions. Ankle Replacement DEMOGRAPHICS OF ANKLE REPLACEMENT TOTAL ANKLE There have been 2,235 ankle replacement procedures reported to the Registry. This includes two partial resurfacing ankle replacements from 2008. These are excluded in any further analyses. Ankle replacement is more commonly undertaken in males (60.8%) with the majority between the ages of 65 and 74 years. The mean age of patients is 66.2 years (Tables A1 A3 and Figure A1). The Registry has recorded 107 bilateral ankle replacements, 11.2% of which were performed within six months of the initial procedure (Table A4). MAJOR TOTAL MAJOR PARTIAL MINOR 1 Pocock SJ, Clayton TC, Altman DG. Survival plots of time to event outcomes in clinical trials: good practice and pitfalls, Lancet 2002; 359: 1686-89. 10 aoa.org.au aoa.org.au 11

Table A1 Number of Ankle Replacements by Gender Male Female TOTAL Ankle Replacement N Row% N Row% N Row% Primary Total 1107 60.0 738 40.0 1845 100.0 Revision 252 64.9 136 35.1 388 100.0 TOTAL 1359 60.8 876 39.2 2235 100.0 Note: Excludes two partial resurfacing procedures Primary Total Ankle Replacement Primary total ankle replacement includes the replacement of the talus and distal tibial articular surfaces with a tibial and talar component. The bearing surface may be fixed to the tibial component or a mobile bearing. Most procedures are performed on males (Figure A2). The median age is 66 years. In 2016, most procedures were undertaken in the 65 to 74 year age group (Table A5 and Figures A2 and A3). Table A2 Number of Ankle Replacements by Age <55 55-64 65-74 75-84 85 TOTAL Ankle Replacement N Row% N Row% N Row% N Row% N Row% N Row% Primary Total 182 9.9 590 32.0 730 39.6 314 17.0 29 1.6 1845 100.0 Revision 46 11.9 94 24.2 170 43.8 74 19.1 4 1.0 388 100.0 TOTAL 229 10.2 685 30.6 900 40.3 388 17.4 33 1.5 2235 100.0 Note: Excludes two partial resurfacing procedures DEMOGRAPHICS OF TOTAL ANKLE REPLACEMENT There have been 1,845 primary total ankle replacements reported to the Registry. The use of total ankle replacement has been declining since its peak in 2011. In 2016, there were 106 fewer total ankle replacements than in 2011. The principal diagnosis is osteoarthritis (92.8%) (Table A6). The most used talar/tibial prostheses are listed in Table A7. The Salto Talaris and Salto were the most commonly used prostheses in 2016. Table A3 Ankle Replacement by Age and Gender Gender Number Percent Minimum Maximum Median Mean Std Dev Male 1359 60.8% 23 94 67 66.7 8.9 Female 876 39.2% 20 89 66 65.5 10.2 TOTAL 2235 100.0% 20 94 67 66.2 9.4 Table A5 Primary Total Ankle Replacement by Age and Gender Gender Number Percent Minimum Maximum Median Mean Std Dev Female 738 40.0% 20 89 66 65.4 10.1 Male 1107 60.0% 23 94 67 66.6 8.8 TOTAL 1845 100.0% 20 94 66 66.1 9.4 Figure A1 Ankle Replacement by Age and Gender 25 20 Figure A2 Primary Total Ankle Replacement by Gender Figure A3 Primary Total Ankle Replacement by Age 200 Male Female 200 <55 55-64 65-74 75 Percent 15 10 100 100 5 0 50-54 45-49 40-44 35-39 30-34 25-29 20-24 <20 60-64 55-59 Male Age 100+ 95-99 90-94 85-89 80-84 75-79 70-74 65-69 Female 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 Table A4 Time Between Procedures for Bilateral Primary Ankle Replacement Same Day 1day-6months 6months TOTAL Bilateral Procedures N Total% N Total% N Total% N Total% Both - Total Ankle 4 3.7 12 11.2 91 85.0 107 100.0 TOTAL 4 3.7 12 11.2 91 85.0 107 100.0 12 aoa.org.au Data Period: 1 September 1999-31 December 2016 aoa.org.au Page 13

