A. This is the finalised protocol. B. Details of review team

Transcription

1 Protocol for rapid review of the clinical effectiveness and cost-effectiveness of metal on metal hip resurfacing (arthroplasty) for treatment of hip disease. A. This is the finalised protocol B. Details of review team Correspondence to: Vale, Luke, Mr. Research Fellow, Lead Reviewer Health Services Research Unit and Health Economics Research Unit University of Aberdeen, Foresterhill, Polwarth Building Aberdeen AB25 2ZD Tel: (01224) Fax: (01224) ; Alphabetical List of Other Review Team Members: Ashcroft, Paddy, Mr. Consultant, Orthopaedic Surgeon Woodend Hospital Grampian University Hospitals, NHS Trust Aberdeen AB15 6LS Tel: (01224) Gibson, Peter, Mr. Consultant, Orthopaedic Surgeon Woodend Hospital Grampian University Hospitals, NHS Trust Aberdeen AB15 6LS Tel: (01224) Fax: (01224) McCormack, Kirsty, Ms. Research Assistant, Systematic Reviewer Health Services Research Unit University of Aberdeen, Foresterhill, Polwarth Building Aberdeen AB25 2ZD Tel: (01224) Fax: (01224) ; McKenzie, Lynda, Ms. Research Fellow, Health Economist Health Services Research Unit and Health Economics Research Unit University of Aberdeen, Foresterhill, Polwarth Building Aberdeen AB25 2ZD Tel: (01224) ext Fax: (01224) ; Stearns, Sally, Dr. Senior Research Fellow, Health Economist 1

2 Health Economics Research Unit University of Aberdeen, Foresterhill, Polwarth Building Aberdeen AB25 2ZD Tel: (01224) Fax: (01224) ; Wallace, Sheila, Ms. Research Fellow, Information Specialist Health Services Research Unit University of Aberdeen, Foresterhill, Polwarth Building Aberdeen AB25 2ZD Tel: (01224) Fax: (01224) ; Steering Committee: Mr. John Cairns, Professor Adrian Grant, Professor Peter Fayers, Professor W. Cairns Smith C. Full title of research question Systematic review of the clinical effectiveness and cost-effectiveness of metal on metal hip resurfacing (arthroplasty) for treatment of hip disease. D. Clarification of research question and scope This review will assess whether metal on metal hip resurfacing (arthroplasty) is more effective and cost-effective than alternative treatments for three groups of people. The alternative treatments include total hip replacement, arthroscopy, intertrochanteric and acetabular osteotomy, bone fusion, and watchful waiting. The three groups are: People with hip disease who are likely to outlive the life of a total hip replacement (e.g. those aged less than 65); People with hip disease who would not be expected to outlive their prosthesis because of age (e.g. those over 65) but who participate in activities predicted to shorten the life of a total hip replacement and therefore would outlive the life of a total hip replacement; People with hip disease who are not suitable for consideration for total hip replacement for reasons other than expected survival or activity. The analysis will focus on long term outcomes (e.g. functional result, pain relief, revision rates, time to revision, and quality of life) but will also assess short term outcomes relating to the operative procedure (e.g. operative complications, length of hospitalisation, and time to return to usual activities). E. Report methods E.1. Search strategy 2

