Int. J. Oral Maxillofac. Surg.

2008; 37: 10891093

doi:10.1016/j.ijom.2008.06.011, available online at http://www.sciencedirect.com

Clinical Paper
Orthognathic Surgery

How accurate is model planning

for orthognathic surgery?

A. Sharifi, R. Jones, A. Ayoub,

K. Moos, F. Walker, B. Khambay,
S. McHugh
Glasgow Dental Hospital & School, Glasgow
University, 378 Sauchiehall Street, Glasgow,
G2 3JZ, UK

A. Sharifi, R. Jones, A. Ayoub, K. Moos, F. Walker, B. Khambay, S. McHugh: How

accurate is model planning for orthognathic surgery?. Int. J. Oral Maxillofac. Surg.
2008; 37: 10891093. # 2008 International Association of Oral and Maxillofacial
Surgeons. Published by Elsevier Ltd. All rights reserved.
Abstract. The purpose of this study was to evaluate the accuracy of model surgery
prediction after orthognathic surgery and to identify possible errors associated with
the prediction process. The study included 46 patients who had undergone
orthognathic surgical procedures; 22 in Group A who had had a Le Fort I osteotomy;
and 24 in Group B who had had a Le Fort I osteotomy and mandibular setback
surgery. The immediate postoperative and preoperative lateral cephalograms were
analysed to calculate surgical changes; these were compared with those obtained
from model surgery prediction and a statistical analysis was undertaken. The
maxilla was more under-advanced and over-impacted anteriorly than predicted by
model surgery. The amount of mandibular setback was more than that predicted by
model surgery. None of the differences between prediction planning and actual
surgical changes was statistically significant at p < 0.05. Inaccuracy with the face
bow recording, the intermediate wafer, and auto-rotation of the mandible in the
supine or anaesthetized patient would appear to be the principal reasons for errors.
Inaccuracies are associated with the transfer of prediction planning to model surgery
planning and prediction, which should be eliminated to improve the accuracy and
predictability of orthognathic surgery.

Current methods of orthognathic surgical

planning involve clinical evaluation,
photographs, freehand surgical simulation based on cephalometric tracing
and then transferred to study model surgery, and computerized prediction software.6,13
Model surgery planning on dental casts
is used for the final correction of facial
deformity and malocclusion. Analysis of
the model surgery allows the transfer of
planned three-dimensional movements for
the surgical correction of complex dentofacial deformities.8
0901-5027/1201089 + 05 $30.00/0

Model surgery depends on the accuracy

of the recording of the dental occlusion in
the retruded position and the face bow
transfer to the articulator. These recordings both have inherent inaccuracies.
BAILY and NOWLIN1 measured the angulation of the occlusal plane to the Frankfort
horizontal on the Hanau articulator and
compared this with lateral cephalograms;
they found a mean difference of 58, which
corresponded to a 70% error. The threedimensional accuracy of the position of
the upper first molar was highly variable
using four different Hanau facebows.16

Keywords: model surgery; orthognathic surgery; articulator; osteotomy.

Accepted for publication 13 June 2008
Available online 28 August 2008

One of the commonest errors in orthognathic model surgery occurs in mounting

the models on the articulator. The accuracy of the face bow transfer may differ
from one type of face bow to another. The
kinematic face bow transfer record is
accompanied by inherent errors in the
adjustment of the instrument to the
patients face and transfer to the articulator.4,11
Assessing the accuracy of orthognathic
surgery by comparing surgical changes
with model surgery is prudent. It is important to recognize why the pre-planned

# 2008 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.

1090

Sharifi et al.

model surgery is not always reflected in

the final outcome.