Table A6 Primary Total Ankle Replacement by Primary Diagnosis Primary Diagnosis Number Percent Osteoarthritis 1712 92.8 Rheumatoid Arthritis 101 5.5 Other Inflammatory Arthritis 11 0.6 Instability 8 0.4 Fracture/Dislocation 3 0.2 Osteonecrosis 3 0.2 Tumour 2 0.1 Other 5 0.3 TOTAL 1845 100.0 Table A7 10 Most Used Talar/Tibial Prostheses in Primary Total Ankle Replacement 2008 2012 2013 2014 2015 2016 N Model N Model N Model N Model N Model N Model 98 Mobility 71 Mobility 59 Salto 52 Salto 62 Salto Talaris 89 Salto Talaris 34 Hintegra 67 Hintegra 51 Mobility 40 Hintegra 34 Hintegra 24 Salto 18 Buechel-Pappas 65 Salto 48 Hintegra 28 Salto Talaris 32 Salto 21 Zenith 11 Salto 23 BOX 15 S.T.A.R 13 Mobility 18 Zenith 12 Hintegra 6 BOX 7 Zenith 11 BOX 12 S.T.A.R 4 S.T.A.R 10 Trabecular Metal Ankle Joint 1 (Eska) 2 S.T.A.R 8 Salto Talaris 6 BOX 4 Trabecular Metal 4 Inbone 1 Salto Talaris 6 Zenith 3 Zenith 1 BOX 4 S.T.A.R 1 Trabecular Metal 1 Inbone 2 Infinity 1 Infinity 1 BOX Custom Made 1 (Mutars) 10 Most Used 168 (6) 100.0% 236 (7) 100.0% 198 (7) 100.0% 155 (8) 100.0% 157 (9) 100.0% 168 (10) 100.0% Remainder 0 (0) 0% 0 (0) 0% 0 (0) 0% 0 (0) 0% 0 (0) 0% 0 (0) 0% TOTAL 168 (6) 100.0% 236 (7) 100.0% 198 (7) 100.0% 155 (8) 100.0% 157 (9) 100.0% 168 (10) 100.0% OUTCOME FOR ALL DIAGNOSES PRIMARY DIAGNOSIS The cumulative percent revision for osteoarthritis at five years is 10.2% and for rheumatoid arthritis is 4.6% (Tables A8 and A9 and Figure A4). REASON FOR REVISION Loosening is the most common reason for revision of primary total ankle replacement. This accounts for 29.2% of all revisions, followed by lysis (11.0%) and instability (10.4%) (Table A10). The cumulative incidence of the five most common reasons for revision is presented in Figure A5. Table A8 TYPE OF REVISION The main type of revision is of the insert component only (50.0%) (Table A11). RE-OPERATION The Registry has recorded 15 procedures where a re-operation was performed on a primary total ankle replacement and no component was removed or inserted. These are not included in the revision analysis. PROSTHESIS TYPES Revision Rates of Primary Total Ankle Replacement by Primary Diagnosis The outcomes of different prosthesis types are listed in Tables A12 and A13. Primary Diagnosis N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Osteoarthritis 154 1712 7430 2.07 (1.76, 2.43) Rheumatoid Arthritis 6 101 485 1.24 (0.45, 2.69) Other Inflammatory Arthritis 0 11 47 0.00 (0.00, 7.87) Other (5) 2 21 89 2.25 (0.27, 8.12) TOTAL 162 1845 8052 2.01 (1.71, 2.35) Note: Only primary diagnoses with over 10 procedures have been listed Table A9 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Primary Diagnosis CPR 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Osteoarthritis 2.5 (1.9, 3.4) 7.3 (6.1, 8.7) 10.2 (8.7, 12.0) 12.1 (10.3, 14.3) 13.2 (11.0, 15.6) Rheumatoid Arthritis 3.1 (1.0, 9.2) 3.1 (1.0, 9.2) 4.6 (1.7, 12.1) Other Inflammatory Arthritis 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) Other (5) 4.8 (0.7, 29.3) 4.8 (0.7, 29.3) 4.8 (0.7, 29.3) Note: Only primary diagnoses with over 10 procedures have been listed 14 aoa.org.au Data Period: 1 September 1999-31 December 2016 aoa.org.au Page 15

