We set up a trial to test whether the KT1000 arthrometer would give consistent measurements of anteroposterior laxity when used by the same and different examiners on the knees of the same subjects on the same day. The results showed substantial inter- and intra-examiner variation in the measurements both of absolute displacement in single knees and of side-to-side differences between pairs of knees. This casts doubt on the reliability of the instrument when used to compare the results of different techniques for reconstructing injured cruciate ligaments.

One of the objectives of knee replacement is to correct flexion deformity, the frequent consequence of rheumatoid arthritis and osteoarthritis. A review of 697 primary and revision replacements carried out between 1969 and 1985 and followed up from 1 to 16 years found that such deformity was present in 61% of knees before the primary operation. Replacement reduced this to 17% and the improvement was usually maintained. The deformity was present in only 21% of the replacements which required revision and the second operation reduced this to 8%. Flexion contractures affected rheumatoid knees more often and more seriously than osteoarthritic knees, but arthroplasty was more successful in correcting the deformity in the former. All of the 11 types of prosthesis used achieved some degree of correction, but the Walldius hinge and the variants of the Freeman condylar design were the most successful. Surprisingly, the best outcome, in terms of pain and reduced need for revision, was found in the rheumatoid knees most seriously deformed before operation, but this association was absent in the osteoarthritic knees. Postoperative deformity in knees without pain or extreme weakness did not appear to influence the patients' ability to walk or to use stairs or a chair, as measured by unexacting tests in the clinic.

Many knee replacement prostheses, embodying various principles of design, are now available and there is need for a method by which valid comparisons of results can be made. An important criterion of success is durability, so the length of time the prostheses have been in situ must be taken into account. Such a method is proposed here and is applied to the results of 673 knee replacements, of nine different types, implanted at the same hospital between 1970 and 1983. A prosthesis was considered to have failed if it had been removed or persistently caused severe pain. Two types of prosthesis were found to be significantly less successful than the other seven, between which none consistently showed significant superiority. Results for the seven types were similar despite the facts that they had been used for knees with different degrees of damage, some as secondary implants, and that they were of different design and at different stages of technical development. The more recently introduced types of prosthesis, designed to have theoretical advantages, were found in practice to be no more successful than the models they superseded.

We examine the hypothesis that a knee replacement is most likely to survive successfully if it is stable with a coronal tibiofemoral angle close to 7 degrees of valgus, the accepted normal. The records of 428 knee replacements followed up for one to nine years were analysed. The highest success rate was indeed found in those so aligned at operation and such knees were most likely to remain stable. Nevertheless, half of the failures occurred in knees correctly aligned at operation and two-fifths in knees which had remained stable in this alignment; many failures must have been caused by factors other than malalignment. Some knees, well aligned at operation, deteriorated into severely varus or valgus positions; their failure rate was significantly higher than that for knees which remained normally aligned and higher also than for knees severely varus or valgus from operation onwards. Malalignment, in itself, may not be the most important cause of failure, though it probably does compound failure from other causes.

A method of analysing the results of knee replacement operations, which makes it possible to estimate the annual failure rate and the proportion of implants which will survive successfully for 10 years, is applied to the post-operative data for 365 operations, using Freeman, Sheehan and Manchester prostheses, carried out from 1972 to 1980 at Harlow Wood Orthopaedic Hospital. Two criteria of success are defined. On both, and for all prostheses, the annual failure rate is found to be much lower in the first two than in the later years. Thus, account must be taken of the period since operation if the success of different prostheses is to be validly compared. This is not possible using conventional statistical methods. But significant trends in annual failure rates and significant differences between prostheses can only be identified from samples larger than are usually available in individual series. Hence it will be necessary for many centres to co-operate in pooling results if reliable conclusions are to be reached and valid comparisons made.

In a consecutive series of 87 operations for knee replacement prospective observations were made of the pattern of articular destruction in the form of a triangle of erosion or osteophyte formation, which is seen on the medial femoral condyle, sometimes with an associated strip on the lateral femoral condyle. These areas represent the surfaces which are normally in contact with the anterior horns of the menisci when the knee is fully extended. The lesions in question were found only in association with a flexion contracture in 84 out of the 87 cases.