3 E.1.1 Studies of metal on metal resurfacing. Extensive electronic searches of the databases listed below will be conducted to identify both published and unpublished information. These searches will aim to identify existing systematic reviews and primary studies evaluating the effectiveness and cost-effectiveness of metal on metal hip resurfacing. The search terms for metal on metal hip resurfacing will be built up by investigating the Medical Subject Headings (MeSH) as well as textword searching. Databases to be searched include: Cochrane database of systematic reviews (CDSR) Database of abstracts of reviews of effectiveness (DARE) Cochrane Controlled Trials Register Cochrane Musculoskeletal Group Specialist Trials Register MEDLINE and PREMEDLINE EMBASE HealthSTAR CINAHL CRD NHS EED Relevant audit databases Worldwide Web Further citations will be sought from the reference list of all retrieved studies. Current research registers will be searched and relevant professional and research organisations will be contacted. All abstracts will be assessed for subject relevance by one reviewer. Full papers will be obtained for all studies that appear potentially relevant and formally assessed. E.1.2 Studies of alternative treatments to metal on metal hip resurfacing. The literature on metal on metal hip resurfacing will be quite limited, so we anticipate using data from studies of the alternative treatments which do not include direct comparisons with metal on metal resurfacing to model the effectiveness and costs of these treatments as alternatives to metal on metal hip resurfacing. The alternative treatments have been in use longer than metal on metal hip resurfacing, so we anticipate utilising additional exclusion criteria to keep the number of studies manageable. For example, for searches for arthroscopy, we may exclude any studies containing the term "hip fracture". The bibliographic databases MEDLINE, EMBASE and The Cochrane Central Controlled Trials Registry will be searched. Further citations will be sought from the reference list of all retrieved studies. We do not anticipate searching for watchful waiting but hope to identify papers on the natural progression of hip disease for the background information of the rapid review. This information along with advice from clinical colleagues will be used to estimate effectiveness and cost-effectiveness of watchful waiting. Table 1 summarises the different approaches we will initially use. 3

4 Table 1: Initial Search Strategy Inclusion Criteria for Alternative Treatments Alternative Type of Studies Study Age or Study Follow-up Treatment Search Criterion Total Hip Replacement Searched through 1995 n.a. Arthroscopy Osteotomy Bone Fusion HTA reports (systematic reviews) NICE update to HTA reports RCTs Geographically based registers RCTs and reviews Case series RCTs and reviews Case series RCTs and reviews Case series Searched through Nov 1999 Dec 1999 onwards onwards Targeted search* n.a. n.a. n.a. n.a Minimum mean or median of 5 years 1998 onwards n.a Minimum mean or median of 5 years Targeted search* n.a. Minimum mean or median of 5 years * Targeted search for reports of studies identified through consultation with clinical colleagues and external advisors E.2. Inclusion and exclusion criteria E.2.1. Types of studies E2.1.1 Metal on metal resurfacing Due to the anticipated scarcity of data on metal on metal hip resurfacing, randomised controlled trials, comparative observational studies, and single prosthesis observational studies will be eligible for inclusion in the review. The following criteria will be applied: Searches for randomised controlled trials will be limited to those that provide patient outcome data and compare metal on metal hip resurfacing with other interventions. Studies not reporting patient outcome data, such as laboratory only studies, will be excluded. There will be no restriction placed on the length of follow-up. Searches for comparative observational studies will be limited to those with concurrent controls that compare metal on metal hip resurfacing with other interventions and provide revision rates, clinical assessment (expressed as a global scale or as a rating of pain or function), or patient-based outcome. There will be no restriction placed on the length of follow-up. 4

5 Searches for single prosthesis observational studies of metal on metal hip resurfacing will be limited to those that provide revision rates, clinical assessment expressed as a global scale or as a rating of pain and function, or patient based outcome. Tentatively, a minimum of 2 years follow-up will be applied, although a longer follow-up criterion will be used (e.g., mean or median of 5 years) if sufficient studies with longer follow-up are found. Studies reported in languages other than English will be identified from their abstracts but will not be included in the review. E2.1.2 Alternative interventions Since two HTA systematic reviews and a NICE update of those reviews have been conducted for total hip replacement, we anticipate relying heavily on those sources and only using additional RCTs of total hip replacement that have been conducted since For the other alternative interventions, searches will be limited to systematic reviews and single intervention observational studies. Intertrochanteric osteotomy and arthroscopy are infrequently performed in the UK, and bone fusion and acetabular osteotomy are generally performed only for specific indications. Therefore, data on arthroscopy will come from the published case series for those surgeons performing the procedure in the UK. Data and bone fusion will be sought from published case series. For both arthroscopy and bone fusion, reports will be identified from targeted searches for reports of studies identified through consultation with clinical colleagues and external advisors. Data on intertrochanteric osteotomy will be derived from published case series from countries where osteotomy is more frequently performed. Included studies will be limited to those that provide revision rates, clinical assessment expressed as a global scale or as a rating of pain and function, or patient based outcome. Tentatively, a minimum of 2 years follow-up will be applied, although a longer follow-up criterion will be used (e.g., mean or median of 5 years) if sufficient studies with longer follow-up are found. Studies reported in languages other than English will be identified from their abstracts but will not be included in the review. E.2.2 Population People with hip disease who: Are likely to outlive the life of a total hip replacement (e.g. those aged less than 65) Would not be expected to outlive their prosthesis because of age (e.g. those over 65) but who participate in activities predicted to shorten the life of a total hip replacement and therefore would outlive the life of a total hip replacement Are not suitable for consideration for total hip replacement for reasons other than expected survival or activity. The patient population will further be split by gender and diagnosis of hip disease, e.g. primary osteoarthritic, secondary osteoarthritic (previous trauma or childhood congenital hip disease) and inflammatory hip disease (rheumatoid arthritis). 5

