J Oral Maxillofac Surg

Zygomatic Fractures Managed by Closed

66:2302-2307, 2008

Reduction: An Analysis With

Postoperative Computed Tomography
Follow-Up Evaluating the Degree of
Reduction and Remaining Dislocation
Birgitta af Geijerstam, MD,* Galina Hultman, MD,
Jakob Bergstrm, Med. Statistician, MSC, and
Pr Stjrne, MD, PhD

Purpose: To evaluate possible associations between the degree of reduction, remaining dislocation
(mm), fracture type, and the sequelae from which the patient may suffer postoperatively in patients with
zygomaticomaxillary fracture managed by closed reduction.
Material and Methods: A 3-year retrospective audit was undertaken to identify all patients who had
sustained a fractured zygoma and were operated on by closed reduction at the Ear, Nose, and Throat
Department, Karolinska University Hospital, Huddinge, Sweden. Patients were followed up by mail
questionnaire and postoperative computed tomography (CT).
Results: The odds of having symptoms (odds ratio [OR] 4.26, confidence interval [CI] 1.09-18.44) was
significantly higher in the group with a reduction less than 100% (n 34) compared with the group with
100% reduction (n 17) (P .035). The odds of having symptoms (OR 9.91, CI 0.89-500) was higher
in the group with remaining dislocation 6 to 10 mm compared with the group with no remaining
dislocation (P .069). The type of fracture (A, B, or C) also influenced the patients postoperative
symptoms. The odds of having symptoms was 48.40 (CI 4.60-500) times higher having fracture C
compared with fracture A (P .001).
Conclusions: The degree of reduction and remaining dislocation of zygomaticomaxillary fractures is
important to achieve a good postoperative result, that is, reducing the patients postoperative symptoms.
Furthermore, the type of fracture also influences the patients long-term sequelae.
2008 American Association of Oral and Maxillofacial Surgeons
J Oral Maxillofac Surg 66:2302-2307, 2008

One of the most common midfacial fractures is fracture
of the zygomaticomaxillary complex. The fracture is
also known as isolated zygoma fracture, malar fracture,
*Division of Ear, Nose, and Throat Diseases, Department of Clin-
ical Science, Intervention and Technology, Karolinska Institutet,
Stockholm, Sweden.
Department of Radiology, Karolinska University Hospital, Hud-
dinge, Sweden.
Medical Statistics, LIME, Karolinska Institutet, Stockholm,
Sweden.
Associate Professor and Director, Division of Ear, Nose, and
Throat Diseases, Department of Clinical Science, Intervention and
Technology, Karolinska Institutet, Stockholm, Sweden.
Address correspondence and reprint requests to Dr af Geijers-
tam: E.N.T. Department, Karolinska University Hospital, SE-141 86
Stockholm, Sweden; e-mail: birgitta.afgeijerstam@karolinska.se
2008 American Association of Oral and Maxillofacial Surgeons
0278-2391/08/6611-0017$34.00/0
doi:10.1016/j.joms.2008.06.050
or classic tripod fracture, more often anatomically
correctly classified as a tetrapod fracture. The epidemi-
ology, fracture pattern, treatment, and complications
have been extensively described in the literature.1,2
Controversy exists regarding the most appropriate
method of reduction of this type of fracture. Some
clinicians favor minimal exposure of the fracture site,
whereas others favor open reduction and fixation.
Marked differences in clinical practice exist about
which method is most frequently used because of
local traditions and the country for the patients treat-
ment.1-3 A majority of these patients suffer from post-
operative complications despite surgical intervention.
The complications are dependent on severity and
location of the fracture, and examples of complica-
tions are malar asymmetry, sensory deficits involving
the infraorbital nerve, limited opening of the mouth,
enophthalmos, and diplopia.
At the Ear, Nose, and Throat Department (ENT
Department) at the University Hospital of Karolinska,

