Editorials

Joint responsibility: the need for a national arthroplasty register

Sixty two different replacement hip joints manufactured by 19 companies are currently available in Britain, half of which have been introduced in the past five years.¹ A minimum follow up of 10 years is normally required to judge a joint replacement successful, but for only eight of these prostheses are there published follow ups of over five years. Only the Charnley low friction arthroplasty, widely regarded as the gold standard, has 20 year follow up results published in a peer reviewed journal.² Since there is no requirement for new implants to be clinically tested before being introduced,³ the power of advertising, rather than a scientific report, may lead a surgeon to select one prosthesis in favour of another."⁴ How then can we find out whether any implant is better than any other?

It is impossible to compare hip and knee implants and their outcomes because different methods are used to collect and analyse data. A sound study would require standardisation of these, together with a large number of patients, requiring a multicentre study. Pilot arthroplasty registers have been established in Trent and to a limited extent in the North West region, but these can reflect only local or regional practice. If several centres are to be included, why not create a national register and produce a true reflection of practice in the United Kingdom? This has already been achieved, with considerable benefits, for organ transplantation in Britain and for total joint replacement in Sweden and Norway.

A national arthroplasty register would enable unbiased scientific evaluation of the outcome of any total joint replacement to be performed. This would allow optimisation of resources and services, leading to improved efficiency and clinical practice. Such a register would need to be prospective, and follow up should include clinical and survivorship studies as well as an assessment of patient satisfaction.

Are there alternatives to a register? There has been much interest recently in the use of stereo radiogrammetry and automated image analysis techniques for the early prediction of implant failure.^{5 6} The clinical application of these methods remains limited, however, as they are time consuming and expensive and used mainly as research tools for analysing individual implants. They could not at present be used to show the superiority of a new design over current implants as they require improved accuracy and further validation studies.⁷ These methods also ignore other factors that influence the outcome of arthroplasty and which a registry could include, such as surgical approach and cementation technique, instrumentation, the skill and experience of the surgeon, and patient related factors such as age and underlying disease.

Radiological follow up is helpful in detecting asymptomatic complications such as acetabular migration and endosteolysis, but who would be responsible for paying for annual follow up radiographs? Even without radiological follow up a national register could monitor orthopaedic outcomes while allowing accurate assessment of complications such as infection and thromboembolism. It would also permit the observation of trends and demography and the effects of the introduction of prophylaxis policies.

The Scandinavian arthroplasty registers have proved successful at monitoring patient demography and have allowed direct comparison of the cumulative revision rates of different implants. This has enabled Swedish surgeons to identify and avoid certain specific complications⁸ and to eliminate inferior implants from the market.⁹ Other factors such as cementation technique have recently been analysed, with valuable results that have altered clinical practice after a mean duration of follow up of 3.2 years.¹⁰

Potential difficulties of a national register include noncompliance, funding, finding appropriate staff to collect the data, and comparisons of NHS and private work. To make a national register work, all units and consultants would have to contribute their data to it, but to ensure this the register will have to maintain confidentiality by coding data for individual surgeons and providing information only on institutions. It will have to reassure clinicians that individual data will not be used for political or medicolegal purposes or for the construction of league tables.

Increasingly, hospital trusts are seeking a role in selecting implants as they market joint replacement operations as part of a package that includes follow up and revision. Some implants have been shown to have promising results at five years but have then developed unacceptably high failure rates between five and 10 years.^{9 11} This move towards offering a package of services provides an economic incentive to define the longevity and modes of failure of an implant so that units can set a realistic price for the package and ensure that both the trust and the patient get the best deal. Prices for implants range from £250 to £2000, and so far the two least expensive designs have the longest published follow ups and the most successful results.^{1 2} The current major influence on choice of implant--the marketing strategies of manufacturers--would be negated by an independent analysis of centrally stored automatically updated registry data, allowing choices to be made on sound clinical grounds.

Now is the time for orthopaedic surgeons to take joint responsibility for prosthesis selection and surgical technique. Most modern implants claim good early results and low failure rates, but long term studies in substantial populations are urgently needed to substantiate their claims of superiority over existing practice. If the art of joint replacement in the United Kingdom is to be advanced, the introduction of a national register is imperative.

Senior orthopaedic research fellow Regional orthopaedic audit coordinator Consultant orthopaedic surgeon Centre for Hip Surgery, Wrightington Hospital for Joint Disease, Wigan WN6 9EP

David H Sochart, Alison J Long, Martyn L Porter 

1. Murray DW, Carr AJ, Bulstrode CJ. Which primary hip replacement? J Bone Joint Surg B 1995;77:520-7.
2. Wroblewski BM, Siney PD. Charnley low friction arthroplasty of the hip: long term results. Clin Orthop Rel Res 1993;292:191-201.
3. Carr AJ, Morris RW, Murray DW, Pynsent PB. Survival analysis in joint replacement surgery. J Bone Joint Surg B 1993;75:178-82. [Medline]
4. Insall JN. Choices and compromises in total knee arthroplasty. Clin Orthop Rel Res 1988;226:43-8.
5. Walker P, Mai S, Cobb A, Bentley G, Hua J. Prediction of clinical outcome of total hip replacement from migration measurements on standard radiographs. J Bone Joint Surg B 1995;77:705-14.
6. Hardinge K, Porter ML, Jones P, Hukins D, Taylor C. Measurement of total hip prostheses using image analysis: the maxima technique. J Bone Joint Surg B 1991;73:724-8.
7. Bulstrode CJ. Total hip replacement: the way forward. Ann R Coll Surg Eng 1996;78:129-32. [Medline]
8. Knutson K, Lewold S, Robertsson O, Lidgren L. The Swedish knee arthroplasty register: a nationwide study of 30 003 knees 1976-92. Acta Orthop Scand 1994;65:375-86. [Medline]
9. Malchau H, Herberts P, Ahnfelt L, Johnell O. Prognosis of total hip replacement in Sweden: follow up of 92 675 operations performed 1978-1990. Acta Orthop Scand 1993;64:497-506. [Medline]
10. Havelin LI, Espehaug B, Vollset SE, Engesaeter LB. Effect of type of cement on early revision of Charnley total hip prosthesis. J Bone Joint Surg B 1995;77:1543-50.
11. Owen TD, Moran CG, Smith SR, Pinder IM. Results of uncemented porous coated anatomic total hip replacement. J Bone Joint Surg B 1994;76:258-62.