6 E.2.3 Types of interventions Metal on metal hip resurfacing, compared with total hip replacement, arthroscopy, osteotomy, bone fusion and watchful waiting. E.2.4 Types of outcome measures The following items will be sought for all included studies: A. Short term outcomes 1. Duration of operation 2. Conversions 3. Serious complications (e.g. nerve palsy, infection) 4. Time in hospital 5. Time to return to normal activities B. Long term outcomes 1 The rate of revision surgery or first operation following conservative treatment 2 Time to revision surgery or first operation following conservative treatment 3 Functional result (e.g. Harris, Mayo, Charnley scores) 4 The percentage of patients pain-free (clinician or analyst assessment) 6. Quality of life (e.g. SF36) 7. Mortality Bone conservation is not included as a review outcome. If the alternative treatments are to delay or prevent total hip replacement, then it is assumed that they would be bone conserving. E.3 Data extraction strategy A draft data extraction form (appendix H1) has been developed to record details of trial methods, participants, interventions, patient characteristics and outcomes. Two reviewers will extract data independently. Where a difference of opinion exists, the two reviewers will consult an arbiter. E.4 Quality assessment strategy All studies that meet the selection criteria will be assessed for methodological quality. The review will adopt a strategy based on a check-list developed by Morris to assess the quality of studies appearing in orthopaedic research journals (appendix H2). 1 This will provide a simple standardised expression of the quality of studies. 1 Morris R. A statistical study of papers in the Journal of Bone and Joint Surg Br J Bone Joint Surg Br 1988; 70:

7 Groups of studies (RCTs, comparative observational studies, single prosthesis observational studies) will be scored in each of the six areas assessed: Clarity of study question and definition of outcome Description of prosthesis and fixation Description of study sample Control of bias in study design Duration of follow-up Statistical and analytical considerations The methodological quality of eligible studies will be assessed by two reviewers independently. Where a difference of opinion exists, the two reviewers will consult an arbiter. E.5 Methods of analysis/synthesis If more than one eligible study is identified, the use of quantitative synthesis will be explored. Where this is not possible, a narrative synthesis of eligible studies will be undertaken. E.6 Methods for estimating qualify of life, costs and cost-effectiveness and/or cost/qaly A model will be developed to estimate the relative cost effectiveness of metal on metal hip resurfacing compared with the alternative treatment strategies. This model will combine data on clinical effectiveness with cost data relevant to the UK NHS. Further details of the modelling and data requirements are summarised below. E.6.1 Cost data The primary perspective for the costing will be the NHS and Personal Social Services. Cost data will therefore include the direct health service costs associated with the treatment options. Quantities of resources used will be identified from consultation with experts, primary data from relevant sources and the reviewed literature. We anticipate that unit cost data will be extracted from the literature or obtained from other relevant sources (manufacturer price lists, NHS reference costs). All cost data will be converted to a single year (2000) in pounds sterling. The following data will be needed to estimate costs incurred by the NHS for a particular procedure. staff time costs, consumables, overheads and capital charges associated with the actual operative procedure length of stay post operative secondary care treatment during the period of hospitalisation and convalescence 7