2302
AF GEIJERSTAM ET AL
Huddinge in Sweden, a large number of zygomatico-
maxillary fractures are treated every year. The most
common treatment method for this type of fracture is
closed reduction using a traction hook. When there is
evidence of misalignment clinically an open approach
is used. To our knowledge, there is still a lack of
prospective studies that compares the outcome of
open versus closed reduction. Few publications de-
scribe reduction with a percutaneous applied traction
hook and then mostly with regard to the method itself
without investigations of the late sequelae. However,
Kaastad et al3 found that a large proportion of dis-
placed zygomaticomaxillary fractures could be
treated successfully solely by closed reduction using a
traction hook.
The aim of the present study was to evaluate closed
reduction in patients with zygomaticomaxillary frac-
tures in terms of degree of reduction (%), remaining
dislocation (mm), and postoperative sequelae. We
wanted to evaluate possible associations among the
degree of reduction, remaining dislocation fracture
type, and the sequelae from which the patient may
suffer postoperatively.
Patients and Methods
A search in the in-patient registry was performed to
identify all patients who were operated on by closed
reduction, operation code EEC 30,4 between the
years 1999 and 2001 at the ENT Department, Karo-
linska University Hospital, Huddinge, Sweden. The
search identified 105 patients in the register. Two
patients had isolated orbital floor fractures, and an-
other was managed first by closed reduction and then
by rigid fixation, and these 3 patients were excluded
from the study. One hundred two consecutive pa-
tients with zygomatic fractures were thus included in
the study. The fractures were classified according to
Zingg et al.5 Their classification system is based on
anatomic points and divides fractures into 3 catego-
ries: category A includes isolated fractures of 1 of the
3 processes of the zygomatic bone. These processes
are the temporal process, which forms the zygomatic
arch (A1), the frontal process, which forms the lateral
orbital wall (A2), and the maxillary process, which
forms the infraorbital rim (A3). Category B represents
fractures of all 3 processes, rendering the zygomatic
bone detached from the facial skeleton. This is the
so-called classic tripod fracture, but anatomically
these fractures are actually tetrapod, because the fron-
tal process of the zygoma also communicates with the
greater wing of the sphenoid in the orbital cavity,
which also requires to be disrupted to technically
render the zygoma free. Category C is the same as
type B, but with fragmentation, including the body of
the zygoma. In the present study, the distribution of
2303
each fracture type was as follows: 31% type A, 51%
type B, and 18% type C (n 102).
After case history and clinical examination, com-
puted tomography (CT) was carried out in 94 patients
and standard radiographs in 8 patients. To assess
initial signs and symptoms case records were re-
viewed. All fractures were performed via closed re-
duction, using either a traction hook (97 cases) or by
the Gillies method (5 cases). Five patients required a
second operation and were retreated with traction
hook. All patients were followed up 24 to 48 months
postoperatively via mail questionnaires describing
possible problems with limited opening of the mouth,
sensory deficits involving the infraorbital nerve, en-
ophthalmos, double vision, malar asymmetry, facial
pain, and increased frequency of maxillary sinusitis or
nasal congestion relative to the status before injury.
The symptoms were evaluated with symptom scores;
0 indicated no symptoms, 1 mild, 2 moderate, and 3
severe symptoms. Seventy-two patients (71%) an-
swered the questionnaire, and these patients were
recalled for a follow-up examination with a renewed
CT of the fractured zygoma. Fifty-one patients (50%)
showed up for renewed CT examination. Of the 21
patients who were not reexamined, 18 failed to show
up and 3 had moved out of the region. Thus, 51