Figure A4 Cumulative Percent Revision 30% 25% 20% 15% 10% 5% Cumulative Percent Revision of Primary Total Ankle Replacement (Primary Diagnosis OA) Osteoarthritis Rheumatoid Arthritis 0% 0 1 2 3 4 5 6 7 8 9 10 Years Since Primary Procedure HR - adjusted for age and gender Osteoarthritis vs Rheumatoid Arthritis Entire Period: HR=1.68 (0.73, 3.85),p=0.219 Number at Risk 0 Yr 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Osteoarthritis 1712 1509 1135 720 299 157 37 Rheumatoid Arthritis 101 88 72 49 25 14 5 Table A10 Reason for Revision of Primary Total Ankle Replacement Revision Diagnosis Number Percent Loosening 45 29.2 Lysis 17 11.0 Instability 16 10.4 Infection 13 8.4 Pain 12 7.8 Implant Breakage Ankle Insert 10 6.5 Fracture 6 3.9 Prosthesis Dislocation 6 3.9 Arthrofibrosis 4 2.6 Heterotopic Bone 2 1.3 Wear Ankle Insert 2 1.3 Incorrect Sizing 2 1.3 Synovitis 2 1.3 Metal Related Pathology 1 0.6 Implant Breakage Tibial 1 0.6 Other 15 9.7 TOTAL 154 100.0 Table A11 Type of Revision of Primary Total Ankle Replacement Type of Revision Number Percent Insert Only 77 50.0 Tibial Only 24 15.6 Tibial/Talar 14 9.1 Arthrodesis 12 7.8 Talar Only 11 7.1 Removal of Prostheses 8 5.2 Minor Components 5 3.2 Cement Spacer 3 1.9 TOTAL 154 100.0 Figure A5 Cumulative Incidence Revision Diagnosis of Primary Total Ankle Replacement Total Ankle Cumulative Incidence 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% Loosening Lysis Instability Infection Pain 1.0% 0.0% 0 1 2 3 4 5 6 7 8 9 10 Years Since Primary Procedure 16 aoa.org.au aoa.org.au 17

Table A12 Revision Rates of Primary Total Ankle Replacement by Prosthesis Type Type N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Agility 0 2 18 0.00 (0.00, 20.25) Ankle Joint (Eska) 0 2 16 0.00 (0.00, 22.75) BOX 12 114 553 2.17 (1.12, 3.79) Buechel-Pappas 10 64 469 2.13 (1.02, 3.93) CCI 1 4 19 5.21 (0.13, 29.03) Custom Made (Mutars) 0 1 0 0.00 (0.00, 7485) Hintegra 33 405 1822 1.81 (1.25, 2.54) Inbone 0 5 3 0.00 (0.00, 119.9) Infinity 0 3 2 0.00 (0.00, 215.2) Mobility 59 567 3253 1.81 (1.38, 2.34) S.T.A.R 7 48 145 4.83 (1.94, 9.94) Salto 35 371 1420 2.47 (1.72, 3.43) Salto Talaris 2 188 224 0.89 (0.11, 3.23) Trabecular Metal 0 15 12 0.00 (0.00, 31.65) Zenith 3 56 97 3.10 (0.64, 9.07) TOTAL 162 1845 8052 2.01 (1.71, 2.35) Table A13 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Prosthesis Type CPR 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs BOX 2.7 (0.9, 8.0) 7.3 (3.7, 14.2) 12.1 (7.0, 20.5) Buechel-Pappas 1.6 (0.2, 10.6) 7.9 (3.4, 17.9) 9.5 (4.4, 19.8) 11.1 (5.4, 21.8) 13.4 (6.9, 25.3) 20.8 (11.0, 37.2) Hintegra 3.2 (1.9, 5.5) 6.4 (4.3, 9.4) 8.8 (6.2, 12.5) 11.0 (7.6, 15.6) Mobility 2.5 (1.5, 4.1) 7.5 (5.6, 10.0) 9.1 (7.0, 11.8) 10.5 (8.1, 13.7) 11.7 (8.9, 15.3) S.T.A.R 4.4 (1.1, 16.4) 16.3 (7.4, 33.6) 16.3 (7.4, 33.6) Salto 2.5 (1.3, 4.8) 6.1 (4.0, 9.3) 12.9 (9.1, 18.2) Salto Talaris 1.4 (0.3, 5.4) Zenith 2.2 (0.3, 14.4) 11.5 (3.4, 35.1) Other (7) 0.0 (0.0, 0.0) 11.1 (1.6, 56.7) 11.1 (1.6, 56.7) OUTCOME FOR OSTEOARTHRITIS AGE AND GENDER There is a difference in the rate of revision with respect to age. Patients aged 75 years or older have a lower rate of revision compared to patients aged less than 55 years and 55 to 64 years (Tables A14 and A15 and Figure A6). Gender is not a risk factor for revision (Tables A16 and A17 