Intertrochanteric fractures of the femur are usually classified into stable and unstable groups, partly to predict the likelihood of displacement with early weight-bearing after internal fixation. Of 82 internally fixed intertrochanteric fractures which would normally be classified as stable, 25 subsequently underwent impaction along the axis of the nail or screw; varus displacement occurred in another six. Of these 31 fractures, 10 were two-part fractures which had been anatomically reduced, and 12 were originally undisplaced. It did not seem possible to predict the subsequent instability from the pre-operative radiographs. Of the 82 fractures, 47 had been treated with a Jewett nail-plate and 35 by a Richards compression screw-plate. Two groups of stable fractures were recognised: Group 1 was undisplaced and Group 2 displaced. Axial impaction occurred in a quarter of all Group 1 fractures however treated. In Group 2 fractures it occurred in a quarter of those treated with a nail-plate, and in over a third of those treated with a sliding screw-plate. The six fractures that underwent varus displacement had all been treated with Jewett nail-plates. Penetration into the joint occurred in 10 patients, all of them in the group treated with Jewett nail-plates; four required re-operation.

One hundred and seven patients with intertrochanteric fractures of the femur treated with a Richards screw-plate were compared retrospectively with 103 patients treated with a Jewett nail-plate. The mortality and morbidity were similar in the two groups. In patients with comparable fractures, those treated with a Richards device mobilised more quickly and left hospital sooner, and more of them returned to their homes. Failures of stabilisation, both clinically and radiographically, were fewer in this group. The reliability of fixation with a correctly positioned Richards screw-plate could justify the omission of outpatient follow-up in all but a small group of patients with severe unstable fractures or grossly defective bone stock. The higher cost of a Richards implant and the slightly longer operative procedure were outweighed by savings in occupancy of acute and long-stay hospital beds.