patients were included in the analysis of degree of
reduction, remaining dislocation, fracture type, and
late sequelae.
EPIDEMIOLOGY
The routes of admission to the hospital were either
patients seeking directly to the ENT Department (37
cases) or via other departments in the hospital or/and
other hospitals (65 cases). Mean age for the patient
group was 41.0 years (SD 15.8, range 16-78
years). Of the 102 patients, 86 were men and 16
women. Cause and incidence of fractures are shown
in Figure 1. Motor vehicle accidents accounted for
only a small part of all fractures (4%). Nine patients
had fractures also in other parts of the facial skeleton,
mainly in the nasal bone and mandible. The patients
were presented to the ENT Department on average
1.7 days after the trauma (range 0-14 days), and
they underwent surgery on average 4.1 days after
trauma (range 0-10 days).
RADIOLOGICAL METHOD
All radiological calculations were made as CT ex-
aminations. The preoperative CT examinations were
made at different hospitals with different CT tech-
niques, most of them with spiral CT and 1- to 3-mm
slices.
The postoperative CT examinations were all made
at the same hospital on a multidetector-row CT scan-
ner (Siemens Somatom Volume Zoom) using the 4
2304
Cause and incidence of
zygomatico-maxillary fractures
50
Number of patients
40
30
20
10
0 separation of cases in the predictor space. In fact, the
er
t
C
t
le
ll
ul
or
Fa
yc
/M
th
sa
Sp
O
C
ar
As
Cause of injury
FIGURE 1. Cause and incidence of zygomaticomaxillary
fracture.
af Geijerstam et al. Zygomatic Fractures Managed by Closed
Reduction. J Oral Maxillofac Surg 2008.
1 mm detector rows, pitch 1.0. Images were first
reconstructed into 1-mm slice thickness at an interval
of 1 mm, and then multiplanary reformatted in the
axial, coronal, and saggital plan, permitting optimal
viewing plan regardless of scan plan. Three-dimen-
sional (3D) imaging was performed.
To measure the degree of operative reduction, the
fractures were radiologically diagnosed according
to the classification system for zygomaticomaxillary
complex fractures based on previously described an-
atomic points. Of the investigated 51 fractures, 15
were fractures of type A, 20 of type B, and 16 of type
C. To measure the displacement, 4 measuring points
were determined: 1) fractures through the anterior
orbital rim and orbital floor, 2) the posterolateral wall
of the maxillary sinus, 3) the zygomatic arch, and 4)
the zygomaticofrontal suture (Fig 2). The remaining
dislocation was measured postoperatively and sum-
marized for all 4 measuring points. The total differ-
ence in all measured parts between pre- and post-
operative displacement was then calculated and
expressed in percentages, that is, the degree of reduc-
tion. The degree of reduction was expressed in 3
categories, 0% to 50%, 50.1% to 99.9%, and 100%
reduction.
STATISTICAL METHOD
The analyses were performed using the LogXact
software, version 6.1 (Cytel Software Corparation,
Cambridge, MA). The Statistica software, version 7.1
(StatSoft Inc, Tulsa, OK) was used for all descriptive
statistics.
Exact methods are useful when data sets are small,
sparse, and highly unbalanced (Cytel User Manual,
2004).6 The LogXact 6.1 enables the use of exact
conditional maximum likelihood to estimate the pa-
ZYGOMATIC FRACTURES MANAGED BY CLOSED REDUCTION
rameters in a model; hence, the large sample assump-
tion underlying an asymptotic approach can be re-
laxed.7
A binary exact logistic regression was conducted to
evaluate the probability or odds of a patient having
Man symptoms depending on the degree of fracture reduc-
Woman tion (%), remaining dislocation of fracture (mm), and
fracture type. The factors were analyzed 1 by 1 to
overcome the problem of collinearity or complete
estimation of the beta-coefficients turned out to be
nonapplicable when trying to include all the factors in
the same model.
Results
INITIAL FINDINGS
At the time of admission, 56 patients (55%) had
difficulties opening their mouth. About one third