8 post operative care in the community, including frequency of GP visits, medication, and any relevant community health care such as physiotherapy. In order to provide a wider societal perspective and to elicit the impact on patients and their families, those costs that fall upon patients will also reported where data are available. The range of costs that will be considered will be: 1) Patients time and travel costs (reported in natural units of resource use) 2) Changes in productivity due to diagnosis, treatment and convalescence as well as gains in productivity as a consequence of successful treatment (reported in natural units) Where appropriate costs will be discounted at 6%, the rate recommended in the NICE guidance to manufacturers and sponsors of submissions. The model will also require data on the following: Incidence and prevalence of hip disease in the UK by age group, gender and primary diagnosis Time to initial operation (watchful waiting only) or revision Outcome data (e.g., health state data in order to estimate QALYs using existing published QOL figures). Mortality rates E.6.2 Assessment of benefits A balance sheet will be constructed to list possible benefits and costs of metal on metal hip resurfacing versus alternative treatments. We anticipate that the main measures of benefit will be improved functioning, reduced pain, increased time to first revision of total hip replacement and improved outcome after revision surgery. If sufficient data are available from the literature, different outcomes will be ascribed utility values and QALYs will be estimated. Where appropriate effectiveness and other measures of benefit will be discounted at 1.5%, the rate recommended in the NICE guidance to manufacturers and sponsors of submissions. E.6.3. Modelling A Markov model will be used to estimate costs and effectiveness over the lifetime of cohorts of patients who initially receive the different management strategies. While the models presented in the recent HTA monographs will, of course, inform development of model, the natures of the comparators specified by NICE precludes its direct adoption. In addition, the precise nature of the model will be constrained by the data available. The results in terms of costs and effectiveness will be presented in terms of a balance sheet. A formal combination of costs and benefits will also be performed, although the type of economic evaluation will depend upon the findings. 8

9 Ideally, the results would be presented as incremental cost per QALY ratios for metal on metal resurfacing compared to the alternatives considered. As noted above, we will attempt to construct QALY measures, but it is unlikely that sufficient data will be available to construct these measures with substantial precision. Therefore, as a further aid to decision-makers, if MOM is more costly but more effective in delaying revision, then the main results will be presented in terms of some measure such as incremental cost per revision avoided at various points in time. As we do not expect to have any data for longer follow-up periods (e.g., ten years or more), the model may be most suitable for estimating the cost effectiveness of the treatment alternatives over a five year period. If the data allow, we will extend the estimates further (e.g., up to 15 to 20 years), though we expect that any such projections will be extremely tentative. E.6.4. Sensitivity analysis Sensitivity analysis will be applied to the model in order to assess the robustness of the results to realistic variations in the levels of the underlying data. Where the overall results are sensitive to a particular variable, the sensitivity analysis will be reported. Finally, the results of the evaluation will be used to estimate the total NHS cost implications under different scenarios of adoption of MOM treatment based upon the three categories of patients and the different diagnostic categories. F. Handling the company submission(s) Given the timetable, we need to receive industry submissions by 12/09/01. We will develop the economic model we judge to be most appropriate, using if necessary data contained in the company submission(s). These economic models will be assessed against the 35 point BMJ guidelines for reviewers. Strengths and weaknesses in terms of methodology adopted, reporting of results and conclusions will be described. It will then be compared with that provided by the model we develop so that differences in results can be highlighted. If the model we develop differs substantively from that put forward by any company, we will justify any assumptions made. Any 'commercial in confidence' data taken from the company submission will be underlined in the HTA report (followed by an indication of the relevant company name in brackets) so that the NICE secretariat can negotiate (before and during the Institute's consultation process) with industry the subsequent inclusion of such data in the HTA monograph publication or subsequent peer-review publications. G. Project management G.1. Timetable/milestones - Final protocol: Originally due 2 July Submitted 20 June

10 Progress report: 19 September 2001 The progress report will address the following matters: 1) Is the report progressing on schedule? 2) Confirmation of external reviewers, including job title and institution 3) Confirmation of date of receipt of industry submission (or notification if still outstanding) 4) Indication of whether extent of data marked in confidence within industry submission is unreasonable i.e. if the whole of the submission is marked in confidence 5) Optional opportunity to comment on any problems encountered in production of the report Draft final report: 28 November