Materials and methods

The most common group of patients treated in the authors unit are those with
maxillary hypoplasia and those with Class
III skeletal deformity due to maxillary
deficiency and mandibular prognathism.
This investigation was carried out on 46
patients: 22 in Group A who had had a Le
Fort I maxillary advancement and impaction; and 24 in Group B who had had
bimaxillary
osteotomies,
maxillary
advancement and impaction with a mandibular setback. None of the cases had
segmental surgery or facial asymmetry.
In all cases the maxillary position was
transferred to the articulator using an average value condylar face bow (Dentatus
face bow). The upper and lower casts were
mounted on a Dentatus semi-adjustable
articulator.
The availability of preoperative and
immediate postoperative lateral cephalograms and casts was mandatory for this
investigation. The following landmarks
were digitized on all the pre- and postoperative radiographs: ANS (anterior
nasal spine), PNS (posterior nasal spine),
A point, UIE (upper incisor edge), UME
(upper molar edge), P (the most concave
point of the palate), B point, LIE (lower
incisor edge), LME (lower molar edge),
Pg (pogonion), Me (menton), Ge (genion:
the most posterior point on the lingual
cortical aspect of the symphysis above
the genial tubercle and usually 12 mm
below the level of the lower incisor apex).
The pre- and postoperative lateral
cephalograms were superimposed on the
best fit of the anterior cranial base8 using
Adobe Photoshop 7.0 (Adobe incorporated 2004 USA).
The Frankfort horizontal (FHP) was
chosen as the reference line for this study
(X axis). A perpendicular line to the FHP
was drawn from the centre of the sella (Y
axis). All pre- and postoperative landmarks were measured to extract X and
Y coordinates. The actual changes following surgery were measured at A and P
points for the Le Fort I osteotomy and at B
and Ge points for mandibular setback
surgery. These changes were compared
with the changes predicted by the model
surgery. Students t-test was used to identify the skeletal significance of the difference between the actual surgical changes
and those predicted. SPSS version 11.5
(Statistical Package for Social Science)
was used to analyse the data.

Fig. 1. Difference in advancement between the actual changes and prediction at point P
(maximal concavity of the palate) in Group A.

Error of methods

The reproducibility of the superimposition

of the lateral cephalographs was assessed.
The pre- and postoperative radiographs of
5 randomly selected cases were superimposed on the cranial base of the skull, and
the following 4 landmarks were placed:
upper incisor edge, lower incisor edge,
lower antero-inferior point of anterior
fixation plate, and the most postero-inferior point of the posterior plate. The procedure was repeated after 1 month. The
Dahlberg formula was applied to assess
the magnitude of errors.5,10
Results

The errors in the landmark locations

traced from the cephalogram were only
0.2 mm across the X axis and 0.3 mm
across the Y axis.
In Group A, in 50% of the cases, the
actual forward movement of the maxilla
was accurate to within 1 standard deviation
of the mean, with that of the prediction
model. In 36% (n = 8) the maxilla was
under-advanced and in 14% (n = 3) it
was over-advanced in comparison with
prediction planning (Fig. 1). Table 1 shows

the tendency toward under-advancement of

the maxilla.
Fig. 2 shows the accuracy of the prediction planning in Group A at point P.
Similar to the horizontal changes, about
50% (n = 11) of the cases had anterior
maxillary impaction similar to the prediction values (within 1 standard deviation of
the mean). In about 33% (n = 7) the maxilla was over-impacted, and it was underimpacted in the remainder (n = 4)
(Table 2). Posterior maxillary impaction
at PNS was similar to model surgery prediction in 64% (n = 14) (Table 2).
In Group B (n = 24), a similar pattern of
differences between the actual surgical
changes and the prediction movement
was also identified (Tables 3 and 4). Accurate maxillary advancement as predicted
was achieved in 46% of cases (n = 11),
with 37% (n = 9) showing under-advancement and 17% (n = 4) over-advancement
(Fig. 3). Anterior over-impaction was
noted in 33% (n = 8) and posterior
under-impaction in 25% (n = 6) of cases
(Figs. 4 and 5, Table 4).
There was no statistically significant
difference between the two groups regarding the magnitude of disparities of actual

Table 1. Differences between the actual changes and those predicted for movements in Group A
along the X co-ordinate; a negative sign means less advancement surgically than that predicted
by model surgery
ANS
PNS
A
UIE
UME
P point
M&M

