[Evaluation of tibial torsion--comparison of clinical methods and computed tomography]

To determine variability between clinical goniometric methods and computed tomography (CT) in measuring posttraumatic malrotation of the tibia.

In a prospective study, absolute tibial torsion of both legs after unilateral fracture of the tibia as well as the difference between both legs (intra-individual torsional difference) was postoperatively determined with two goniometric and two CT methods in 40 patients (female : male = 16 : 24, mean age = 46 +/- 34 years).

The mean difference between goniometric and CT methods in determining intra-individual torsional difference was not significant. Nevertheless, variance of values was higher in goniometric measurement (up to +/- 11 degree), with the measurements of torsional difference showing two times greater standard deviation in interobserver variability of goniometric methods as compared to CT.

Goniometric methods do not significantly differ from the CT methods for determination of intraindividual torsional difference. The higher variance of interobserver values limits accuracy of goniometric methods. Therefore, goniometric determination of tibial torsion can only be considered an estimate but not a precise measurement.