(30%) had observed visual flattening of the cheek and
asymmetry of the face, 26 patients (25%) had local
sensory disturbances such as hyposensibility. Double
vision was found in 8 patients (8%), and enophthal-
mos in none (0%) (Table 1).
FOLLOW-UP FINDINGS
At follow-up, 26% of the patients complained of
problems opening their mouth compared with 55%
before treatment. The prominence of the cheek was
considered well-shaped in 72% of cases. Sensory dis-
turbances and enophthalmos seemed to have in-
creased and double vision remained unchanged at
follow-up (Table 1).
The influence of the degree of reduction on the
patients postoperative signs/symptoms can be seen
in Figure 3. In the statistical analysis the odds of
having symptoms (odds ratio [OR] 4.26, confidence
interval [CI] 1.09-18.44) were significantly higher in
the group with a reduction less than 100% (n 34)
compared with the group with 100% reduction (n
17) (P .035). When dividing the degree of reduc-
tion into 3 subgroups (group 1 0%-50%, group 2
50.1%-99.9%, group 3 100%) and comparing the
groups with one another, the odds of having symp-
toms were 5.05 (CI 1.26-28.01) times higher at the
reduction degree of 50.1% to 99.9% than at the reduc-
tion degree of 100% (P .02).
The type of fracture (A, B, or C) also influenced the
patients postoperative symptoms. The influence of
fracture type on patients postoperative signs/symp-
toms at follow-up is shown in Figure 4. The odds of
having symptoms were 48.40 (CI 4.60-500) times
higher for fracture C compared with fracture A (P
.001) and comparing fracture B to A 5.68 (CI 1.06-
AF GEIJERSTAM ET AL
41.67) (P .041). When looking at the remaining
dislocation (mm) instead of the degree of reduction
(%) and its influence on patients postoperative symp-
toms, we observed that the odds for symptoms (OR
9.91, CI 0.88-500) were higher in the group with
the remaining dislocation 6 to 10 mm compared with
the group with no remaining dislocation (P .069).
There was a trend toward statistical difference be-
tween the group with greater than 10 mm dislocation
and the group with no remaining dislocation (OR
8.29, CI 0.70-471.93) (P .12).
Discussion
The results in the present study show that the
degree of reduction and remaining dislocation of zy-
gomaticomaxillary fractures are important factors for
2305
FIGURE 2. Zygomaticomaxillary complex (ZMC) fracture on the
left side of the face. A, Axial CT scan demonstrates 3 of 4
components of the ZMC fracture. 1, Fracture through the anterior
wall of the maxillary sinus at the level of the infraorbital canal; 2,
fracture of the posterolateral wall of the maxillary sinus; 3, zygo-
matic arch fracture. B, Axial CT scan demonstrates a fracture
through the zygomaticofrontal suture and lateral orbital wall (4).
C, Coronal CT scan demonstrates inferior and medial displace-
ment of the body of the left zygoma. There is a disruption of the
orbital floor (white arrow) and lateral wall of the maxillary sinus
(2), and a fracture through the zygomaticofrontal suture and lat-
erial orbital wall (4).
af Geijerstam et al. Zygomatic Fractures Managed by Closed
Reduction. J Oral Maxillofac Surg 2008.
a good postoperative result, that is, reduced postop-
erative symptoms. Furthermore, the type of fracture
strongly influences the patients long-term sequelae.
Because of the retrospective design of this study
the finding upon acute admission was not standard-
ized. Some of the case records were inconclusive, and
not all of the patients were submitted for a CT scan.
At follow-up via mail questionnaire the respond rate
was 71%. Even though this is a good respond rate,
there is always a risk for selection bias, meaning that
patients with very good or bad postoperative results
were not included in the study. Of those who under-
went renewed CT there was a mix of patients with
both high and low symptom scores.
The fractures were most often caused by assault,
which correspond well to the result found in compa-
rable studies.1,2,8 Motor vehicle accidents only ac-
2306 ZYGOMATIC FRACTURES MANAGED BY CLOSED REDUCTION

