Braun et al.

Journal of Orthopaedic
Journal of Orthopaedic Surgery and Research (2023) 18:69
https://doi.org/10.1186/s13018-023-03552-8 Surgery and Research

RESEARCH ARTICLE Open Access

Intra‑ and interobserver reliability analysis

of pediatric lower limb parameters on digital
long leg radiographs
Sebastian Braun1*†, Marco Brenneis1†, Jana Holder1,4, Andrea Meurer1,3† and Felix Stief1,2†

Abstract
Background Malalignments of the lower extremity are common reasons for orthopedic consultation because it may
lead to osteoarthritis in adulthood. An accurate and reliable radiological assessment of lower limb alignment in chil-
dren and adolescents is essential for clinical decision-making on treatment of limb deformities and for regular control
after a surgical intervention.
Objective First, does the analysis of full-length standing anteroposterior radiographs show a good intra- and inter-
observer reliability? Second, which parameter is most susceptible to observer-dependent errors? Third, what is the
Standard Error of Measurement ­(SEM95%) of the absolute femoral and tibial length?
Methods Two observers evaluated digital radiographs of 144 legs from 36 children and adolescents with patho-
logical valgus alignment before a temporary hemiepiphysiodesis and before implant removal. Parameters included
Mechanical Femorotibial Angle (MFA), Mechanical Axis Deviation (MAD), mechanical Lateral Distal Femoral Angle
(mLDFA), mechanical Medial Proximal Tibial Angle (mMPTA), mechanical Lateral Proximal Femoral Angle (mLPFA),
mechanical Lateral Distal Tibial Angle (mLDTA), Joint Line Convergence Angle (JLCA), femur length, tibial length. Intra-
and interobserver reliability ­(ICC2,1), ­SEM95% and proportional errors were calculated.
Results The intra- and interobserver reliability for almost all measurements was found to be good to excellent (Intra-
ICC2,1: 0.849–0.999; Inter-ICC2,1: 0.864–0.996). The ­SEM95% of both observers was found to be ± 1.39° (MFA), ± 3.31 mm
(MAD), ± 1.06° (mLDFA) and ± 1.29° (mMPTA). The proportional error of MAD and MFA is comparable (47.29% vs.
46.33%). The relevant knee joint surface angles show a lower proportional error for mLDFA (42.40%) than for mMPTA
(51.60%). JLCA has a proportional error of 138%. Furthermore, the ­SEM95% for the absolute values of the femoral and
tibial length was 4.53 mm for the femur and 3.12 mm for the tibia.
Conclusions In conclusion, a precise malalignment measurement and the knowledge about S­ EM95% of the respec-
tive parameters are crucial for correct surgical or nonsurgical treatment. The susceptibility to error must be consid-
ered when interpreting malalignment analysis and must be considered when planning a surgical intervention. The
results of the present study elucidate that MAD and MFA are equally susceptible to observer-dependent errors. This

†
Sebastian Braun and Marco Brenneis have contributed equally to this work
and share first authorship
†
Andrea Meurer and Felix Stief have contributed equally to this work and
share last authorship
*Correspondence:
Sebastian Braun
sebastian.braun@kgu.de
Full list of author information is available at the end of the article

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​
mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 2 of 12

study shows good to excellent intra- and interobserver ICCs for all leg alignment parameters and joint surface angles,
except for JLCA.
Trial registration: This study was registered with DRKS (German Clinical Trials Register) under the number
DRKS00015053.
Level of evidence I, Diagnostic Study.
Keywords Lower limb deformities, Leg axis, Genu valgum, Radiological assessment, Reliability, Pediatric orthopedic,
Guiding growth

Background line (MAD) deviates more than 10 mm lateral or 15 mm

Osteoarthritis (OA) of the knee is a common cause for medial to the center of the knee [14], which is approxi-
knee pain, reduced joint motion and consecutive muscle mately 3° deviation of the physiological MFA. Further-
weakness leading to physical inactivity, reduction in qual- more, knowledge of the joint surface angles and the Joint
ity of life, disability and eventually a total joint replace- Line Convergence Angle (JLCA) is important to assess
ment [1]. Varus or valgus malalignment of the lower on which bones the guided growth intervention should
extremity has a major impact on the development and be performed. In addition, an estimate of the residual
progression of knee OA [2]. The valgus deformity leads growth can be made based on the absolute lengths of the
to increased load on the lateral compartment of the knee femur or tibia in relation to age and sex and thus the cor-
joint and the varus malalignment increases the load on rection potential of the guided growth intervention can
the medial compartment [3–6]. be determined [15–17].
To prevent the development of OA, growth control To the best of our knowledge no study has shown which
should be considered in children with severe axial devia- parameter or combination of parameters is more reliable
tion in the frontal plane. For children and adolescents, for setting the indication for a THE in children and ado-
there is a wide range of normal values concerning axis lescents. Therefore, we first aim to elucidate if the analy-
deviation of the lower limb, which need to be differen- sis of a full-length standing anteroposterior radiograph
tiated from pathological and pre-arthritic deformities of both legs shows a good intra- and interobserver reli-
prior to surgical intervention [7]. As a surgical treatment, ability concerning lower limb alignment measurements
temporary hemiepiphysiodesis (THE) of the distal femur in the frontal plane. A misjudgment or incorrect determi-
or proximal tibia using tension band plates, depending nation may result in a delayed or unnecessary indication
on the location of the pathologic joint surface angles, for THE and thus in an incorrect manipulation of the leg
has become established for axis correction of children axis. Second, it is important to clarify which of the inves-
and adolescents [8, 9]. However, the exact patient age for tigated alignment parameters is less prone to observer-
the best possible correction and the lowest probability of dependent errors in determining lower limb alignment.
rebound remains unclear. Furthermore, there is a lack of Third, as mentioned earlier, the correct and reproducible
defined parameters for the indication of THE. The defi- measurement of the absolute tibial and femoral length
nition of parameters like the Mechanical Axis Deviation is essential to estimate the growth potential. Therefore,
(MAD) or the Mechanical Femorotibial Angle (MFA) the measurement error of the absolute femoral and tibial
is complicated by the interindividual variation during length needs to be elaborated.
growth, the uncertainty of when MAD leads to patho-
logical joint moments and pressure loads as well as the Methods
determination method and the associated susceptibility Patients
to errors [4, 5, 10, 11]. Patients older than 8 years of age who showed an inter-
Malalignment of the lower extremity is one of the most malleolar distance of more than 10 cm received an
common reasons for an orthopedic consultation. An anterior–posterior-full length X-ray of both legs in the
accurate and reliable radiological assessment of lower standing position for objective assessment of the leg axis
limb alignment is essential for clinical decision-making [7]. From the 40 patients recruited, four patients were
on treatment of limb deformities and for regular con- excluded from this study due to unusable radiographs.
trol after a surgical intervention [8, 12]. The full-length This homogenous cohort finally consists of 36 patients
standing anteroposterior radiograph of both legs is the (median age of 12.81 years at baseline; Table 1) with
gold standard for assessment of lower limb alignment an idiopathic pathological valgus alignment of at least
in frontal plane [13]. Most commonly the indication for one knee according to the MAD of the lower limb were
a THE is set when the physiological mechanical bearing recruited between September 2014 and October 2021
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 3 of 12

