Foot & Ankle International - June 2017 - 687

Lintz et al

687

Table 1. Inter- and Intraobserver Reliability of Foot and Ankle
Offset Measurement in Different Groups.

Overall
Interobserver
Intraobserver
Normal
Interobserver
Intraobserver
Varus
Interobserver
Intraobserver
Valgus
Interobserver
Intraobserver

ICC

95% CI

0.99
0.97

0.00
0.02

0.99-1.00
0.92-1.02

0.95
0.96

0.06
0.04

0.83-1.07
0.88-1.05

0.98
0.93

0.02
0.09

0.94-1.02
0.75-1.11

0.99
0.95

0.00
0.05

0.99-1.00
0.84-1.07

Abbreviations: FAO, Foot and Ankle Offset; ICC, intraclass correlation
coefficient; SD, standard deviation.

was Gaussian. Additionally, intra- and interobserver reliability has proven excellent in all groups of normal, valgus,
and varus cases.
In the authors' opinion, the originality of this study lies
in the description of an entirely new concept (3D biometrics) for measuring HFA that avoided projection, rotation,
and operator errors related to traditional 2D methods. This
is concurrent with previous literature in which Richter
et al35 stated that WBCT imaging is a more accurate method
of measuring angles because it is not subject to rotation and
projection bias and because it is weightbearing.
Our technique's purpose was to limit the drawbacks of
2D measurement relating to rotational, projectional and
operative issues. With respect to HFA, there is no Gold
Standard or "true" measurement.3,6,22 Rotation and other
potential bias influence the value of HFA measurements.3
However, in most studies, reliability is assessed using intra
and interobserver reliability carried out on the same radiograph for each patient. So it is logical that the same observer
(intra observer), or 2 observers (inter observer), given the
same instructions, would find the same value. The real
question is rather if 2 radiographs of the same patient taken
at different times would result in the same value. To really
assess reliability, the radiographs would therefore have to
be repeated on each patient, which would be ethically difficult to explain. This was done in a previous study using a
cadaveric setup3 where a 30-degree difference in rotation of
the foot could result in a 50% difference in HFA value. So
repeatability truly lies in the radiographic setting. In setups
where the position of the foot, the height of the x-ray source
(angle between the horizontal and the direction of the
X-Ray beam), the distance of the x-ray source, the individual practice of the radiographer, and the measurement technique all influence the end result, repeatability may be
impaired. WBCT may permit better control over these

variables through regular assessment, as is enforced by
international regulation by scanning templates that incorporate markers of known length, angle, and spatial distribution so that it may be checked, without consequences for the
patients, that measurements made by the machine are reliable. There are other advantages, including reduced spatial
footprint, radiation dose (comparable to a series of 5 conventional radiographs), and time for acquisition (52 seconds). However, the clinical and economic efficiency of
this technology still remains to be proven.
The clinical relevance for this study is therefore to report
the possibility of developing dedicated 3D biometrics to
provide more accurate measurement tools for planning foot
and lower limb surgery in the future. This technology
should also enable reliable and accurate data recording for
the purposes of research and clinical audit. The software
here described is a framework, in which 3D biometric tools
such as FAO may be adapted suitably for WBCT.
The authors acknowledge some limitations in this study
such as the absence of a post hoc sample size calculation,
and the absence of a gold standard to compare to, which is
always an issue when developing new measurements. We
did not compare our measurements with a traditional HFA
measurement, for example, as determined by the Saltzman
view, which is generally accepted as the main method to
assess HFA alignment. This would have required additional
irradiation of patients, which could not be justified in this
study setting. In addition, because there is no "true" or gold
standard measure for HFA, it would not necessarily be an
appropriate comparison.
We believe that establishment of the "true" HFA measurement that has relevance in clinical practice requires
fully automatic measurement tools. A fully automated system may enable the gathering of data in great quantities in
order to correlate varying pathologies. A "better" HFA measure will be one that improves the discrimination of a normal case from a pathologic one. In the meantime, an
international collaborative effort has to be made to adapt
traditional 2D measurements to WBCT and validate their
use in a clinical setting. This would provide the basic tools
and guidelines to evaluate and validate 3D biometrics. In
the future, the involvement of reference centers will be
required to conduct cost and clinical-effectiveness analyses
for this technology. Such an evolution has been seen over
the last 15 years in the dental area.43,45

Conclusion
In conclusion, a semiautomatic software was successfully
used to assess HFA using a 3D biometric measurement,
FAO. This new concept may represent the way forward to
make the best of WBCT. Further research is warranted in
order to properly validate such tools for clinical use. An
international effort is required in order to adapt traditional

Table of Contents for the Digital Edition of Foot & Ankle International - June 2017