Abstract
Arthrography in twenty-five painful hips of twenty-one patients after total hip replacement and in fifty-three asymptomatic hips of forty-two patients after this procedure showed that arthrographic evidence of loosening was not always associated with pain. More than one-fifth of the asymptomatic hips had arthrographic evidence of loosening. Furthermore, the loosening shown by arthrogram could be confirmed in only seven of twelve hips that were subsequently explored. In addition, there was no apparent relationship between arthrographic loosening and the presence of a Trendelenburg sign. This study casts doubt on the value of an arthrogram in the diagnosis of the cause of pain after total hip replacement.