and Figure A7). FIXATION Fixation in total ankle replacement is predominantly cementless (Tables A18 and A19). It is not possible to assess if cement fixation on one or both sides of the joint has a different outcome, as too few of these procedures have been undertaken. MOBILITY This is the first year that the Registry has compared revision rates of fixed and mobile total ankle replacements used for the treatment of osteoarthritis. The majority of total ankle replacements are mobile (88.1%). The follow up of the fixed ankle group is also much shorter than for the mobile group. Early indications are that fixed total ankle replacement may have a lower risk of revision. However, increased procedure numbers and longer follow up are required to determine if this is the case (Table A20). BONE SURFACE INTERFACE Again, this is the first year that the Registry has analysed outcomes for primary total ankle replacement based on bone surface interface of the prostheses. The majority of total ankle replacements have supplementation of the cementless fixation surface using either HAp or CaP. At this time there is no evidence to suggest that this provides any advantage (Tables A21 and A22 and Figure A8. Note: Only prosthesis types with over 40 procedures have been listed 18 aoa.org.au aoa.org.au 19

Table A14 Revision Rates of Primary Total Ankle Replacement by Age (Primary Diagnosis) Table A16 Revision Rates of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA) Age N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) <55 20 151 699 2.86 (1.75, 4.42) 55-64 61 550 2493 2.45 (1.87, 3.14) 65-74 56 681 2919 1.92 (1.45, 2.49) 75 17 330 1319 1.29 (0.75, 2.06) TOTAL 154 1712 7430 2.07 (1.76, 2.43) Gender N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Male 102 1059 4636 2.20 (1.79, 2.67) Female 52 653 2794 1.86 (1.39, 2.44) TOTAL 154 1712 7430 2.07 (1.76, 2.43) Table A17 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA) Table A15 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Age (Primary Diagnosis OA) CPR 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs <55 5.5 (2.8, 10.7) 11.0 (6.7, 17.6) 14.7 (9.6, 22.3) 55-64 2.5 (1.5, 4.3) 7.8 (5.8, 10.6) 11.8 (9.1, 15.3) 14.7 (11.2, 19.1) 16.3 (12.4, 21.3) 65-74 2.2 (1.3, 3.7) 7.0 (5.2, 9.4) 9.1 (6.9, 11.9) 11.3 (8.6, 14.7) 11.3 (8.6, 14.7) 75 2.0 (0.9, 4.4) 5.0 (2.9, 8.4) 7.4 (4.6, 11.8) 7.4 (4.6, 11.8) CPR 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Male 2.9 (2.0, 4.1) 8.0 (6.4, 10.0) 10.7 (8.7, 13.0) 12.8 (10.4, 15.6) 13.8 (11.1, 16.9) Female 2.0 (1.1, 3.4) 6.1 (4.4, 8.4) 9.5 (7.2, 12.6) 11.1 (8.3, 14.7) 12.2 (8.9, 16.5) Figure A7 Cumulative Percent Revision of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA) 30% Male Female HR - adjusted for age Male vs Female Entire Period: HR=1.22 (0.87, 1.71),p=0.243 Figure A6 Cumulative Percent Revision of Primary Total Ankle Replacement by Age (Primary Diagnosis OA) 25% Cumulative Percent Revision 30% 25% 20% 15% 10% 5% <55 55-64 65-74 75 HR - adjusted for gender <55 vs 75 Entire Period: HR=2.35 (1.23, 4.50),p=0.010 55-64 vs 75 Entire Period: HR=1.90 (1.11, 3.26),p=0.019 65-74 vs 75 Entire Period: HR=1.49 (0.87, 2.57),p=0.148 Cumulative Percent Revision 20% 15% 10% 5% 0% 0 1 2 3 4 5 6 7 8 9 10 Years Since Primary Procedure 0% 0 1 2 3 4 5 6 7 8 9 10 Years Since Primary Procedure Number at Risk 0 Yr 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Male 1059 930 707 453 199 103 25 Female 653 579 428 267 100 54 12 Number at Risk 0 Yr 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs <55 151 130 108 80 35 17 3 55-64 550 497 383 242 117 58 11 65-74 681 605 447 278 98 57 19 75 330 277 197 120 49 25 4 Data Period: 1 September 1999-31 December 2016 Page 21 20 aoa.org.au aoa.org.au 21

