The results of 251 operations done for insertion of an internal hip prosthesis were evaluated by examination of the patient in ninety-one cases and by questionnaire in forty-eight. Of these 139 operations, 112 were considered successful (excellent, good, or satisfactory result) and twenty-two unsuccessful (poor or failure). In five cases the patient was too senile to permit evaluation of the result.

Operations done for fresh fractures had significantly better results than those done as salvage procedures for hip fractures with unsuccessful initial treatment. Although the over-all complication rate was not significantly different in the two groups, the high incidence of fracture of the femur during salvage operations on old unsuccessfully treated fractures was a significant hazard not found in any other group of cases. The high postoperative mortality in the patients operated on for fresh fracture might suggest that the procedure carries too great a risk, but it must be remembered that patients with fresh fracture constituted a biased group since this procedure was elected in poorer-risk patients. Thus, the mortality rate may be distorted. The fact that the Hueter approach gave significantly less pain in the patients with fresh fractures may indicate that this approach is preferable if the operation can be done through this rather limited exposure. It appears that use of a stem prosthesis to replace the femoral head in fresh fractures of the femoral neck is a justifiable procedure. However, only relatively short follow-up periods are available; and, therefore, the operation should still be limited to the more elderly patient.

In the patients operated on for malum coxae senilis, the end result was not as good as in those treated for fresh fracture or for salvage of an unsuccessfully treated old fracture of the femoral neck. For malum coxae senilis, a Judet prosthesis inserted through a Gibson approach gave a significantly poorer result with respect to hip pain than did a Frederick R. Thompson prosthesis inserted through a Smith-Petersen approach. From this study it appears that the former operation should be abandoned. Where the results with Eicher prostheses could be compared with those with Frederick R. Thompson prostheses, no significant difference could be found with respect to hip pain, muscle power, or hip-joint motion. Hip pain is less, and the end result is significantly better in the slender patient when the operation is done for malum coxae senilis.

Although the complication rate is high, this series includes our earliest experiences with hip prostheses. It is to be hoped that the incidence of complications will decrease with increasing experience. It is quite significant that when a major complication occurs the chances of a successful outcome are significantly reduced.