Incidence of dislocation after hip arthroplasty. Comparison of different registration methods in 408 cases.

We studied four different methods of registration of dislocations after total hip arthroplasty (THA) carried out at Malmö General Hospital during 1979-1988. Established registers failed to incorporate more than half of the dislocations and approximately one third of the patients, compared with a manual retrospective review of the original operating cords and the patient files. In 22 percent of the cases the first dislocation occurred more than one year postoperatively. Therefore different studies must use similar methods of registration concerning all details about dislocations in order to allow an adequate comparison of THA dislocation parameters.