Twenty-eight fresh fractures of the femoral neck in which an autogenous-bone graft was used in conjunction with a nail as the primary treatment were reviewed. The ages of the patients at operation ranged from fifty-seven to ninety-one, with an average of seventy-two. There were no operative deaths, but one patient died of an acute heart attack seventeen days after operation. Another patient could not be followed because she went to another city directly from the hospital. Of the twenty-six fractures remaining, twenty-two (84 per cent) came to solid bone union. Partial weight-bearing was started at an average of four months. The average time to full weight-bearing was six and a half months.

Of the twenty-two patients who had bone union, nineteen lived for more than four years after operation. Twelve showed no roentgenographic evidence of necrosis for at least five years; five of these patients had roentgenograms at ten years, still without evidence of necrosis.

The seven remaining united fractures showed some degree of avascular necrosis by roentgenogranms, an approximate over-all incidence of 36 per cent. Five of these patients had no complaints and the necrosis was picked up only by follow-up roentgenograms. In one of these patients, the necrosis improved appreciably as the years went on and the head became essentially normal in appearance. Only two of these seven patients (10 per cent of those with united fractures followed four years or more) had clinically disabling necrosis. Obviously it is not possible to give an accurate incidence of necrosis because so much depends on the interpretation of the roentgenograms and the length of follow-up. In this study the incidence of some degree of avascular necrosis was 36 per cent, a percentage based on long-term roentgenographic findings.

There was one infection in this series which cleared up in twenty-five days with antibiotics and no surgical intervention.

Degenerative joint changes were not seen in the absence of avascular necrosis of the head. These changes were manifest first in the head and only secondarily, if at all, in the acetabulum.

There was no appreciable shortening of the neck.

The behavior of the graft varied. In the cases of non-union it broke presumably at the time the fracture came apart. In all the other cases the graft was well incroporated although there was great variation in the reaction between the graft and host, particularly in the speed of absorption. Some grafts disappeared in the early months and some were clearly visible at ten years. In over half the cases the graft began to disappear in the head as early as elsewhere, or earlier. Most often the last portion to be replaced was the middle third. The speed of absorption did not seem to be related to the speed of union or the incidence of avascular necrosis.

This study is not yet conclusive. It is being continued under the same conditions except that the grafts are of fresh-frozen bank bone rather than of autogenous origin and a two-hole sideplate is used to prevent retraction of the nail. The results will be reported later when long-term roentgenographic studies are completed. I am now using this method of treatment in most displaced fractures of the femoral neck in which the patients are under eighty and too healthy to warrant use of a replacement prosthesis.