11 G.2. Competing interests The NHS Orthopaedic Department in which our clinical collaborators (Paddy Ashcroft and Peter Gibson) practice is intending to conduct research on short term clinical outcomes of metal on metal resurfacing. It is the intention for this project to be substantially funded by the Scottish Executive Health Department. However, Corin Medical will supply the implants and provide a small proportion of the support costs (amount to be identified once arrangements are finalised). G.3. External reviewers: The rapid review will be subject to external peer review by at least two experts. These reviewers will be chosen according to academic seniority and content expertise and will be agreed with NCCHTA. We recognise that methodological review will be undertaken by the NICE secretariat and Appraisal Committee, but if the rapid review encounters particularly challenging methodological issues, we will organise independent methodological reviews. External expert reviewers will see a complete and near final draft of the rapid review and will understand that their role is part of external quality assurance. We will require peer reviewers to sign a copy of the NICE Confidentiality, Acknowledgement and Undertaking. We will return peer reviewers signed copies to NCCHTA. Comments from external reviewers and our responses to these will be made available to NCCHTA in strict confidence for editorial review and approval. The names of external reviewers will be supplied with the progress report submitted to NCCHTA. 11

12 Appendix H1: Draft Data Extraction Form for Technology Assessment Review (TAR): Metal on Metal Resurfacing Hip Arthroplasty Reviewer ID: Study Details Study ID: Refman ID: Authors: Title: Publication year or date of interim data collection: Publication source: Study Design RCT Comparative observational study Single prosthesis observational study Other: Participants Number of participants randomised or included in study: Criteria for Inclusion: Criteria for Exclusion: Relevance to principal question: Were participants Yes No Unclear Not applicable (a) Likely to outlive the life of a total hip replacement? (b) Unlikely to outlive the life of a total hip replacement? (c) Unsuitable for total hip replacement 12

13 Setting and Timing Setting of study: Recruitment period: Were the majority of procedures performed by the principal researcher? Yes No If Yes, what proportion of procedures did the principal researcher perform? Intervention Surgical technique N o of patients Intervention 1 Intervention 2 Intervention 3 Intervention 4 Patient Characteristics 13

14 Intervention 1 Intervention 2 Intervention 3 Intervention 4 Age (years) a Sex (M/F) Primary Osteoarthritis (%) Secondary Osteoarthritis (%) Rheumatoid arthritis (%) Comments: Outcomes Short term outcomes Followup period Intervention 1 Intervention 2 Intervention 3 Interventio n 4 Duration of operation (min) a Conversions (N o & specify) Time in hospital (days) a Time to return to normal activities (days) a Serious complications (specify) e.g dislocation, infection, reoperation with in 6 months, nerve palsy. a Mean (s.d) and/or median (I.Q.R.) 14

15 Long term outcomes Number and rate of revision surgery Number and rate of operation (conservative treatment groups only) Functional result (e.g. Harris, Mayo, Charnley scores) The % of patients pain-free Quality of life (e.g SF36) Mortality Other: Comments 15

16 Contact with Author Date: / / Signature: 16

17 Appendix H2: Draft Quality Assessment Form for Technology Assessment Review (TAR): Study ID: Metal on Metal Resurfacing Hip Arthroplasty Assessment item Yes No Unable to judge 1. Clarity of study question and definition of outcome a) Is the purpose of the study clearly stated? b) Is the definition of prosthesis failure clear? c) Is there a clear definition of primary outcome(s)? d) Are standardised outcome measures used? e) Are the outcome measures used appropriate for the purpose of the study? Not applicable No agreed rating 2. Description of prosthesis and method of fixation a) Is the prosthesis design adequately described? b) Is the method of fixation adequately described? 3. Description of study sample a) Is the method of selection of the sample adequately described? b) Are the study exclusion and inclusion criteria stated? c) Is the baseline sample clearly described in terms of basic characteristics (age, sex etc)? d) Is the study sample sufficiently homogenous in terms of disease/diagnosis? e) Is the study sample sufficiently homogenous in terms of co-morbidity? 4. Control of bias in study design a) Is the method of allocation random? b) Is the method of masking the patient to the intervention allocated stated? c) Were outcome assessors blind to intervention allocation? d) Are baseline values for groups compared? e) Has the study adequately controlled for confounding factors? 5. Duration and completeness of follow-up a) Are intervals between surgery and follow-up assessment clearly stated? b) Are reasons for loss of patients at follow-up stated? c) Are those lost to follow-up compared to the rest of the sample? d) Is there an appropriate length of follow-up? e) Is the length of follow-up at least 5 years? 6. Statistical and analytical considerations a) Has the study sample size been justified? b) Are the data clearly presented? c) Was the data analyst masked to interventions? d) Has type of statistical test and actual probability value been stated? e) Are statistical tests appropriate to study? f) Has the data been analysed by intention to treat? g) Is the sample on which failures are assessed adequate? h) Are conclusions justified by evidence? 17

