We report the 12 to 74 month results of our mark I saddle prosthesis after its use as a salvage device for gross loss of pelvic bone stock in 76 patients with failed hip arthroplasties. The implant transmits load between iliac bone and bare polish chrome-cobalt. Our clinical and radiological results indicate that a useful and stable articulation can be achieved in most cases, provided that continued deep infection can be avoided. The appearance of radiological sclerosis at the bearing site in successful cases seems to indicate that significant late migration will not occur. Based on our experience with the mark I prosthesis we have designed and developed a mark II model which has freedom of axial rotation of the saddle. Our early results in 40 cases show a significant improvement over the results which could have been predicted for the mark I device.

We reviewed a series of 91 patients with deep infection of a cemented total hip arthroplasty caused by coagulase-negative staphylococci (C-NS). Of these, 72 were treated by one-stage exchange arthroplasty with a failure rate of 13% due to recurrence of infection. The other 19 patients have started or completed treatment by a two-stage exchange without failure to date. In 27 of the 91 patients multiple strains of C-NS were discovered, many being resistant to previously used antibiotics. The use of gentamicin-containing cement in the primary arthroplasty was significantly associated with the emergence of gentamicin-resistant C-NS in subsequent deep infection. Bacteriological diagnosis of such infections must take into consideration the possibility that multiple strains of the organism are involved.

Closed rupture of the middle slip of the extensor hood of a finger is easily missed until the late appearance of a buttonhole deformity. Early diagnosis gives the best chance of satisfactory treatment, but Boyes' test becomes positive only at a late stage. A new test is described in which, from a 90 degree flexed position over the edge of a table, the patient tries to extend the proximal interphalangeal joint of the involved finger against resistance. The absence of extension force at the proximal joint and fixed extension at the distal joint are immediate signs of complete rupture of the central slip. The theoretical basis and the method of performing the test are discussed.

Exchange operation is recommended as the treatment of choice for most deep infections involving a total hip replacement. This revision arthroplasty comprises, in one stage, excision of soft tissue, removal of implant and cement, replacement with an appropriate implant using Palacos R acrylic cement loaded with an appropriate antibiotic and, more recently, systemic antibiotics. During our first 10 years without systemic antibiotics we have achieved an overall 77 per cent success rate from a first attempt in 583 patients and a 90 per cent success rate after subsequent exchange procedures. Morbidity is significant but acceptable. Success is defined as control of infection, no loosening, and useful function. The factors associated with failures include, in particular, specific infections (Pseudomonas group, Streptococcus group D, Proteus group, and Escherichia coli), delay in operation and inadequate antibiotic dosage in the cement.

In thirty-one rat tibiae, plugs of plain acrylic cement were inoculated with Staphylococcus aureus; these all remained contaminated at the end of two weeks when the animals were killed. Inoculation with known strains of Pseudomonas, Proteus and Gp. G Streptococcus resulted in 70 to 93 per cent persisting contamination. Gentamicin, to which the organisms were fully sensitive, was efficacious in controlling the infection (90 per cent plugs proving sterile after two weeks). Fucidin was less successful against Staphylococcus aureus although effective in vitro. Intravenous inoculation with a suspension of Staphylococcus aureus succeeded in contaminanting 70 per cent of sixty plain cement plugs when injected into the tail vein half an hour after closure of the leg wounds. Only 11 per cent of sixty-four plugs were so contaminanted when the injection was delayed for two weeks. This animal model is submitted as a possible future means of testing different antibiotic-cement combinations against infection.

Laboratory experiments and clinical investigations have confirmed the various claims made originally by Buchholz and Engelbrecht (1970) that antibiotic-loaded acrylic cement releases the antibiotic into the surroundings in useful concentrations. Palacos R cement released higher concentrations than CMW, Simplex and Sulfix brands of cement and over longer periods. Concentrations of gentamycin and fucidin were sufficient to penetrate dead cortical bone. These conclusions need to be assessed with animal studies, mechanical testing and clinical results before the ideal place of antibiotic-loaded acrylic cement is established.