Biases in a randomized comparison of three types of screw fixation in displaced femoral neck fractures

We studied fixation of displaced femoral neck fractures prospectively in a randomized multicenter study, comparing 2 Olmed screws, 2 Tronzo screws and 3 Ullevaal hip screws. The study population consisted of 482 women and 125 men, of whom 432 women and 100 men were older than 65 years of age. Their median age was 80 (54-97) years. Despite agreement on criteria, the rates of reoperations for pain and failure--salvage (prosthesis replacement) and other reoperations (removal of implant)--differed significantly between the 3 hospitals regardless of type of fixation. In total, the percentages of salvage operations were: Olmed screw 17/175, Tronzo 17/130 and Ullevaal screw 11/302 (n.s.); the percentages of other reoperations were 11, 6 and 13, respectively (n.s.). In the whole series, the 2-year rate of salvage operations was 14%. No differences between the implants were found in patients older than 65 years of age. We conclude that an agreed, common definition of a hard end-point (reoperation) does not ensure comparability of results, because of differences in clinical decision making.