Mean

SD

Median

1.16
0.20
0.20
0.14
0.32
0.86
0.34

1.98
2.06
2.21
2.34
2.36
2.33
2.15

0.80
0.47
0.10
0.25
0.21
1.03
0.21

IQ (25%:75%)
2.97:0.24
1.04:1.44
2.36:1.40
2.72:1.08
2.33:1.40
2.85:0.66
2.49:0.86

Abbreviations: A: Maxillary A Point; ANS: Anterior Nasal Spine; IQ: Interquartile; M&M:
Mean of all the maxillary landmarks; P: The maximum concavity of palate; PNS: Posterior
Nasal Spine; SD: Standard Deviation; UIE: Upper Incisor Edge; UME: Upper Molar Edge.

How accurate is model planning for orthognathic surgery?

Fig. 2. Difference in impaction between the actual changes and prediction at point P (maximal
concavity of the palate) in Group A.
Table 2. The differences in mm between the actual changes and those predicted for movements
in Group A along the Y axis; a negative sign means more impaction surgically than that
predicted by model surgery
ANS
A
UIE
P point

Mean

SD

Median

IQ (25%:75%)

0.29
0.37
0.50
0.94

2.38
2.97
2.13
2.01

0.28
0.83
1.13
1.16

1.96:1.35
1.69:1.87
2.42:1.10
2.5:0.22

Table 3. The differences between the actual changes and those predicted for maxillary
advancement in Group B along the X axis; a negative sign means less surgical advancement
than that predicted by model surgery
ANS
PNS
A
UIE
UME
P point
M&M

Mean

SD

Median

0.37
0.73
0.50
0.20
0.39
0.63
0.06

2.20
2.10
2.68
1.98
2.77
2.92
2.1

0.05
0.21
0.22
0.47
0.62
1.17
0.06

maxillary surgical movement and the predicted values from orthognathic models.
In Group B, 46% (n = 11) had an accurate
mandibular setback as predicted in the
study models. Over-correction was
detected in 37% (n = 9) and under-correction in 17% (n = 4) (Fig. 6). The mean
values showed a tendency towards setting
the mandible more posteriorly than predicted.
None of the differences detected in
Groups A and B between the actual surgical movements and the predictions from

IQ (25%:75%)
1.94:1.19
2.41:-0.21
1.79:2.63
1.13:1.44
1.87:2.38
2.68:1.16
1.67:1.07

study models was statistically significant

(p < 0.05). The difference at ANS was
close to statistical significance at
p = 0.051.
Discussion

In this study, the maxilla showed a tendency to under-advancement compared

with the predicted movement in about
33% of the cases in both groups. This
could be a result of the inaccuracy of
transferring the maxillary plane angle

Table 4. The differences between the actual changes and those predicted for maxillary
impaction in Group B along the Y axis; a negative value means over impaction
ANS
A
UIE
P point

Mean

SD

Median

IQ (25%:75%)

0.42
0.25
0.06
0.65

1.85
2.30
2.45
2.09

0.10
0.31
0.14
0.78

1.90:0.81
1.94:0.97
1.64:0.49
2.58:0.57

1091

and the inaccuracy of recording the mandibular position.

