2002 - University of Washington Bone and Joint Sources

Accuracy and Reproducibility of Step and Gap Measurements
from Plain Radiographs After Intra-Articular Fracture of the
Distal Radius
WREN V. MCCALLISTER, M.D., JEFFREY M. SMITH, M.D., JEFF KNIGHT, B.S.,
AND THOMAS E. TRUMBLE, M.D.
Recent reports have highlighted
the important relationship
between residual intra-articular
displacement and the long-term
outcome of distal radius fractures.
There appears to be a general consensus
that accelerated development of
arthritis and worsened severity of
degenerative changes is associated with
residual step-off displacements greater
than 1 to 2 millimeters. Therefore, to
improve clinical outcome after intraarticular
fracture of the distal radius,
many reports in the literature strongly
advocate surgical reduction of
fragments with articular incongruities
greater than 1 to 2 millimeters. In each
of these studies, the intra-articular stepoff
and gap displacements were
measured in millimeters using plain
radiographs.
Definitive standards for describing
and quantifying articular surface
incongruity do not exist. Kreder et al.
in 1996 attempted to standardize the
technique for measuring intra-articular
displacement of distal radius fractures.
Sixteen observers examined six
radiographs of healed intra-articular
distal radius fractures and found that,
“…even experienced clinicians did not
readily agree on the size of step and gap
deformity.” Specifically, they found that
two experienced observers chosen at
random to measure step-off and gap
displacement would differ by 3
millimeters or more at least 10% of the
time, while the same observer making
repeat measurements would be
expected to differ by 2 millimeters at
least 10% of the time. Singer also
demonstrated significant interobserver
variability in the measurement of intraarticular
distal radius fracture fragment
displacement using plain radiographs.
To date, the accurate and reliable
measurement of intra-articular fracture
displacements of less than 2 millimeters
from plain radiographs has not been
established.
The purpose of this study was to
determine the accuracy and
reproducibility of measured step-off
and gap displacements involving the
articular surface of the distal radius.
Our hypothesis was that step-off and
gap displacements of 2 millimeters or
less would show poor accuracy and
reproducibility for both intra- and
inter-observer measurements. This
study differs from others in that by
using cadaver specimens, we have
created a gold standard for the fracture
displacements. Earlier studies, lacking
such a gold standard, were limited to
comparisons among observers and
could not comment on the accuracy of
measurement. Further, our cadaver
model mimics intra-articular fracture
in the acute setting. As Kreder et al.
acknowledge, “It may well be that some
parameters (such as step and gap
deformity) are more readily quantified
in the acute setting, since fracture lines
have not been obscured by the healing
process.” Finally, by computing
tolerance limits, we attempt to define a
reliable threshold for the accuracy and
reliability of plain film radiography in
the setting of intra-articular distal
radius fractures. Such information has
value to the clinician in the decisionmaking
process regarding the use of CT
imaging.
METHODS
A Melone three-part fracture was
created in cadaver forearms to generate
a total of 12 combinations of step and
gap displacement (Figure 1). The range
of step deformity was 3.5mm (0.5 min,
4.0 max) and the range of gap
deformity was 3.9mm (0.1 min, 4.0
max). The radiographs were examined
by 22 physicians (divided into three
groups based on skill level: Group 1: six
R1 and R2 level orthopaedic surgery
residents; Group 2: four R4 senior
orthopaedic surgery residents on the
Hand service, five Hand Fellows; Group
3: six experienced Attending hand
surgeons and one musculoskeletal
radiologist) in a blinded, randomized
fashion using a standardized technique.
Twice each physician (range 3 weeks to
Step-off
Gap
Inter Intra Inter Intra
ICC * 0.78 0.81 0.85 0.89
κκ ** 0.64 0.65 0.65 0.76
Tolerance (mm) 2.58 1.76 2.51 1.61
Table 1: Results for inter- and intra-observer measurements (see Methods for explanation). * Landis & Koch agreement criteria: 0.80 almost perfect.
2002 ORTHOPAEDIC RESEARCH REPORT 25