Table 1 Patient characteristics (n = 36)

Time of Investigation Age (range) in years Height (range) in cm Weight (range) in kg BMI (range) in kg/m2

Implantation of tension band plates using THE 12.81 (10.50–15.38) 165.89 (145.20–182.90) 63.49 (31.80–89.30) 22.86 (14.95–29.29)
Removal of tension band plates using THE 13.72 (11.32–16.35) 171.83 (150.60–187.40) 71.25 (33.80–94.50) 23.88 (14.90–32.79)
Table 1 shows the anthropometric data of the collective at the time of implantation and removal of tension band plates using the temporary hemiepiphysiodesis
(THE). Values are presented as median and interquartile range in parenthesis

on a voluntary basis. Patients were included if the MAD or knee condyles, history of major trauma or sport injury
deviates more than 10 mm lateral [14] and were planned of the lower extremity, knee surgery within 12 months
to receive THE of one or both knees [3, 12, 18]. The ten- prior to inclusion in this study, chronic joint infec-
sion band plates used in this study were the Eight-Plates tions or previously received intraarticular corticosteroid
(Orthofix, Lewisville, TX, USA) or Pedi-Plates (Orthope- injections.
diatrics Inc., Warsaw, IN, USA). The aim of this growth
guidance technique is to straighten the leg axis. After a Radiographic malalignment analysis
successful operation by means of THE is an MFA of 0° The standardized, scalable, digital long leg X-rays were
(± 2°) or an MAD of 0 mm (± 6 mm). All patients were taken in an anterior–posterior orientation. A 25.4-mm
analyzed twice, directly before guided growth interven- (1 inch)-diameter metal ball was placed between the
tion (72 legs) and at the time of implant removal (the legs at the level to the bones next to the joint line of the
same 72 legs). Accordingly, patients were deliberately knees and used as a reference for determining the indi-
measured both at a time when a lower limb malalignment vidual magnification factor. The radiographic measure-

templating program, TraumaCad® (version 2.3.4.1; Voy-

was present and with corrected leg alignment to test the ments were performed with a commercially available
reliability during the entire treatment period (see Fig. 1).
In total 144 legs/data sets could be evaluated. ant Health, Petach-Tikva, Israel). Criteria for a valid full
Exclusion criteria were: rheumatoid arthritis, anterior leg length anteroposterior radiographic image were [18]:
cruciate ligament deficiency, neuromuscular disorders, patient standing in an upright, weight-bearing position,
achondroplasia or hypochondroplasia, sagittal plane both legs parallel to each other a shoulder width apart,
deformities (genu pro- and recurvatum), flexion contrac- fully extended knees, patellae centered over the femoral
tures in the hip or knee joint, leg length discrepancy of condyles pointing straight forward in order to avoid rota-
more than 10 mm, avascular necrosis of the femoral head tional errors [19, 20].

Fig. 1 Diagram of the prospective study design. THE—temporary hemiepiphysiodesis

Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 4 of 12

Definitions of the angles measured are described The Mechanical Axis Deviation (MAD) is defined
below (Fig. 2): as the distance between the center of the knee and the
Malalignment analysis was performed due to the mechanical axis of the limb (Fig. 2a–f ). Medial deviation
principles described by Paley et. al [12, 18]: The radi- (varus) is depicted as positive and lateral deviation (val-
ographic Mechanical Femorotibial Angle (MFA) is gus) as negative values.
defined as the angle formed by the line from the center The mechanical Lateral Distal Femoral Angle (mLDFA)
of the hip to the center of the knee (mechanical femur shows the angle between the mechanical femoral axis
line) and the line from the center of the knee to the and the tangent through the most convex points of each
center of the ankle (mechanical tibia line) (Fig. 2b) [21]. femoral condyle (Fig. 2g).
To find the center of the hip, an auxiliary circle around The mechanical Medial Proximal Tibial Angle
the femoral head was needed. The center point of the (mMPTA) depicts the angle between the mechanical tib-
circle was determined as center of the hip. The center ial axis and the line along the subchondral bone from the
of the knee joint was defined as the midpoint between tibial plateau (Fig. 2h).
the center of the intercondylar region and the center of The mechanical Lateral Proximal Femoral Angle
the eminentia intercondylaris. The center of the ankle (mLPFA) shows the angle between the mechanical femo-
was determined as the midpoint of the talar dome. ral axis and the line from the tip of the greater trochanter
Neutral alignment was defined as 0°, varus malalign- to the center of the hip (Fig. 2d).
ment as positive angles and valgus malalignment as The mechanical Lateral Distal Tibial Angle (mLDTA)
negative angles. is defined as the angle between the mechanical axis of

Fig. 2 Left (Image a–c) full-length standing anterior–posterior radiograph of the right leg; a – d center of the hip (circle); a Mechanical axis of
the limb and Mechanical Axis Deviation (MAD) (red lines); b Mechanical Femorotibial Angle (MFA) between mechanical femur line (blue line)
and mechanical tibia line (red line); c length of the femur and tibia (red brackets); d mLPFA mechanical Lateral Proximal Femur Angle; e mLDTA
mechanical Lateral Distal Tibia Angle; f detailed magnification from image a, Mechanical axis of the limb and MAD; g mLDFA mechanical Lateral
Distal Femur Angle; h mMPTA mechanical Medial Proximal Tibia Angle; i JLCA Joint Line Convergence Angle [18]
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 5 of 12

