Canal Flare Index in the Canine Femur Is Influenced by the Measurement Method.

Abstract

We evaluated whether differences exist among the canal flare index (CFI) values obtained by different calculation methods in the veterinary literature. The endosteal widths were measured from radiographic images of canine cadaveric femora. Three different formulae were used to calculate the CFI. The CFILT-M was the ratio of endosteal width at the medial aspect of the lesser trochanter (LT) to the midshaft (M), while the CFILT-I was the ratio of LT to the isthmus (I). The CFIPLT-I was the ratio of endosteal width at the proximal aspect of the lesser trochanter (PLT) and the I. The widths at each level and the CFI calculation methods were compared. Using Rashmir-Raven's method, the femora were typed as stovepipe, normal and champagne fluted. The limits of agreement were also evaluated. The endosteal width at the proximal aspect of the lesser trochanter was 12% wider than at the medial aspect and 8% wider at the midshaft than at the isthmus. The CFILT-M was less than CFILT-I and CFIPLT-I by 9 and 20%, respectively. By Rashmir-Raven's classification, the CFILT-M method provided 18% stovepipe, 79% normal and 3% champagne fluted femora. The CFILT-I method showed the stovepipe, the normal and the champagne fluted as 6, 82 and 12%, respectively. The CFIPLT-I method classified the femora either normal (55%) or champagne fluted (45%). The comparison of CFILT-M with the other methods using Bland-Altman analysis showed lower mean difference for the CFILT-I than the CFIPLT-I. The level of width measurements at proximal femora might have an impact on the CFI values, likewise, preoperative planning procedures and the selection of a stem type in total hip arthroplasty.