Table 1. SIGNS/SYMPTOMS BEFORE TREATMENT AND AT FOLLOW-UP IN PATIENTS WITH ZMC FRACTURES

Treated Cases Initial Follow-up

Findings (n 102) (n 72)

Signs/symptoms Fracture type A, B, and C A (1-3) B C Total

Limited opening of the mouth 56 (55%) 4 9 6 19 (26%)
Hyposensibility 26 (25%) 2 18 9 29 (40%)
Enophthalmos 0 (0%) 1 7 8 16 (22%)
Double vision 8 (8%) 0 3 3 6 (8%)
Asymmetry of cheek 31 (30%) 3 9 8 20 (28%)
Nasal congestion 2 7 6 15 (21%)
Sinus symptoms 3 5 4 12 (17%)
Facial pain 4 8 4 16 (22%)
af Geijerstam et al. Zygomatic Fractures Managed by Closed Reduction. J Oral Maxillofac Surg 2008.

counted for 4%, and the reason is probably that few than 100% in comparison with the 100% reduction
multitraumas are presented at our hospital. group. Comparing the 3 subgroups, the odds of hav-
Several other studies2,3,5,9 report data on complica- ing symptoms were significantly higher for the group
tions and postoperative symptoms in patients with with a 50.1% to 99.9% degree of reduction versus
zygomatic fractures. Previous studies of closed reduc- 100%. When examining remaining dislocation (mm),
tion using either a percutaneously applied traction the odds of having symptoms were substantially
hook3 or the Gillies method8,10-12 have shown satis- higher for the group with a remaining dislocation of 6
factory to excellent results on long-term evaluation in to 10 mm compared with the group with no remain-
zygomaticomaxillary fractures. In this study the aim ing dislocation. These results are clinically not sur-
was to investigate whether there is an association prising, and with a larger sample the results may be
between the patients postoperative symptoms and expected to be even more evident.
CT findings, that is, the degree of reduction, remain- In the present study, 51% of the fractures were type
ing dislocation, and fracture type. B fractures, which correspond well to the findings of
Our results show that there was a statistically sig- Tadj et al2 and Zingg et al5 (60.6% and 57%, respec-
nificant association between the degree of reduction tively). In addition, 18% were comminuted fractures
in a zygomatic fracture and the patients postopera- type C, which also correspond quite well to the find-
tive symptoms. Although the number of patients in ings of Tadj et al and Zingg et al (14.9% and 35%,
the study was limited and was unevenly distributed in respectively).
each subgroup, the odds of having symptoms were In addition, the type of fracture also plays an im-
4.26 times higher in the group with a reduction less portant role in terms of the patients postoperative

Degree of reduction and postoperative Fracture type and postoperative

signs/symptoms signs/symptoms
% of patients with symptoms

70%
60%
% of patients with symptoms

60%
50% Degree of
reduction
40% 50%
100%
30% 50,1-99,9% 40%
0-50%
20% Fracture type A
30%
(n=23)
10%
20%
0% Fracture type B
10% (n=34)
)

0)

9)

9)

6)

