Classification of acetabular fractures. A systematic analysis of the relevance of computed tomography

The classification of acetabular fractures and especially the diagnosis of additional lesions can be misleading, when the personal experience is limit ed and the decisions are based only on conventional radiographs. The introd uction of Spiral-CT with multiplanar reformations and 3-D views has improve d the quality of visualization. Due to their higher costs, the need of thes e additional diagnostic tools is frequently questioned. This paper discusse s the relevance of plain radiographs, 2-D-CTs, 3-D-CTs and Femursubtraction -CTs (FsCT) for the classification of aceta bu lar fractures, based on a co ntrolled study. Methods: Thirty physicians with different levels of experie nce in acetabular surgery were divided in three groups of 10 each: group I comprised residents without operative experience in acetabular surgery, gro up II was physicians with 3-10 years of operative experience, and group III was experts in acetabular surgery. A total of 10 complete radiographic cas es of high quality providing all levels of preoperative diagnostics (plain radiographs, 2-D-CT, CT with multiplanar reformation, 3-D-CT, Fs-CT) of dif ferent acetabular fracture types were prepared. The task for each candidate was to classify the fracture according to Letournel and to identify all ad ditional injuries within the hip joint (e. g. marginal impaction, head frac tures, etc.). The different diagnostic "levels" could be ordered stepwise a ccording to personal need and no time limit was given. The case was finishe d when the candidate presented his final diagnosis. The use of the differen t radiographs, the preliminary diagnosis, the changes in diagnosis, and the final decisions were recorded. These findings were correlated with the dif ferent levels of experience and against a "consensus classification" which was generated by thorough discussion, and the use of intraoperative informa tion and postoperative radiographs not accessible to the candidates. Result s: The "correct" fracture classification based on plain radiographs was: gr oup I, 11%; group II, 32%; group ill, 61%. Based on 2-D-CT a "correct" diag nosis was reached by 30% in group I, by 55% in group II, and by 76% in grou p Iii. With consideration of the "transient forms" in acetabular fractures based on Letournel and the 3-D-CT used mainly by group I, the rate of "corr ect" classifications rose to 65% in group I, 64% in group II and 83% in gro up III. The modifiers were diagnosed "correctly" in group I by 37%, in grou p II by 56%, and in group III by 73%. The use of the 3-D-CT and especially the Fs-CT by group I resulted in an improvement in the rate of correct clas sifications to 61%, whereas in group II the Fs-Ct was used only exceptional ly. The 2-D-CT was the basis for the diagnosis of the additional lesions in acetabular fractures within all groups resulting in 73% complete diagnoses in group III. This study showed the importance of CT for the exact analysi s and classification of acetabular fractures. In particular, the secondary reformations in CT and the 3-D-views dramatically improved the rate of "cor rect" classifications in the group of surgeons with limited personal expere nce in acetabular surgery. This allows the less experienced an acceptable l evel of "correct" diagnoses, so that the treatment options can be weighed c orrectly. Among the "experts" a rate of divergent classifcations of approxi mately 20% was observed, especially in "transient" forms of acetabular frac tures.