18 Appendix H3: Background Overview Treatment of younger and more active people with disease of the hip joint is limited mainly because of the high activity loads that increase the rate of wear and lead to early implant failure with a conventional total hip replacement. Metal on metal hip resurfacing may be a suitable alternative to THR for this younger more active patient group. Approximately 37,700 THRs (7,100 of which were revisions) were performed in 1989/90 in England and Wales. Around 28% of these were for patients under 65 years of age, and 2% for people under Long term results for THR among younger patients indicate that the rate of revision is much greater than that among older patients, with approximately 15-20% more patients requiring revision among those under 65 years. 2 Even higher rates of revision following THR have been reported for the subgroup of young patients with osteoarthritis. 3 Irreversible bone loss occurs with THR, and the options for revision are limited following a failure. Revision hip replacements are more difficult to perform and generally have poorer outcomes. Younger patients with disease of the hip joint are currently advised not to have a THR until absolutely necessary, as prolonging the time before primary THR is felt to lead to better lifetime outcomes and does not appear to make future THR more difficult. Metal on metal (MOM) hip resurfacing is one alternative treatment for this patient group. Other alternatives include conservative management (including pain management and reduction in level of activity), and other surgical procedures such as arthroscopy, osteotomy and bone fusion. Technology MOM hip resurfacing involves removal and replacement of the surface of the femoral head with a hollow metal hemisphere, which fits into a metal acetabular cup. The technique is claimed to preserve femoral bone and therefore improve the outcome of 18

19 future replacements. MOM resurfacing has been developed from earlier metal on polyethylene implants that, despite early good results, had a high failure rate. The devices currently available include the Cormet 2000, Birmingham Hip, Conserve Plus and Wagner Resurfacing. There is to date a lack of long term evidence on the effectiveness or cost-effectiveness of MOM, although preliminary evidence suggests that patients receiving MOM are less likely to suffer from the problems associated with wear that troubled the earlier polyethylene prostheses. 4 Projected costs metal on metal resurfacing The Cormet 2000 resurfacing implant is priced at 1,500-2,000 per device, excluding staff time and theatre costs. The cost of the whole procedure has been estimated at 4,380 and is therefore roughly 10% more costly than the mean of 3,900 for a primary total hip replacement (range 3404 to 4434) (NHS reference costs). 5 Effectiveness There is a lack of randomised trials or other well-controlled studies for hip resurfacing implants. In general the reported trials have, at present, inadequate follow-up periods. A critical review of evidence reports that resurfacing prostheses were inferior to all cemented total hip replacements, 6 although most of the studies appraised used the old metal on polyethylene implants rather than the current metal on metal implants. Other uncontrolled observational studies of metal on metal resurfacing have found more encouraging results. For example a preliminary short-term follow up 4 of 70 patients implanted with metal on metal (Wagner) prosthesis found that sixty-four of the 70 patients had an outcome assessed as excellent or good. A larger study reporting outcome over five years over the course of the development of the McMinn hip found that peripherally expanded hydroxyapatite coated cup and cement metal head (116 patients implanted over 19 months) gave excellent outcomes. 19

20 References 1 Anonymous (2001). Metal on metal resurfacing hip arthroplasty (hip resurfacing). NCCHTA briefing paper. 2 Northmore-Ball MD Young adults with arthritic hips should be offered alternatives to total hip replacement. BMJ 315: Alec Miner (2001). Draft scope: metal on metal resurfacing. National Institute for Clinical Excellence 4 Wagner M, Wagner H. Preliminary results of uncemented metal on metal stemmed and resurfacing hip replacement arthroplasty. Clin Orthopaed Related. Res 1996;329S:S78-S88. 5 NHS National reference costs. Accessed 22 May Faulkner A, Kennedy JG, Baxter K et al. Effectiveness of hip prostheses in primary total hip replacement: a critical review of evidence and an economic model. HTA 1998;2(6). 20