8)
15

13

13
2

=
=

=
(n

(n

(n

(n
(n

(n

(n

(n

0%
n

in

on

s
om
k

tio
ty

os

th

pa

Fracture type C
ee

si
ili

ou
lm

es

vi

pt

y ry os th n in s on
ch

ib

al

lit (n=15)
m

tio
ha

ng

et
m

pa om si
e

ou
ns

ci

bi m
bl
of

al vi
sy

es
Fa

pt
e
ht

co

i al
e

m
ou
th

ns m
os
ry

op

th ng ci le
s

m
al

e m
D

nu

e
g
yp

y
et

h a y b
En

h o
as

os As
in

op t F s
c ou
m

Si
H

en

al
N

g s
yp En
ym

in as nu D
op

H
en Si
As

N
s

op
m
le

s
ob

m
le
Pr

ob
Type of signs/symptoms Pr Type of signs/symptoms

FIGURE 3. Signs and symptoms at follow-up depending on the
degree of reduction (n 51). Each patient may be represented in
more than 1 symptom group.
af Geijerstam et al. Zygomatic Fractures Managed by Closed
Reduction. J Oral Maxillofac Surg 2008.
FIGURE 4. Signs and symptoms at follow-up depending on frac-
ture type (n 72). Each patient may be represented in more than
1 symptom group.
af Geijerstam et al. Zygomatic Fractures Managed by Closed
Reduction. J Oral Maxillofac Surg 2008.
AF GEIJERSTAM ET AL
symptoms. For example, patients with fracture type C
had 48.40 times higher odds to have the remaining
symptoms postoperatively compared with patients
with fracture type A (isolated fracture of the zygo-
matic arch). The finding that a more severe fracture
was associated with late sequelae was significant. In
contrast to other studies,1,2 in which all patients with
zygomatic fractures were included despite surgical
method, the present study only included patients
with fractures managed by closed reduction. This may
have affected the results because we may have had a
larger portion of type A and B fractures, that is, less
complicated fractures.
When comparing the degree of reduction in the 3
different types of fractures, we observed that type C
fractures more often had a lower degree of reduction.
Of the type C fractures only 6% (n 1/16) had a 100%
degree of reduction, whereas in type B and A frac-
tures the degree of reduction was 35% (n 7/20) and
60% (n 9/15), respectively. Our finding that 60% of
patients with a type A fracture had a 100% reduction
in combination with a generally low symptom score
at follow-up indicates that closed reduction is an ex-
cellent method of choice in type A fractures. In type
B fractures one should, depending on radiological
findings, fracture configuration, and the patients ini-
tial symptoms in some cases consider open reduction
instead of closed reduction. We feel that for more
advanced fractures, and certainly for type C fractures,
closed reduction is not to be recommended.
Even though there are some studies1,2 that retro-
spectively evaluated closed reduction versus open
fixation in zygomaticomaxillary fractures, there are no
randomized prospective studies. Open reduction
with internal fixation is associated with a significantly
more extensive surgical trauma, and earlier studies
have shown that this involves a risk of increased
postoperative sequelae. We are aware that our study
has a high frequency of postoperative complications,
although an initial frequency of almost 30% hyposen-
sibility postoperatively in both open and closed re-
duction has been reported earlier,2 which is equiva-
2307
lent to our findings. Furthermore, we used a
questionnaire for the 24- to 48-month follow-up. In
our opinion, using this method patients tend to report
higher frequency of symptoms compared with when
evaluated by physicians.
Prospective studies that compare open and closed
reduction would be valuable, because closed reduc-
tion means small structural trauma, short anesthesia,
and short hospitalization compared with open fixa-
tion.
In conclusion, the degree of reduction and remain-
ing dislocation of zygomaticomaxillary fractures cor-
relates well with the patients long-term sequelae.
Also, the type of fracture strongly correlates with the
patients postoperative symptoms. Closed reduction
is an excellent method of choice in type A fractures.
In more complicated fractures open reduction should
always be considered.
References
1. Ellis E, El-Attar A, Moos FK: An analysis of 2,067 cases of
zygomatico-orbital fracture. J Oral Maxillofac Surg 43:417, 1985
2. Tadj A, Kimble FW: Fractured zygomas. Aust N Z J Surg 73:49,
2003
3. Kaastad E, Freng A: Zygomatico-maxillary fractures: Late results
after traction-hook reduction. J Cranio Maxillofac Surg 17:210,
1989
4. Classification of Surgical Procedures. Stockholm, NOMESCO,
1997, p 53
5. Zingg M, Laedrach K, Chen J, et al: Classification and treatment
of zygomatic fractures. J Oral Maxillofac Surg 50:778, 1992
6. LogXact. Version 6 with Cytel Studio-Disrete Regression Soft-
ware Featuring Exact Methods, User Manual. Cambridge, MA,
Cytel Software Corparation, 2004, p 7
7. Agresti A: Catagorical Data Analysis (2nd ed). Hoboken, NJ,
Wiley & Sons, 2002, p 250
8. Ogden GR: The gillies method for fractured zygomas: An anal-
ysis of 105 cases. J Oral Maxillofac Surg 49:23, 1991
9. Balle V, Christensen P-H, Greisen O, et al: Treatment of zygo-
matic fractures: a follow-up study of 105 patients. Clin Otolar-
yngol 7:411, 1982
10. Larsen DO, Thomsen M: Zygomatic fractures II. A follow-up
study of 137 patients. Scand J Plast Recontr Surg 12:59, 1978
11. Lund K: Fractures of the zygoma: A follow-up study on 62
patients. J Oral Surg 29:557, 1971
12. Kristensen S, Tveteraas K: Zygomatic fractures: Classification
and complications. Clin Otolaryngol 11:123, 1986