The semi-adjustable articulator used for
model surgery in this study was originally
created for prosthetic dentistry. Its face
bow was designed to transfer the relationship of the maxilla to the terminal hinge
axis of the mandible. To accomplish this,
the posterior end of the face bow is aligned
to the terminal hinge axis (middle of condyle), and the anterior end is aligned to the
orbitale. These points define a plane called
the axis-orbital plane,7,9 which is related
to the upper cross arm of the articulator to
mount the maxillary occlusal model.
The upper cross member of the articulator represents the Frankfort horizontal
plane, which may not be horizontal
(Fig. 6), as assumed in designing the existing semi-adjustable articulator. This could
cause errors in model planning because
most vertical measurements made during
model surgery are parallel to a reference
line, usually the upper arm of the articulator. When the maxilla is impacted in a
horizontal plane superiorly on the articulator, the magnitude of the upward movement is related to a horizontal reference
line. The extent of the impaction is different when the cast has been mounted
with an angle between the occlusal plane
and the upper member of the articulator,
which is different from the real angle
between the occlusal plane and the Frankfort horizontal in that patient, for which a
different amount of impaction would be
required.7 A more anterior maxillary
impaction would be needed surgically
than that predicted by model surgery to
compensate for the steepness of the maxillary plane angle on the articulator; this
was true in 33% of the cases in this study.
When using a conventional articulator
for orthognathic surgery, it is essential that
the angle between the occlusal plane and
the Frankfort horizontal for the patient is
the same as the angle between the occlusal
plane and the upper member of the articulator on the maxillary model. If this is
incorrect, the result of the model surgery
is erroneous.
The authors suggest that every individual lateral cephalogram should be
checked for the accuracy of the mounting
of the maxillary cast on the articulator.
Further research is needed to prove and
confirm the accuracy of this hypothesis.
GATENO et al.9 showed that if there were
a 128 difference between the Frankfort
horizontal plane and the axis-orbital plane,
with bimaxillary model surgery, the maxilla would move 15% less than desired for
maxillary advancement. In the present
study, in about 33% of cases, the maxilla

1092

Sharifi et al.

Fig. 3. Difference in advancement between the actual changes and predictions at point P
(maximal concavity of the palate) in Group B.

Fig. 4. Difference in impaction between the actual changes and predictions at point P (maximal
concavity of the palate) in Group B.

Fig. 5. Difference in impaction between the actual changes and predictions at PNS in Group B.

showed under-advancement from that predicted for both groups.

The other source of error is the difference in the patients mandibular position
when supine and upright; the mandible
tends to be positioned more posteriorly
when the patient is lying down and the
mouth has been actively closed into the
relaxed position of centric occlusion.12
BOUCHER and JACOBY3 reported that the
mandible was positioned up to 2 mm more
posteriorly in the anaesthetized patient. A
slight posterior displacement, with a vertical drop of the condyle (mean 2.4 mm)
under general anaesthesia was reported.15
BAMBER et al2 recommended recording the
centric relationship in the supine conscious position when planning bimaxillary
osteotomies. All the cases in the present
study had the wax bite and face bow taken
in the upright position and in centric
occlusion. This would have registered
the mandible in a more anterior position
than its location when the patient was
lying flat under general anaesthesia. Once
the maxilla was down-fractured at a Le
Fort I level, the mandible would be the
only reference point to adjust the maxillary position in an antero-posterior direction using the occlusal wafer. Therefore,
less maxillary advancement would be
achieved than predicted on the articulator
owing to the more posterior mandibular
position. The mandible has been overcorrected (more setback) to compensate
for maxillary under-advancement.
The other possible source of errors in
planning orthognathic surgery is inherent
in the nature of mandibular hinge axis and
the consequences of its application in
prediction planning. The true hinge axis
of mandibular rotation is the axis exactly
located as a result of the combined rotation
and translation of the condyles. It has been
shown that an insignificant gliding of the
condyles of 0.5 mm significantly displaces
the located hinge axis. The inaccuracy in
registering and transferring the true hinge
axis of the condyle to the articulator would
cause errors in predicting mandibular rotation secondary to maxillary surgery. Using
the current instrumentation, discrepancies
between the true hinge axis and the transferred axis for the mounted casts on semiadjustable articulators are inevitable.14
Intermediate wafers are another major
source of inaccuracy. After a Le Fort I
osteotomy and mobilization of the maxilla, the wafer places the maxilla in the
pre-planned position relating it to the
mandible for fixation. Under general
anaesthesia, the muscles of mastication
are relaxed and the mandible would
not serve as a fixed reference plane for

How accurate is model planning for orthognathic surgery?