Table A18 Revision Rates of Primary Total Ankle Replacement by Fixation (Primary Diagnosis OA) Fixation N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Cemented 0 11 43 0.00 (0.00, 8.48) Cementless 154 1693 7363 2.09 (1.77, 2.45) Hybrid (Tibial Cemented) 0 3 8 0.00 (0.00, 47.74) Hybrid (Talus Cemented) 0 5 17 0.00 (0.00, 22.32) TOTAL 154 1712 7430 2.07 (1.76, 2.43) Table A19 Yearly Cumulative Percent Revision of Cementless Primary Total Ankle Replacement (Primary Diagnosis OA) CPR 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Cementless 2.6 (1.9, 3.5) 7.3 (6.1, 8.8) 10.3 (8.8, 12.1) 12.2 (10.4, 14.4) 13.3 (11.1, 15.8) Table A22 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Bone Interface Surface (Primary Diagnosis OA) Cumulative Percent Revision 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Calcium Phosphate CaP Coating 2.2 (0.3, 14.4) 11.5 (3.4, 35.1) Double coated surface - porous titanium + HAP 3.2 (1.8, 5.5) 6.5 (4.4, 9.7) 9.1 (6.3, 12.9) 11.3 (7.8, 16.2) HAP Coated 2.9 (0.9, 8.6) 7.8 (4.0, 15.1) 12.8 (7.4, 21.7) Porous Coating 2.2 (1.3, 3.8) 7.7 (5.8, 10.2) 9.3 (7.2, 12.0) 10.8 (8.4, 13.9) 12.3 (9.5, 15.9) Titanium Plasma Spray + CaP 4.3 (1.1, 16.2) 16.6 (7.5, 34.4) 16.6 (7.5, 34.4) Titanium plasma spray and HAP 2.3 (1.2, 4.2) 5.7 (3.7, 8.7) 13.5 (9.4, 19.3) Other (2) 0.0 (0.0, 0.0) Note: Only Bone Interface Surfaces with 50 or more procedures have been listed Table A20 Revision Rates of Primary Total Ankle Replacement by Mobility (Primary Diagnosis OA) Mobility N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Fixed 2 203 246 0.81 (0.10, 2.94) Mobile 152 1509 7184 2.12 (1.79, 2.48) TOTAL 154 1712 7430 2.07 (1.76, 2.43) Table A21 Revision Rates of Primary Total Ankle Replacement by Bone Interface Surface (Primary Diagnosis OA) Figure A8 Cumulative Percent Revision of Primary Total Ankle Replacement by Bone Interface Surface (Primary Diagnosis OA) Cumulative Percent Revision 20% 18% 16% 14% 12% 10% 8% 6% Double coated surface - porous titanium + HAP HAP Coated Porous Coating Titanium plasma spray and HAP HR - adjusted for age and gender Double coated surface - porous titanium + HAP vs Titanium plasma spray and HAP Entire Period: HR=0.87 (0.53, 1.41),p=0.562 HAP Coated vs Titanium plasma spray and HAP Entire Period: HR=1.09 (0.56, 2.10),p=0.805 Porous Coating vs Titanium plasma spray and HAP Entire Period: HR=0.84 (0.55, 1.28),p=0.411 4% Bone Interface Surface N Revised N Total Obs. Years Revisions/100 Obs. Yrs (95% CI) Calcium Phosphate CaP Coating 3 56 101 2.98 (0.61, 8.70) Double coated surface - porous titanium + HAP 32 382 1724 1.86 (1.27, 2.62) HAP Coated 12 106 519 2.31 (1.19, 4.04) Porous Coating 64 587 3444 1.86 (1.43, 2.37) Titanium Plasma Spray + CaP 7 50 139 5.03 (2.02, 10.37) Titanium plasma spray and HAP 35 507 1469 2.38 (1.66, 3.31) Other (2) 1 24 34 2.95 (0.07, 16.42) TOTAL 154 1712 7430 2.07 (1.76, 2.43) Note: Only Bone Interface Surfaces with over 50 procedures have been listed. 2% 0% 0 1 2 3 4 5 6 7 8 9 10 Years Since Primary Procedure Number at Risk 0 Yr 1 Yr 3 Yrs 5 Yrs 7 Yrs 8 Yrs 9 Yrs Double coated surface - porous titanium + HAP 382 357 272 162 61 27 5 HAP Coated 106 102 86 51 20 5 0 Porous Coating 587 572 522 398 198 117 31 Titanium plasma spray and HAP 507 392 219 98 18 7 1 Note: Only Bone Interface Surfaces with over 60 procedures have been listed 22 aoa.org.au aoa.org.au 23