the tibia and the line through the medial and lateral talus The SEM estimates the distribution of repeated
shoulder (Fig. 2e). measurements with the same instrument around their
The Joint Line Convergence Angle (JLCA) depicts the "true" value. It quantifies the precision of individual
angle between the tangent of the most convex points of scores on tests and provides an absolute index of reli-
each femoral condyle and the line along the subchondral ability, in contrast to ICC, which is a relative measure
bone from the tibial plateau (Fig. 2i). of reliability [25]. The SEM is directly related to the reli-
The length of the femur describes the distance between ability of a test, i.e., the greater the SEM, the lower the
the center of the femoral head and the center of a line reliability of the test. The range of the ­ SEM95% (e.g.,
connecting the two most distal points of the medial and MAD ­SEM95% =  ± 3.31 mm → ­SEM95%Range = 6.62 mm)
lateral femoral condyle (Fig. 2c) [22]. was divided by the range of the respective norm val-
The length of the tibia describes the distance between ues considered as acceptable (e.g., ­MADAcptRange 15 mm
the center of a line connecting the two most proximal to 1 mm = 14 mm). This proportional error shows for
points of the medial and lateral proximal tibial plateau which parameter, the S ­ EM95% is proportionately lowest
and the center of a line. or highest.
To assess intra- and interobserver reliability, radio- Statistical data analysis was performed with SPSS (ver-
graphs for each patient were blinded and templated twice sion 26, IBM Corporation, New York, NY, USA). The sig-
by two independent and experienced observers (SB, MB) nificance level for all statistical tests was set at p ≤ 0.05.
familiar with the templating software. The two observers The following data were presented as mean and standard
are orthopedic residents with special focus on pediatric deviation (parametric variables) or median ± 25–75 per-
orthopedics and more than five years of experience ana- centiles (nonparametric variables).
lyzing lower limb alignments. To avoid recollection bias, The level of evidence of this study is a Level I, Diagnos-
the second evaluations from the same set of measure- tic Study.
ments were repeated at least 4 weeks apart, blinded to
the measurements provided previously. Results
From September 2014 to October 2021, full-length
Statistical analysis standing anteroposterior radiographs of the legs at two
To evaluate the reliability of the measurements, the different time points (implantation of tension band plates
intra- and interobserver intraclass correlation coeffi- and removal of tension band plates) from 72 legs of 36
cient ­(ICC2,1) was calculated. An important advantage of children and adolescents (16 girls and 20 boys) were
­ICC2,1 is that the coefficient can detect bias by the order included in this study. All the children and adolescents
in which pairs of data are compared and is therefore pref- had a valgus malalignment on both sides. Table 1 shows
erable to other correlation coefficients [23, 24]. For inter- the anthropometric data such as age, height, weight, and
observer ­ICC2,1 the results of the first measurement of BMI of the collective at the time of implantation and at
observer one are compared with the results of the first the time of removal of the tension band plates.
measurement of observer two. One-way random tests
with absolute agreement were used for each observer to Does the analysis of a full‑length standing anteroposterior
estimate intraobserver ICCs and two-way random tests radiograph of the lower limb show a good intra‑
with absolute agreement per set of measurements for and interobserver reliability concerning lower limb
each parameter were used to determine interobserver alignment measurements?
ICCs. ­ICC2,1 values < 0.70 indicate poor, 0.70 to 0.79 fair, The values MFA, MAD, mLPFA, mLDFA, mMPTA,
0.80 to 0.89 good, and 0.90 to 1.00 excellent reliability mLDTA, JLCA, length of femur, and length of tibia were
[23, 24]. measured twice on 144 legs by each observer (Table 2).
The Shapiro–Wilk test was used to test normal distri- For both observers, the intraobserver I­CC2,1 (Table 3)
bution of the analyzed parameters. Differences between in alignment analysis was good to excellent and likewise
observers were tested with Wilcoxon test (nonparamet- high, ranging from 0.849 to 0.999. In contrast, fair results
ric dependent variables) or paired t-test (parametric were only found for intraobserver ­ICC2,1 in analysis of
dependent variables). To assess the observer-dependent JLCA (0.607–0.676).
errors, we calculated the Standard Error of Measure- Interobserver ­ICC2,1 (Table 4) for alignment parameters
ments ­(SEM95%) from the interobserver ­ICC2,1. (MFA, MAD, mLPFA, mLDFA, mMPTA, mLDTA, femur
SEM = SD × 1 − ICC2,1 length and tibia length) ranged from 0.864 to 0.996. Simi-
lar to intraobserver findings, interobserver ­ICC2,1 analy-
sis of JLCA was slightly out of line and showed fair results
(0.488). It is also shown that the intraobserver ­ICC2,1 and
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 6 of 12

Table 2 Lower limb measurement parameters (N = 144)

Measures Observer one Observer two p Value
Mean ± SD (range) Median (25 – 75 Mean ± SD (range) Median (25 – 75
percentiles) percentiles)

MFA (°) − 2 (− 5 to 0) − 3 (− 5 to − 1) p < 0.001

MAD (mm) − 8.69 ± 11.17 -9.01 ± 10.60 p = 0.013
(− 38 to 22) (− 37 to 22)
mLPFA (°) 88 (85–91) 88 (85–92) p = 0.173
mLDFA (°) 86 (84–89) 86 (84–89) p < 0.001
mMPTA (°) 90 (89–92) 90 (89–91) p < 0.001
mLDTA (°) 87 (84–89) 87 (85–89) p < 0.001
JLCA (°) 2 (1–2) 1 (1–2) p < 0.001
length of femur (mm) 474.24 ± 36.58 (387–572) 474.20 ± 36.28 (390–570) p = 0.067
length of tibia (mm) 379.5 (359–402) 378 (358–401) p < 0.001
Table 2 shows the mean measures from both analyses of both observers (SD = standard deviation) and the median with 25–75 percentiles
MFA Mechanical Femorotibial Angle; MAD Mechanical Axis Deviation; negative values of MFA and MAD show valgus malalignment; mLPFA mechanical Lateral
Proximal Femur Angle; mLDFA mechanical Lateral Distal Femur Angle; mMPTA mechanical Medial Proximal Tibia Angle; mLDTA mechanical Lateral Distal Tibia Angle;
JLCA Joint Line Convergence Angle, length of the femur and tibia. Parametric values: MAD and femur length; nonparametric values: MFA, mLPFA, mLDFA, mMPTA,
mLDTA, JLCA and tibia length.

Table 3 Intraobserver reliability: Intraclass correlation coefficient (SD ± 11.17 mm) for observer one and −9.01 mm
­ICC2,1 (95% confidence interval in parenthesis) (SD ± 10.60 mm) for observer two. The other mean
Measures Observer one Observer two values for observer one and observer two are shown in
Intraobserver-ICC2,1 Intraobserver-ICC2,1 Table 2.
The mLDFA, mMPTA, mLDTA, JLCA, MFA, MAD
MFA (°) 0.974 (0.963–0.981) 0.982 (0.972–0.989) and tibia length measurements of observer one and
MAD (mm) 0.994 (0.992–0.996) 0.995 (0.993–0.996) observer two were significant different (p < 0.05).
mLPFA (°) 0.977 (0.967–0.984) 0.961 (0.946–0.972) No significant difference between both observers
mLDFA (°) 0.972 (0.962–0.980) 0.981 (0.974–0.987) was found for femur length (p = 0.067) and mLPFA
mMPTA (°) 0.889 (0.845–0.920) 0.949 (0.929–0.963) (p = 0.173).
mLDTA (°) 0.849 (0.796–0.889) 0.902 (0.866–0.929) To assess the observer-dependent errors, the S ­ EM95%
JLCA (°) 0.607 (0.492–0.701) 0.676 (0.612–0.757) was calculated (Table 4). The ­SEM95% of both observ-
Femur length (mm) 0.997 (0.996–0.998) 0.999 (0.999–0.999) ers was found to be ± 1.39° for MFA, ± 3.31 mm for
Tibia length (mm) 0.997 (0.996–0.998) 0.996 (0.994–0.997) MAD, ± 1.69° for mLPFA, ± 1.06° for mLDFA, ± 1.29°
Table 3 shows the intraobserver-ICC2,1 from both observers for mMPTA, ± 2.17° for mLDTA and ± 1.38° for JLCA
MFA Mechanical Femorotibial Angle, MAD Mechanical Axis Deviation, mLPFA with a 95% confidence interval. In order to determine
mechanical Lateral Proximal Femur Angle, mLDFA mechanical Lateral Distal
Femur Angle, mMPTA mechanical Medial Proximal Tibia Angle, mLDTA
for which parameter the ­ SEM95% is proportionately
mechanical Lateral Distal Tibia Angle, JLCA Joint Line Convergence Angle, length lowest/highest, we divided the range of the respective
of the femur and tibia. ­SEM95% ­(SEM95%Range) by the range of the different val-
ues considered acceptable (accepted range). The results
(proportional error) are shown in Table 4. The pro-
interobserver ­ICC2,1 are slightly lower for the joint sur- portional error of MAD was slightly higher, but com-
face angles at the tibia than for the joint surface angles at parable to the proportional error of MFA (47.29% vs.
the femur. 46.33%). Looking at the relevant joint surface angles at
the knee joint, we noticed that the proportional errors
for the femur (mLPFA 33.80% and mLDFA 42.40%) are
Which parameter is most susceptible
lower than for the tibia (mMPTA 51.60% and mLDTA
to observer‑dependent errors in determining lower limb
72.33%). JLCA shows the highest proportional error
alignment?
of 138%. Consistent to the results from the intraob-
As shown in Table 2, the mean MFA was −2.40°
server- and interobserver ­ICC2,1, the SEMs and the pro-
(SD ± 3.13°) for observer one and −2.81° (SD ± 3.12°)
portional errors are also larger for the articular surface
for observer two. The MAD values averaged -8.69 mm
angles at the tibia than at the femur.
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 7 of 12

