In a randomised trial involving 598 patients with 600 trochanteric fractures of the hip, the fractures were treated with either a sliding hip screw (n = 300) or a Targon PF intramedullary nail (n = 300). The mean age of the patients was 82 years (26 to 104). All surviving patients were reviewed at one year with functional outcome assessed by a research nurse blinded to the treatment used. The intramedullary nail was found to have a slightly increased mean operative time (46 minutes (sd 12.3) versus 49 minutes (sd 12.7), p < 0.001) and an increased mean radiological screening time (0.3 minutes (sd 0.2) versus 0.5 minutes (sd 0.3), p <  0.001). Operative difficulties were more common with the intramedullary nail. There was no statistically significant difference between implants for wound healing complications (p = 1), or need for post-operative blood transfusion (p = 1), and medical complications were similarly distributed in both groups. There was a tendency to fewer revisions of fixation or conversion to an arthroplasty in the nail group, although the difference was not statistically significant (nine versus three cases, p = 0.14). The extent of shortening, loss of hip flexion, mortality and degree of residual pain were similar in both groups. The recovery of mobility was superior for those treated with the intramedullary nails (p = 0.01 at one year from injury).

In summary, both implants produced comparable results but there was a tendency to better return of mobility for those treated with the intramedullary nail.

A total of 455 patients aged over 70 years with a displaced intracapsular fracture of the proximal femur was randomised to be treated either by hemiarthroplasty or internal fixation. The preoperative characteristics of the patients in both groups were similar.

Internal fixation has a shorter length of anaesthesia (36 minutes versus 57 minutes, p < 0.0001), lower operative blood loss (28 ml versus 177 ml, p < 0.0001) and lower transfusion requirements (0.04 units versus 0.39 units, p < 0.0001). In the internal fixation group 90 patients required 111 additional surgical procedures while only 15 additional operations on the hip were needed in 12 patients in the arthroplasty group. There was no statistically significant difference in mortality between the groups at one year (61/226 versus 63/229, p = 0.91), but there was a tendency for an improved survival in the older less mobile patients treated by internal fixation. For the survivors assessed at one, two and three years from injury there were no differences with regard to the outcome for pain and mobility. Limb shortening was more common after internal fixation (7.0 mm versus 3.6 mm, p = 0.004).

We recommend that displaced intracapsular fractures in the elderly should generally be treated by arthroplasty but that internal fixation may be appropriate for those who are very frail.

There are a number of classification systems for intracapsular fractures of the proximal femur, but none has been shown to be practical with satisfactory reproducibility and accurate predictive value. We have investigated the AO classification and evaluated intra-and interobserver accuracy and its value in predicting treatment and outcome.

We found it to have very poor intra- and interobserver reliability and to be of limited predictive use for the outcome of treatment. A simplified system in which the subdivisions were allocated to one of three groups of undisplaced, displaced and basal fractures was found to be of value. We conclude that this is the only division which is appropriate for these fractures and that the AO system for intracapsular fractures is too complicated and should not be used.