1093

Fig. 6. Frankfort plane. (After DOWNS WB. Analysis of the dentofacial profile. Angle Orthod 1956: 26: 192212.)

maxillary surgery. In addition, the wafer

cannot be placed exactly in the preplanned position because there are no
anatomical references points outside the
osteotomy cuts against which a check can
be made. An instrument that would place
the intermediate wafer in its correct relation to the Frankfort horizontal or visual
axis plane should be designed. This
instrument should be able to transfer reliable anatomical references (e.g. external
auditory meatus and nasion) from patient
to articulator. If there were clear anatomical references, which act as external
frames of reference, these could guide the
wafer into the planned position more
accurately.
There were no statistically significant
differences between the predicted and
actual changes, mainly due to the small
sample size, however, this study has identified clear differences in some cases
between what was planned and achieved.
These discrepancies may produce an unacceptable occlusion, therefore every effort
should be made to address the disparity
between model planning and surgical
planning.
This article did not investigate the clinical significance of the mismatch between
model prediction and actual surgical
change, but it did prove its existence
and highlight ways of improving orthognathic model surgery. The use of a specially designed orthognathic articulator
that takes into consideration anatomical
variations, accurately locating the mandible in three dimensions and reproducing
its autorotation, should improve orthognathic model surgery.

Acknowledgements. The authors would

like to thank Mrs A Maguire and Ms A
McCormack for their help in compiling
this manuscript.

References
1. Bailey JO, Nowlin TP. Accuracy of
Frankfort plane transfer of the Hanau
articulator. J Dent Res 1981: 60: 531.
2. Bamber MA, Firouzai R, Harris M,
Linney AD. A comparative study of
two arbitrary face bow transfer systems
for orthognathic surgery planning. Int J
Oral Maxillofac Surg 1996: 25: 339343.
3. Boucher L, Jacoby J. Posterior border
movement of the human mandible movements. J Pros Dent 1961: 11: 836.
4. Bowley JF, Michaels GC, Lai TW, Lin
PP. Reliability of face bow transfer procedure. J Pros Dent 1992: 67: 491498.
5. Dahlberg G. Statistical method for medical and biological students. Interscience
New York 1940.
6. Donatsky O, Hillerup S, Jorgensen
JB, Jacobsen PU. Computerised cephalometric orthognathic surgical simulation, prediction and postoperative
evaluation of precision. Int J Oral Maxillofac Surg 1992: 21: 199203.
7. Ellis E, Tharanon W, Gambrell K.
Accuracy of face bow transfer: Effects on
surgical prediction and post surgical
result. J Oral Maxillofac Surg 1992: 50:
562567.
8. FONSECA RJ. Oral Maxillofac Surg. Saunders: Philadelphia, 2nd Vol. 2000, Chapter 5, p. 98.
9. Gateno J, Forrest KK, Camp B. A
comparison of three methods of face
bow transfer recording: Implication for
orthognathic surgery. J Oral Maxillofac
Surg 2001: 59: 635640.

10. Ghafari J, Engel FE, Laster LL.

Cephalometric superimposition on the
cranial base: A review and a comparison
of four methods. Am J Orthod Dentofac
Orthopod 1987: 91: 403413.
11. Gold BR, Setchell DJ. An investigation of the reproducibility of face bow
transfer. J Oral Rehab 1983: 10: 495
503.
12. Helkimo M, Ingervall B, Carlsson
GE. Variation of retruded and muscular
position of mandible under recording
conditions. Acta Odonto Scand 1971:
29: 423427.
13. Loh S, Yow M. Computer prediction of
hard tissue profile in orthognathic surgery. Int J Adult Ortho Orthognat Surg
2002: 17: 342347.
14. Lotzmann U. Considerations of precision and consistence of mandibular transverse hinge axis Zentrum fur ZahnMund-und Kieferheilkunde, 99. der
Georg-August-Universitat
Gottingen
1990: p. 372379.
15. McMillan LB. Border movements of
human mandible. J Pros Dent 1972: 27:
524.
16. OMalley MA, Milosevic A. Comparison of three face bow/semi adjustable
articulator systems for planning orthognathic surgery. Br J Oral Maxillofac Surg
2000: 38: 185190.
Address:
Ashraf F. Ayoub
Glasgow Dental Hospital & School
Professor of Oral & Maxillofacial Surgery
378 Sauchiehall Street
Glasgow
G2 3JZ
United Kingdom
Tel: +44 141 211 9604
Fax: +44 141 211 9601
E-mail: a.ayoub@dental.gla.ac.uk