Table 4 Standard values and the accepted range of each lower limb parameter [12, 18], interobserver-ICC2,1, Standard Error of
Measurement (SEM) and proportional error
Measures Standard value Accepted range Interobserver-ICC2,1 SEM95% SEM95%Range Proportional
error in %

MFA (°) 1.3 ± 3 6 0.949 (0.859–0.975) ± 1.39 2.78 46.33

MAD (mm) 8±7 14 0.976 (0.966–0.983) ± 3.31 6.62 47.29
mLPFA (°) 90 ± 5 10 0.974 (0.964–0.981) ± 1.69 3.38 33.80
mLDFA (°) 88 ± 2.5 5 0.969 (0.954–0.979) ± 1.06 2.12 42.40
mMPTA (°) 87 ± 2.5 5 0.864 (0.816–0.901) ± 1.29 2.58 51.60
mLDTA (°) 89 ± 3 6 0.876 (0.831–0.909) ± 2.17 4.34 72.33
JLCA (°) 1±1 2 0.488 (0.350–0.605) ± 1.38 2.76 138.00
Femur length (mm) N/A N/A 0.996 (0.994–0.997) ± 4.53 N/A N/A
Tibia length (mm) N/A N/A 0.994 (0.990–0.998) ± 3.12 N/A N/A
Interobserver-ICC2,1 (95% confidence interval in parenthesis) with Standard Error of Measurement (SEM) with a 95% confidence interval. For interobserver-ICC2,1 first
measurement results of observer 1 were compared with first measurement results of observer 2
MFA Mechanical Femorotibial Angle; MAD Mechanical Axis Deviation; mLPFA mechanical Lateral Proximal Femur Angle; mLDFA mechanical Lateral Distal Femur Angle;
mMPTA mechanical Medial Proximal Tibia Angle; mLDTA mechanical Lateral Distal Tibia Angle; JLCA Joint Line Convergence Angle, length of the femur and tibia
For values of femur length and tibia length there are no standard values and therefore no accepted ranges, S­ EM95%Range or proportional errors are not applicable (N/A)

What is the SEM95% of the determination of the absolute present study the intra- and interobserver reliability for
femoral and tibial length, which are needed almost all lower limb alignment measurements in full-
for the estimation of the existing correction potential length standing anteroposterior radiographs of the fron-
for each bone? tal plane were found to be good to excellent.
The mean femur length was 474.24 mm (SD ± 36.58 mm) Our study is the first that investigates all relevant mala-
for observer one and 474.20 mm (SD ± 36.28 mm) for lignment parameters in the frontal plane in children and
observer two. The mean tibia length was 379.80 mm adolescents regarding intra- and interobserver reliability,
(SD ± 29.85 mm) for observer one and 378.38 mm ­SEM95% and proportional errors in order to make the best
(SD ± 29.67 mm) for observer two. The statistical analy- possible and the least error-prone decision on the indica-
sis did detect significant differences between the meas- tion and timing of THE, on postoperative follow-up and
urements of those two observers for the tibia length on the right time for removal of the tension band plates
(p < 0.001), but not for the femur length (p = 0.067) after completion of growth guidance.
(Table 2). The S ­EM95% for the absolute values of the According to the study of Specogna et al. [34], reliabil-
femoral and tibial length were ± 4.53 mm (femur) ity of repeated lower limb frontal plane alignment meas-
and ± 3.12 mm (tibia) (Table 4). If the S ­EM95% of the ures is high for planning a high tibial osteotomy with
respective bone is related to its absolute length, it can OA affecting the medial compartment of the knee due
be shown that the ­SEM95% of ± 4.53 mm for the femur is to varus malalignment (MFA Intra-ICC2,1 = 0.98; 95%
about 1.92% of the length of the femur, and ± 3.12 mm for CI = 0.97 − 0.99, Inter-ICC2,1 = 0.98; 95% CI = 0.91–0.99;
the tibia is about 1.64% of the length of the tibia in this MAD Intra-ICC2,1 = 0.98; 95% CI = 0.97 − 0.99, Inter-
collective. ICC2,1 = 0.97; 95% CI = 0.90–0.99). Specogna et al. [34]
elucidated that the estimates of error for the measure-
Discussion ments of MFA need to be considered with ± 1.50° and
The digital measurement of biometric parameters as well for the MAD ± 4.3 mm. These findings are largely con-
as the preoperative planning of orthopedic interventions sistent with the results of our investigation. In contrast
was shown to have good to excellent reliability, e.g., for to our findings, the mentioned study investigated MFA
Cobb angle in scoliosis [26, 27], planning of hip prosthe- and MAD on radiographs collected from adult patients
ses [28, 29] or leg axis determinations in long-leg X-rays (mean age 44 (21–65) years). When analyzing leg axis
[30–33]. The correct and reliable analysis of static param- on skeletally immature patients, Gordon et al. [31] were
eters of the lower extremity is essential for an adequate able to show that intra- and interobserver reliabilities
treatment of pediatric orthopedic diseases. In addition for each of the measurements (mLDFA, mMPTA and
to the clinical appearance, these measurements play a MAD) were ≥ 0.90 (0.90–0.99) in children (mean age
significant role in determining the indication for guided 11.2 (7.1–14.7) years) with neutral alignment and both
growth intervention of the lower limb using THE. In the varus or valgus malalignment regardless of the level of
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 8 of 12

