Sources of Error in the Production and Measurement of Standard Radiographs of the Foot

Abstract

In procuring radiographs of the human foot for comparative purposes it was found that, by using the methods described, measurements of a chosen dimension (arch-height) on 36 films showed a standard deviation of ± 1·17 mm. In 10 contactcopies of a given radiograph the standard deviation of the measurements taken by two observers of this same dimension was ± 0·212 mm. It was not possible definitely to allocate the difference between these deviations to any particular source of variation. In the plantar view, measurements of angle between the axes of the first two metatarsals showed a standard deviation of ± 0·436 deg. (16 observations on contact-copies). In a series of 10 subjects each subjected to 10 different sets of circumstances under otherwise standard conditions, certain large alterations in the arch-height dimension were observed. Thus, flexing the weight-bearing knee in an otherwise standard position leads to a mean depression of the archheight of 3·9 mm. ± 0·933 mm. If in this position the trunk and leg are rotated internally (amounting to a forced, passive eversion of the foot in a weightbearing position) a mean diminution of 7·55 mm. ± 0·837 mm. is observed (approximately 14·5 per cent. of the total). Both these diminutions are very highly significant (probably < ·001). Active contraction of tibialis posterior with knee extended in a weightbearing position (standard stance) raises the "archheight" by a mean distance of 2·475 mm. ± 0·386 mm. This is also very highly significant (probably < ·001). A half-inch wedge under the medial border of the heel produces no significant raising of the arch. These results confirm the necessity for strict observation of the standard conditions laid down and themselves serves as a test of the validity of the method in a behavioural study. In the dimension used as an illustration the standard deviation of observations made under the conditions laid down may be expected not to exceed ± 1·5 mm.