List of Tables Categories of Ankle Replacement... 11 Table A1 Number of Ankle Replacements by Gender... 12 Table A2 Number of Ankle Replacements by Age... 12 Table A3 Ankle Replacement by Age and Gender... 12 Table A4 Time Between Procedures for Bilateral Primary Ankle Replacement... 12 Demographics of Total Ankle Replacement... 13 Table A5 Primary Total Ankle Replacement by Age and Gender... 13 Table A6 Primary Total Ankle Replacement by Primary Diagnosis... 14 Table A7 Most Used Talar/Tibial Prostheses in Primary Total Ankle Replacement... 14 Outcome for All Diagnoses... 15 Table A8 Revision Rates of Primary Total Ankle Replacement by Primary Diagnosis... 15 Table A9 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Primary Diagnosis... 15 Table A10 Reason for Revision of Primary Total Ankle Replacement... 17 Table A11 Type of Revision of Primary Total Ankle Replacement... 17 Table A12 Revision Rates of Primary Total Ankle Replacement by Prosthesis Type... 18 Table A13 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Prosthesis Type... 18 Outcome for Osteoarthritis... 19 Table A14 Revision Rates of Primary Total Ankle Replacement by Age (Primary Diagnosis OA)... 20 Table A15 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Age (Primary Diagnosis OA)... 20 Table A16 Revision Rates of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA)... 21 Table A17 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA)... 21 Table A18 Revision Rates of Primary Total Ankle Replacement by Fixation (Primary Diagnosis OA)... 22 Table A19 Yearly Cumulative Percent Revision of Primary Total Ankle Replacement by Fixation (Primary Diagnosis OA)... 22 Table A20 Revision Rates of Primary Total Ankle Replacement by Mobility (Primary Diagnosis OA)... 22 Table A21 Revision Rates of Primary Total Ankle Replacement by Bone Interface Surface (Primary DiagnosisOA.22 Table A22 Yearly Cumulative Percent Revision of Primary Total Ankle by Bone Interface Surface (Primary Diagnosis OA).23 List of Figures Demographics of Ankle Replacement... 11 Figure A1 Ankle Replacement by Age and Gender... 12 Demographics of Total Ankle Replacement... 13 Figure A2 Primary Total Ankle Replacement by Gender... 13 Figure A3 Primary Total Ankle Replacement by Age... 13 Outcome for All Diagnoses... 15 Figure A4 Cumulative Percent Revision of Primary Total Ankle Replacement (Primary Diagnosis OA)... 16 Figure A5 Cumulative Incidence Revision Diagnosis of Primary Total Ankle Replacement... 17 Outcome for Osteoarthritis... 19 Figure A6 Cumulative Percent Revision of Primary Total Ankle Replacement by Age (Primary Diagnosis OA)... 20 Figure A7 Cumulative Percent Revision of Primary Total Ankle Replacement by Gender (Primary Diagnosis OA)... 21 Figure A8 Cumulative Percent Revision of Primary Total Ankle Replacement by Interface Surface(Prim Diagnosis OA)..22 24 aoa.org.au aoa.org.au 25

26 aoa.org.au