observer experience. Gordon et al. [31] pointed out that axis in children and adolescents show a high physiologi-
the overall mean interobserver differences were ± 1.4° cal range in younger patients and a typical course. While
for the mLDTA, ± 1.6° for the mMPTA and ± 3.1 mm children under 2 years of age show a physiological varus
for the MAD measurement. In a similar study, Schmale alignment, a valgus leg axis develops until about 7 years
et al. [32] have shown that the intraobserver ICC val- of age. With advancing skeletal maturation (about 8 to
ues ranged from 0.27 to 0.94 for the mLDFA and from 10 years of age), the range decreases significantly, which
0.88 to 0.97 for the mMPTA in patients with open dis- is also reflected in a decrease in the spontaneous correc-
tal femur and proximal tibia physes at the time of a tion of the existing leg axis deformities [36]. The results
transphyseal anterior cruciate ligament reconstruction. of this study refer to children aged 10 years and older, in
Furthermore, the reliability (interobserver ICC: MAD whom the spontaneous correction potential is already
0.92, mLDFA 0.86 and mMPTA 0.98) of leg axis meas- reduced. Radtke et al. [8] indicate a THE of the knee at
urements were rated as good to excellent. Nowicki et al. a MAD of > 10 mm, regardless of whether the deviation
[33] evaluated the MAD, mLDFA, mMPTA and JLCA of is medial or lateral. They were not using the MFA for
pediatric lower extremities (mean age 13.5 years) with indication. The same applies to other studies. Gupta et al.
neutral, varus or valgus (mal)alignment but not the MFA [37] considered a MAD of 3 mm both medial and lateral
and they did not calculate the S ­ EM95% for all relevant as a normal value and described a deviation of the MFA
parameters. They could also show comparable values for lateral to the lateral intercondyloid tubercle as valgus
the intraobserver ICC of the four measurement param- malalignment and a deviation medial to the medial inter-
eters: 0.756–0.990 for mLDFA, 0.489–0.958 for mMPTA, condyloid tubercle as varus malalignment. In contrast,
0.831–0.927 for JLCA and 0.974–0.993 for MAD. Inter- Stevens et al. [38] divide the knee joint in an anteropos-
observer ICC values of 0.732–0.977 could be presented terior radiograph into three zones lateral and three zones
for the respective parameters. Feldman et al. [35] came medial, mirrored at the knee joint center. Here, neither
to a similar conclusion that planning guiding treatment the MFA nor the MAD is directly considered. They see
on those digital radiograph measurements is reliable and an indication for a THE in the case of a deviation of the
reproducible with intraobserver ICC of 0.77 (0.65–0.86) mechanical bearing line in zones 2 and 3 [38].
for mMPTA and mLDTA 0.80 (0.77–0.84). In this study, Regarding the results of the present study, we can pos-
however, only the relevant joint surface angles of the tibia tulate that the ­SEM95% of MFA, MAD, mLPFA, mLDFA,
were comparable to our finding, because they chose to mMPTA and mLDTA are likewise high. Thus, the
measure and analyze the anatomical angles instead of the results suggest that if a MFA measurement of 3° valgus
mechanical angles. Furthermore, MAD and MFA were is assumed, an S ­ EM95% of ± 1.39° means that the actual
not investigated as well. Table 5 shows a summary of all value will range between 1.61° (non-pathologic) and 4.39°
the studies just mentioned. This is to illustrate the inves- (pathologic) valgus. In addition, a MAD measurement of
tigated parameters and the respective differences in our −10 mm lateral to the center of the knee with a ­SEM95%
manuscript. of ± 3.31 mm means that the actual value ranges between
When looking at the results of the present study, it −6.69 mm (non-pathologic) and -13.31 mm (pathologic)
becomes obvious that nearly all ICCs represent good to valgus. This susceptibility to error must be taken into
excellent correlations. Only the intra- and interobserver account not only for leg alignment parameters but also
ICCs for JLCA are out of line and provide solely poor for joint surface angles when interpreting malalignment
correlations. At a closer look at the standard values of analysis and must be considered when planning a surgi-
the JLCA (1 ± 1°), it becomes apparent that a planning cal intervention. An inaccurate and wrong interpreted
discrepancy of 1.38° in two successive examinations of malalignment analysis may result in an incorrect diagno-
two different observers of the same X-ray results in a sis, which in turn leads to an unnecessary indication for
relatively large discrepancy (proportional error = 138%; surgical therapy. Furthermore, it could lead to termina-
Table 4). This might explain why even small measure- tion of the guided growth intervention by a removal of
ment inaccuracies become noticeable in a relatively the tension band plates at the wrong point of time. Con-
smaller ICC. In accordance with objective 1, it can be sidering the ­SEM95% of the MAD and the MFA, a wrong
concluded that the reliability of the individual parameters interpretation could also be an explanation for the high
is good, except of the JLCA. rebound rates of up to nearly 50% [39]. Therefore, it
Furthermore, it was investigated which parameter is must be discussed whether an overcorrection of the leg
most susceptible to observer-dependent errors in deter- axis with a minimum of 1.39° MFA and 3.31 mm MAD
mining lower limb alignment in the frontal plane. To should be aimed to prevent the common phenomenon of
date, there are no clear criteria for the indication of THE. rebound by taking the respective S ­ EM95% into account.
The growth of the child and the development of the leg This slight overcorrection is already recommended by
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 9 of 12

Table 5 Summary of the results of the leg axis and joint angle analyses of different studies investigating intra- and interobserver
reliability and Standard Error of Measurements of lower limb radiographs
Parameters Specogna et al.[34] Gordon et al.[31] Schmale et al.[32] Nowicki et al.[33] Feldman et al.[35]
N = 42 N = 56 N = 15 N = 32 N = 60

MFA
Intra ICC 0.98 (0.97–0.99) N/A N/A N/A N/A
Inter ICC 0.98 (0.91–0.99) N/A N/A N/A N/A
SEM 1.50 N/A N/A N/A N/A
MAD
Intra ICC 0.98 (0.97–0.99) 0.99 (0.99–0.99) 0.93–0.99 (0.84–1.00) 0.974–0.993 (0.946–0.996) N/A
Inter ICC 0.97 (0.90–0.99) 0.99 (0.99–0.99) 0.92 (0.81–0.97) 0.977 (0.964–0.987) N/A
SEM 4.3 3.1 N/A N/A N/A
mMPTA
Intra ICC N/A 0.99 (0.99–0.99) 0.88–0.97 (0.73–0.99) 0.489–0.958 (0.183–0.979) 0.78 (0.65–0.86)
Inter ICC N/A 0.98 (0.97–0.98) 0.98 (0.94–0.97) 0.778 (0.670–0.869) 0.77
SEM N/A 1.6 N/A N/A N/A
mLDFA
Intra ICC N/A 0.96 (0.95–0.97) 0.27–0.94 (-0.16–0.97) 0.756–0.990 (0.545–0.995) 0.80 (0.77–0.84)
Inter ICC N/A 0.91 (0.90–0.92) 0.86 (0.68–0.95) 0.732 (0.423–0.877) 0.70
SEM N/A 1.4 N/A N/A N/A
mLPFA
Intra ICC N/A N/A N/A N/A N/A
Inter ICC N/A N/A N/A N/A N/A
SEM N/A N/A N/A N/A N/A
mLDFA
Intra ICC N/A N/A N/A N/A N/A
Inter ICC N/A N/A N/A N/A N/A
SEM N/A N/A N/A N/A N/A
JLCA
Intra ICC N/A N/A N/A 0.831–0.927 (0.659–0.964) N/A
Inter ICC N/A N/A N/A 0.812 (0.678–0.899) N/A
SEM N/A N/A N/A N/A N/A
Intra ICC Intraobserver reliability using the intraclass correlation coefficient; Inter ICC Interobserver reliability using the intraclass correlation coefficient; SEM Standard
Error of Measurement; MFA Mechanical Femorotibial Angle; MAD Mechanical Axis Deviation; mLPFA mechanical Lateral Proximal Femur Angle; mLDFA mechanical
Lateral Distal Femur Angle; mMPTA mechanical Medial Proximal Tibia Angle; mLDTA mechanical Lateral Distal Tibia Angle; JLCA Joint Line Convergence Angle

several authors [40, 41]. Thus, a precise malalignment In addition to the observer-dependent errors, the
measurement and the knowledge about the presented radiological acquisition technique also influences the
­SEM95% of the respective parameters is crucial for a cor- result of the respective angles [13, 42, 43]. When tak-
rect surgical or nonsurgical treatment. In addition, we ing the full-leg radiograph, the patellae of the extended
were able to show that by measuring the MAD and MFA, legs have to be aligned frontally (not the feet) so that the
the ­SEM95% are comparable (proportional error = 47.29% femoral condyles are positioned parallel to the frontal
vs. 46.33%, respectively). Hereby, we consider the MAD plane. Rotation of the leg at the time the image is taken
and MFA to be equally (im)precise parameters for the should be avoided, as should flexion of the knee joint. In
determination of an axial malalignment and initiation of the case of increased internal rotation of the knee joints,
a growth-guiding therapy. In order to decide on which the leg axis appears more valgus; in the case of external
bones to perform growth guidance, mLDFA and mMPTA rotation, it appears correspondingly more varus [13].
show distinct observer-dependent differences. The pro- In our study cohort, an internal quality assessment was
portional error for mLDFA is lower at 42.2% than for performed first by the executing radiology assistants
mMPTA at 51.60%. and subsequently by the two analyzing observers. All
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 10 of 12

included patients showed qualitatively unobjectionable susceptibility to observer-dependent errors. In addi-

radiographs. tion, due to no significant difference in determining the
Furthermore, ­SEM95% of tibial and femoral length was femoral length between two different observers and a
determined. Those values are essential for a proper esti- lower ­SEM95% for tibial length by two different observ-
mation of residual growth and correction potential using ers, both femoral and tibial length may be considered
the multiplier method according to Paley [15, 16, 44]. The precise when analyzing the residual growth and cor-
multiplier method need an exact determination of the rection potential for the respective bone, for instance,
bone length (either tibia or femur) to give a precise infor- with the multiplier method. This study shows good
mation about the correction potential. We could show to excellent intra- and interobserver ICCs for all leg
that the ICCs of the length determination were good and alignment parameters and joint surface angles, except
the ­SEM95% were relatively high with ± 4.53 mm for the for JLCA. Furthermore, according to the ­SEM95% and
femur and ± 3.12 mm for the tibia. Since the femur length proportional errors the determination of joint surface
shows no significant difference in the measurement of angels of the femur seems to be more precise than for
two different examiners, the measurement method for the tibia. In unclear or borderline cases with marginal
determining the center of the femoral head with the aid MAD and MFA, instrumented gait analysis and the
of a circle around the femoral head with the determina- determination of dynamic joint loading [4, 5, 11] should
tion of the most distal points of the two femoral condyles be included to determine whether surgical or nonsurgi-
does not seem to be examiner-dependent and could be cal treatment should be preferred.
seen as precise when analyzing the residual growth and
correction potential with the multiplier method. Due to
Abbreviations
the slightly lower S
­ EM95% in determining the tibial length, OA Osteoarthritis
which were 1.64% of the total tibia length and ± 1.96% of THE Temporary hemiepiphysiodesis
the total femur length, the tibial length can be considered SEM Standard Error of Measurement
MAD Mechanical Axis Deviation
as precise as well. Again, it is important to know the error MFA Mechanical Femorotibial Angle
and include it in the analysis in predicting growth poten- JLCA Joint Line Convergence Angle
tial for each bone. mLPFA Mechanical Lateral Proximal Femur Angle
mLDFA Mechanical Lateral Distal Femur Angle
One limitation of this study is that it did not aim to mMPTA Mechanical Medial Proximal Tibia Angle
identify differences in the experience of the two exam- mLDTA Mechanical Lateral Distal Tibia Angle
iners. However, Vaishya et al. [45] showed that more ICC Intraclass correlation coefficient
SD Standard deviation
experienced examiners tend to perform more precise
measurements. Other studies (e.g., Gordon et al. [31]) did Acknowledgements
not find any experience-dependent differences in intra- Not applicable.
or interobserver reliability with excellent agreement for Author contributions
all observers. In accordance with Vaishya et al. [45], we SB, MB, JH, FS and AM were involved in the conception; SB, MB, and FS
consider the measured values of this study to be highly designed the work; SB and MB contributed to the acquisition and analysis; SB,
MB, JH and FS assisted in the acquisition, analysis; SB and MB contributed to
reliable and reproducible and may not be improved with writing—original draft preparation; SB, MB, JH, FS and AM helped in writing—
further experience of the observers. review and editing. All authors read and approved the final manuscript.
In conclusion, a precise malalignment measurement
Funding
and the knowledge about the presented ­SEM95% of the Open Access funding enabled and organized by Projekt DEAL. The institution
respective parameters are crucial for a correct surgi- of one or more of the authors (SB, MB, JH, FS, AM) has received funding from
cal or nonsurgical treatment. The susceptibility to error the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation).
Project number: 403837822. This study was registered with DRKS (German
must be included when interpreting malalignment Clinical Trials Register) under the number DRKS00015053. The funders had no
analysis and must be considered when planning a sur- role in the design of the study, in the collection, analyses, or interpretation of
gical intervention. The ICC and S ­ EM95% values in this data, in the writing of the manuscript or in the decision to publish the results.
study indicate how reproducible, reliable and precise Availability of data and materials
radiological assessment of variables are obtained from The datasets used and/or analyzed during the current study are available from
an X-ray. In general, with any individual measurement, the corresponding author on reasonable request.
it should be understood that a measurement can never
be determined exactly but is always subject to a cer- Declarations
tain measurement error or measurement inaccuracy. Ethics approval and consent to participate
The same applies to anatomical standard values, which This study was performed in line with the principles of the Declaration of
also have their physiological range. The results of this Helsinki. Approval was granted by the Ethics Committee of Goethe University
Frankfurt (protocol code: 182/16). Informed consent was obtained from all
study elucidate that MAD and MFA show no different individual participants included in the study.
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 11 of 12

Consent for publication 15. Paley D, Bhave A, Herzenberg JE, Bowen JR. Multiplier method for predict-
Not applicable. ing limb-length discrepancy. J Bone Joint Surg Am. 2000;82(10):1432–46.
https://​doi.​org/​10.​2106/​00004​623-​20001​0000-​00010.
Competing interests 16. Birch JG, Makarov MR, Sanders JO, Podeszwa DA, Honcharuk EM, Esparza
The authors declare that they have no competing interests. M, et al. Lower-extremity segment-length prediction accuracy of the
Sanders multiplier, Paley multiplier, and White-Menelaus formula. J Bone
Author details Joint Surg Am. 2021;103(18):1713–7. https://​doi.​org/​10.​2106/​jbjs.​20.​
1
Department of Orthopedics (Friedrichsheim), University Hospital Frankfurt, 01923.
Goethe University, 60528 Frankfurt/Main, Germany. 2 Dr. Rolf M. Schwiete 17. Sanders JO, Howell J, Qiu X. Comparison of the Paley method using
Research Unit for Osteoarthritis, Department of Orthopedics (Friedrichsheim), chronological age with use of skeletal maturity for predicting mature
University Hospital Frankfurt, Goethe University, Frankfurt/Main, Germany. limb length in children. J Bone Joint Surg Am. 2011;93(11):1051–6.
3
Medical Park St. Hubertus Klinik, Bad Wiessee, Germany. 4 Department https://​doi.​org/​10.​2106/​jbjs.J.​00384.
of Sports and Exercise Science, University of Salzburg, 5020 Salzburg, Austria. 18. Paley D. Normal Lower Limb Alignment and Joint Orientation. Principles
of Deformity Correction. Berlin, Heidelberg: Springer; 2002.
Received: 7 June 2022 Accepted: 19 January 2023 19. Hunt MA, Fowler PJ, Birmingham TB, Jenkyn TR, Giffin JR. Foot rotational
effects on radiographic measures of lower limb alignment. Can J Surg.
2006;49(6):401–6.
20. Jamali AA, Meehan JP, Moroski NM, Anderson MJ, Lamba R, Parise C. Do
small changes in rotation affect measurements of lower extremity limb
References alignment? J Orthop Surg Res. 2017;12(1):77. https://​doi.​org/​10.​1186/​
1. Sharma L. Osteoarthritis of the Knee. N Engl J Med. 2021;384(1):51–9. s13018-​017-​0571-6.
https://​doi.​org/​10.​1056/​NEJMc​p1903​768. 21. Moreland JR, Bassett LW, Hanker GJ. Radiographic analysis of the axial
2. Tanamas S, Hanna FS, Cicuttini FM, Wluka AE, Berry P, Urquhart DM. Does alignment of the lower extremity. J Bone Joint Surg Am. 1987;69(5):745–9.
knee malalignment increase the risk of development and progression of 22. Reina-Bueno M, Lafuente-Sotillos G, Castillo-Lopez JM, Gomez-Aguilar E,
knee osteoarthritis? A systematic review Arthritis Rheum. 2009;61(4):459– Munuera-Martinez PV. Radiographic assessment of lower-limb discrep-
67. https://​doi.​org/​10.​1002/​art.​24336. ancy. J Am Podiatr Med Assoc. 2017;107(5):393–8. https://​doi.​org/​10.​
3. Farr S, Kranzl A, Pablik E, Kaipel M, Ganger R. Functional and radiographic 7547/​15-​204.
consideration of lower limb malalignment in children and adolescents 23. Streiner D. Health measurement scales: a practical guide to their develop-
with idiopathic genu valgum. J Orthop Res. 2014;32(10):1362–70. https://​ ment and use. Aust N Z J Public Health. 2016;40(3):294–5. https://​doi.​org/​
doi.​org/​10.​1002/​jor.​22684. 10.​1111/​1753-​6405.​12484.
4. Böhm H, Stief F, Sander K, Hösl M, Döderlein L. Correction of static axial 24. Gamble P, de Beer J, Petruccelli D, Winemaker M. The accuracy of
alignment in children with knee varus or valgus deformities through digital templating in uncemented total hip arthroplasty. J Arthroplasty.
guided growth: Does it also correct dynamic frontal plane moments 2010;25(4):529–32. https://​doi.​org/​10.​1016/j.​arth.​2009.​04.​011.
during walking? Gait Posture. 2015;42(3):394–7. https://​doi.​org/​10.​1016/j.​ 25. Weir JP. Quantifying test-retest reliability using the intraclass correlation
gaitp​ost.​2015.​06.​186. coefficient and the SEM. J Strength Cond Res. 2005;19(1):231–40. https://​
5. Holder J, Feja Z, van Drongelen S, Adolf S, Böhm H, Meurer A, et al. Effect doi.​org/​10.​1519/​15184.1.
of guided growth intervention on static leg alignment and dynamic knee 26. Mok JM, Berven SH, Diab M, Hackbarth M, Hu SS, Deviren V. Comparison
contact forces during gait. Gait Posture. 2020;78:80–8. https://​doi.​org/​10.​ of observer variation in conventional and three digital radiographic
1016/j.​gaitp​ost.​2020.​03.​012. methods used in the evaluation of patients with adolescent idiopathic
6. Stief F, Böhm H, Schwirtz A, Dussa CU, Döderlein L. Dynamic loading of scoliosis. Spine. 2008;33(6):681–6. https://​doi.​org/​10.​1097/​BRS.​0b013​
the knee and hip joint and compensatory strategies in children and ado- e3181​66aa8d.
lescents with varus malalignment. Gait Posture. 2011;33(3):490–5. https://​ 27. Wills BP, Auerbach JD, Zhu X, Caird MS, Horn BD, Flynn JM, et al.
doi.​org/​10.​1016/j.​gaitp​ost.​2011.​01.​001. Comparison of Cobb angle measurement of scoliosis radiographs with
7. Vogt B, Schiedel F, Rödl R. Guided growth in children and adolescents. preselected end vertebrae: traditional versus digital acquisition. Spine.
Correction of leg length discrepancies and leg axis deformities. Ortho- 2007;32(1):98–105. https://​doi.​org/​10.​1097/​01.​brs.​00002​51086.​84420.​d1.
pade. 2014;43(3):267–84. https://​doi.​org/​10.​1007/​s00132-​014-​2270-x. 28. Shaarani SR, McHugh G, Collins DA. Accuracy of digital preoperative tem-
8. Radtke K, Goede F, Schweidtmann K, Schwamberger T, Calliess T, Fregien plating in 100 consecutive uncemented total hip arthroplasties: a single
B, et al. Temporary hemiepiphysiodesis for correcting idiopathic and surgeon series. J Arthroplasty. 2013;28(2):331–7. https://​doi.​org/​10.​1016/j.​
pathologic deformities of the knee: a retrospective analysis of 355 cases. arth.​2012.​06.​009.
Knee. 2020;27(3):723–30. https://​doi.​org/​10.​1016/j.​knee.​2020.​04.​024. 29. Brenneis M, Braun S, van Drongelen S, Fey B, Tarhan T, Stief F, et al.
9. Hosseinzadeh P, Ross DR, Walker JL, Talwalkar VR, Iwinski HJ, Milbrandt TA. Accuracy of preoperative templating in total hip arthroplasty with special
Three methods of guided growth for pediatric lower extremity angular focus on stem morphology: a randomized comparison between com-
deformity correction. Iowa Orthop J. 2016;36:123–7. mon digital and three-dimensional planning using biplanar radiographs.
10. Stief F, Feja Z, Holder J, van Drongelen S, Adolf S, Braun S, et al. Non- J Arthroplasty. 2021;36(3):1149–55. https://​doi.​org/​10.​1016/j.​arth.​2020.​10.​
invasive determination of frontal plane lower limb alignment using 016.
motion capture technique—An alternative for full-length radiographs in 30. Sled EA, Sheehy LM, Felson DT, Costigan PA, Lam M, Cooke TD. Reliability
young patients treated by a temporary hemiepiphysiodesis? Gait Posture. of lower limb alignment measures using an established landmark-based
2020;79:26–32. https://​doi.​org/​10.​1016/j.​gaitp​ost.​2020.​04.​011. method with a customized computer software program. Rheumatol Int.
11. Stief F, Holder J, Böhm H, Meurer A. Dynamic analysis of joint loading 2011;31(1):71–7. https://​doi.​org/​10.​1007/​s00296-​009-​1236-5.
due to leg axis deformity in the frontal plane: relevance of instrumented 31. Gordon JE, Chen RC, Dobbs MB, Luhmann SJ, Rich MM, Schoenecker PL.
gait analysis. Orthopade. 2021;50(7):528–37. https://​doi.​org/​10.​1007/​ Interobserver and intraobserver reliability in the evaluation of mechani-
s00132-​021-​04121-9. cal axis deviation. J Pediatr Orthop. 2009;29(3):281–4. https://​doi.​org/​10.​
12. Paley D, Pfeil J. Principles of deformity correction around the knee. Ortho- 1097/​BPO.​0b013​e3181​9b9188.
pade. 2000;29(1):18–38. https://​doi.​org/​10.​1007/​s0013​20050​004. 32. Schmale GA, Bayomy AF, O’Brien AO, Bompadre V. The reliability of full-
13. Radtke K, Gómez Dammeier B, Braun S. Radiographic analysis of limb length lower limb radiographic alignment measurements in skeletally
malalignment in the frontal plane. Orthopade. 2021;50(7):520–7. https://​ immature youth. J Child Orthop. 2019;13(1):67–72. https://​doi.​org/​10.​
doi.​org/​10.​1007/​s00132-​021-​04125-5. 1302/​1863-​2548.​13.​180087.
14. Schnurr C, König DP. Achsfehlstellungen Knie. Orthopädie und Unfallchir- 33. Nowicki PD, Vanderhave KL, Farley FA, Kuhns LR, Dahl W, Caird MS. Reli-
urgie Essentials. 2nd ed. Stuttgart: Georg Thieme; 2013. ability of digital radiographs for pediatric lower extremity alignment. J
Pediatr Orthop. 2012;32(6):631–5. https://​doi.​org/​10.​1097/​BPO.​0b013​
e3182​694e07.
Braun et al. Journal of Orthopaedic Surgery and Research (2023) 18:69 Page 12 of 12

34. Specogna AV, Birmingham TB, DaSilva JJ, Milner JS, Kerr J, Hunt MA, et al.
Reliability of lower limb frontal plane alignment measurements using
plain radiographs and digitized images. J Knee Surg. 2004;17(4):203–10.
https://​doi.​org/​10.​1055/s-​0030-​12482​22.
35. Feldman DS, Henderson ER, Levine HB, Schrank PL, Koval KJ, Patel RJ, et al.
Interobserver and intraobserver reliability in lower-limb deformity correc-
tion measurements. J Pediatr Orthop. 2007;27(2):204–8. https://​doi.​org/​
10.​1097/​01.​bpb.​00002​42441.​96434.​6f.
36. Ehnert M, Stief F, Meurer A. Axial deformities of the lower extremity in
the frontal plane: from physiological development to pathological and
possible long-term consequences. Orthopade. 2021;50(7):511–9. https://​
doi.​org/​10.​1007/​s00132-​021-​04124-6.
37. Gupta P, Gupta V, Patil B, Verma V. Angular deformities of lower limb in
children: Correction for whom, when and how? J Clin Orthop Trauma.
2020;11(2):196–201. https://​doi.​org/​10.​1016/j.​jcot.​2020.​01.​008.
38. Stevens PM, Maguire M, Dales MD, Robins AJ. Physeal stapling for idi-
opathic genu valgum. J Pediatr Orthop. 1999;19(5):645–9.
39. Stief F, Holder J, Böhm H, Meurer A. Prevalence and predictors of
rebound deformity in the frontal plane: a literature review. Orthopade.
2021;50(7):548–58. https://​doi.​org/​10.​1007/​s00132-​021-​04118-4.
40. Saran N, Rathjen KE. Guided growth for the correction of pediatric lower
limb angular deformity. J Am Acad Orthop Surg. 2010;18(9):528–36.
https://​doi.​org/​10.​5435/​00124​635-​20100​9000-​00004.
41. Stevens PM. Guided growth for angular correction: a preliminary series
using a tension band plate. J Pediatr Orthop. 2007;27(3):253–9. https://​
doi.​org/​10.​1097/​BPO.​0b013​e3180​3433a1.
42. Marques Luís N, Varatojo R. Radiological assessment of lower limb align-
ment. EFORT Open Rev. 2021;6(6):487–94. https://​doi.​org/​10.​1302/​2058-​
5241.6.​210015.
43. Radtke K, Becher C, Noll Y, Ostermeier S. Effect of limb rotation on radio-
graphic alignment in total knee arthroplasties. Arch Orthop Trauma Surg.
2010;130(4):451–7. https://​doi.​org/​10.​1007/​s00402-​009-​0999-1.
44. Whitaker AT, Vuillermin C. Lower extremity growth and deformity. Curr
Rev Musculoskelet Med. 2016;9(4):454–61. https://​doi.​org/​10.​1007/​
s12178-​016-​9373-4.
45. Vaishya R, Vijay V, Birla VP, Agarwal AK. Inter-observer variability and its
correlation to experience in measurement of lower limb mechanical axis
on long leg radiographs. J Clin Orthop Trauma. 2016;7(4):260–4. https://​
doi.​org/​10.​1016/j.​jcot.​2016.​05.​010.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions