G Model

JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS

Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Oral and Maxillofacial Surgery,

Medicine, and Pathology
journal homepage: www.elsevier.com/locate/jomsmp

Original Research

Factors related to the treatment outcome of maxillofacial fascia space

infection
Pornchai Jansisyanont a,∗ , Watinee Kasemsai b , Pimol Bamroong b
a
Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, Chulalongkorn University, Bangkok, Thailand
b
Dental Department, Police General Hospital, Bangkok, Thailand

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To identify the predisposing factors related to maxillofacial space infection treatment out-
Received 21 August 2013 comes.
Received in revised form 23 March 2014 Methods: A retrospective study was performed. Our sample comprised 112 patients admitted to the
Accepted 16 April 2014
hospital with fascial space infections. The study variables were classified. Descriptive statistics were
Available online xxx
computed. The chi-squared test, t test, ANOVA, multivariate linear regression, and logistic regression
were used to analyze statistical differences between the groups.
Keywords:
Results: One hundred and twelve patients (65 males, 47 females) with a mean age of 33.29 ± 19.68 years
Maxillofacial space infection
Complication outcome
were admitted to the hospital with infections, mostly due to odontogenic causes. The mean time from
the first onset of symptoms to admission was 6.22 ± 4.94 days. Patients often presented with more than
one infected space (68.7%). The submandibular space was most commonly involved (46.4%), followed
by the pterygomandibular space (28.6%). The white blood cell count at admission ranged from 10,000
to 15,000 cell/mm3 . The length of hospital stay averaged 5.93 ± 5.39 days. The multiple linear regression
showed (statistical significance p = 0.002 and p = 0.003, respectively) that patients aged 41–60 years or
who had at least four involved spaces were positively involved with having longer hospitalizations. The
logistic regression showed that patients aged more than 40 years (p = 0.021) or who had at least four
involved spaces (p = 0.001) were more likely to develop complications.
Conclusion: The factors that are related significantly to the treatment outcome of maxillofacial infection
are increasing age, more than four involved fascial space infection, and immune compromised condi-
tions. The better understanding of these factors and close monitoring of the patients will prevent serious
complications.
© 2014 Asian AOMS, ASOMP, JSOP, JSOMS, JSOM, and JAMI. Published by Elsevier Ltd. All rights reserved.夽

1. Introduction are more common in children [1]. Caries is the most frequent
odontogenic etiology followed by pericoronitis and periodontal
Maxillofacial infection may range from low-grade, well localized disease [2]. Furthermore, there were several reported cases of
infections to severe life-threatening fascial space infection. Oral and cervicofacial necrotizing fasciitis occurring from dental infections
maxillofacial space infections have several etiologies such as odon- [3–6].
togenic cause, surgical treatment, upper respiratory tract infection, The majority of odontogenic infections are easily managed by
penetrating trauma, and malignancies. Odontogenic etiology is the minor surgical procedures and medical treatment. Occasionally,
most common cause in adults whereas upper respiratory infections these infections become severe in a very short time and cause more
serious complications. If the infection is not adequately treated,
it may develop into an abscess. This abscess may track along
夽 Asian AOMS: Asian Association of Oral and Maxillofacial Surgeons; ASOMP: Asian the fascial planes deeper into the neck and potentially extend
Society of Oral and Maxillofacial Pathology; JSOP: Japanese Society of Oral Pathol- into the mediastinitis leading to life-threatening complications [7].
ogy; JSOMS: Japanese Society of Oral and Maxillofacial Surgeons; JSOM: Japanese When life-threatening complications such as mediastinitis, pleu-
Society of Oral Medicine; JAMI: Japanese Academy of Maxillofacial Implants. ral emphyema, pericarditis, pericardial effusion, epidural abscess,
∗ Corresponding author at: Faculty of Dentistry, Chulalongkorn University, 34
jugular vein thrombosis, venous septic embolus, adult respiratory
Henri-Dunant Rd., Patumwan, Bangkok 10330, Thailand. Tel.: +66 8 9920 9260;
fax: +66 2 218 8581. distress syndrome, and septic shock occur, the mortality rate may
E-mail address: [email protected] (P. Jansisyanont). reach 40–50% [8].

http://dx.doi.org/10.1016/j.ajoms.2014.04.009
2212-5558/© 2014 Asian AOMS, ASOMP, JSOP, JSOMS, JSOM, and JAMI. Published by Elsevier Ltd. All rights reserved.夽

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
2 P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx

Patients with medical conditions that compromise their Locations of involved spaces were classified into two groups, deep
immune response, such as uncontrolled metabolic diseases (ure- and superficial groups depending on the cervical fascia that covered
mia, severe diabetes, alcoholism, and malnutrition), immune each space. The spaces that are covered by deep cervical fascia will
compromised conditions (leukemia, lymphoma, and malignant be categorized to the deep group, while the spaces that are covered
tumors), and immunosuppressive drugs (cancer chemotherapeu- by superficial cervical fascia will be categorized to the superficial
tic agents or immunosuppressive agents) often develop serious and group. Patients affected both deep and superficial groups will be
more rapidly progressive infection [9]. Several studies have com- category to deep group.
pared diabetic and nondiabetic patients with deep neck infection The outcome variables were the length of hospital stay and the
[10,11]. They found significant increases in age, length of hospital absence or development of complications: mediastinitis, pericar-
stay, more frequent complications, operation, and space involve- dia effusion, airway obstruction, or spread of the infection to other
ment in the diabetic group. In addition, other studies have shown spaces that required re-incision and drainage (re-I&D).
that the patients with medical conditions had higher risk for life
threatening deep neck infections [8,12]. Patients with immune dys- 2.3. Data management and analysis
function such as diabetes mellitus (DM) or HIV infection are at risk
for atypical and more complicated cases of deep neck infection Descriptive statistics were used for all the predictor variables.
[13,14]. The independent t test, chi-squared test, ANOVA, multivariate
A better understanding of the risk factors of fascial space infec- linear regression, and logistic regression were used to evaluate
tion complications would allow the members of these high-risk the relationships between study variables and outcome variables.
groups to be kept under close observation, thereby reducing the p < 0.05 was considered statistically significant.
possibility of complications and length of hospital stay. The objec-
tive of this study was to determine the factors related to the 3. Results
treatment outcome of maxillofacial space infections.
3.1. Demographics and clinical findings
2. Materials and methods
One hundred and twelve patients were admitted to hospital and
A retrospective study was performed by reviewing the records treated for maxillofacial space infections during the study period.
of 142 infected cases in the Department of Oral and Maxillofacial The patients comprised 65 males and 47 females with mean ages of
Surgery at the Faculty of Dentistry, Chulalongkorn University and 33.35 ± 19.07 and 33.21 ± 20.71 years, respectively. The descriptive
Police General Hospital from 2006 to 2010. The inclusion criterion statistics and characteristics of the patients are shown in Table 1.
was patients with fascial space infection admitted to the hospital by The ages of the patients ranged from 1 to 76 years.
the Department of Oral and Maxillofacial Surgery. Thirty cases were The mean time from the first onset of symptoms to admission
excluded because they had peritonsillitis, orocutaneous fistula or found in the present study was 6.22 ± 4.94 days (range 1–28 days).
flap necrosis from osteomyelitis, osteoradionecrosis, or soft tissue
reconstructive operation, rather than fascial space infection. Table 1
Patient characteristics.
2.1. Variables
Variables % (no. of patients, N = 112)

The predictor variables were age, gender, cause of infection, Sex

Male 58.04 (65)
immunocompromised condition, onset, number of space involve-
Female 41.96 (47)
ment, location of space involvement, and white blood cell (WBC) Age
count at admission. The outcome variables were the length of stay 0–20 years 27.68 (31)
and development of complications. 21–40 years 36.61 (41)
41–60 years 25.89 (29)
≥61 years 9.82 (11)
2.2. Data collection Cause of infection
Odontogenic 74.11 (83)
Data were classified into six major groups: demographic data, Surgery 21.43 (24)
cause of infection, immunocompromised condition, admission Others 4.46 (5)
Onset
data, fascial space involvement, and treatment type. The demo-
≤3 days 35.71 (40)
graphic variables were age and sex. The cause of infection was >3 days 64.29 (72)
defined as either odontogenic, surgical (including extraction and Immunocompromised condition
surgical removal), or others (pathology, trauma, and radiation). The Present 11.61 (13)
immunocompromised conditions were diabetes, steroid therapy, Number of involved spaces
1 space 31.25 (35)
organ transplant, malignancy, chemotherapy within the previous 2 spaces 36.61 (41)
year, chronic kidney disease, malnutrition, human immunodefi- 3 spaces 18.75 (21)
ciency virus, and the use of immunosuppressive medications [9,15]. ≥4 spaces 13.39 (15)
Information on the duration of onset (the first onset of symptoms Side of involved spaces
Right side 37.50 (42)
to admission), WBC count on the first day of admission, and the
Left side 58.04 (65)
space(s) involved were also obtained. The treatments received by Both sides 3.57 (4)
the patients were analyzed based on the type of antibiotic, type Midline 0.89 (1)
of treatment (divided to medication only or medication combined Location of involved spaces
with surgery), route and number of incision and drainage (I&D), or Deep 71.43 (80)
Superficial 28.57 (32)
type of anesthesia. Admission white blood cell count
Space involvement will be defined by operative note of oper- ≤10,000 29.11 (27)
ation room, if the diagnosis was cellulitis or abscess they will be 10,001 to ≤15,000 33.04 (37)
classified to inclusion criteria. In nonoperation case, the spaces >15,000 32.68 (31)
Unknown 15.18 (17)
were identified from definitive diagnosis in the medical record.

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx 3

Thirteen patients were immunocompromised with diabetes melli- Table 4

Distribution of overlapped space involvement.
tus found as the most common condition (10 cases with DM, 2 cases
with steroid therapy and 1 HIV case). Among the immunocom- Space % (no. of patients, N = 112)
promised patients, three patients developed complications. One Submandibular 46.43 (52)
patient developed necrotizing fasciitis and facial nerve palsy. One Pterygomandibular 28.57 (32)
had an airway obstruction before the operation. And one patient Buccal 25.00 (28)
who had been treated conservatively had a fistula tract develop- Submasseteric 17.86 (20)
Canine 16.96 (19)
ment postoperatively at the site.
Submental 16.07 (18)
Our study determined that 77 of 112 patients (68.75%) had more Lateral pharyngeal 15.18 (17)
than one involved space (Table 1); 10 patients had four involved Sublingual 13.39 (15)
spaces, 3 patients had five involved spaces, and 1 patient each had Vestibular 10.71 (12)
Preseptal 10.71 (12)
six involved spaces and seven involved spaces. The mean num-
Others 22.32 (25)
ber of involved space was 2.21 ± 1.196 (range 1–7 spaces). The
immunocompromised patients had 3.15 ± 1.46 spaces involved and
the noncompromised patients had 2.09 ± 1.10 spaces involved. The 3.3. Treatment
difference between these groups was significant (p = 0.002, t test).
Approximately 60% of infected spaces were on the patient’s left Based on our review, surgery (I&D and/or extraction) was per-
side. Thirty-seven patients presented with a WBC count in the formed on 101 (90.2%) patients, 92 of whom received I&D. The
range of 10,000–15,000 cells/mm3 , while approximately the same frequency and route of I&D are presented in Table 5. Nine patients
number, 29 and 32 presented with counts <10,000 cells/mm3 , and required repeat I&D. The majority of the patients required an
>15,000 cells/mm3 , respectively. intraoral approach for the first I&D. Intravenous penicillin plus
We identified that most infections occurred in patients in the metronidazole was the most commonly prescribed antibiotic regi-
21–40 years old group. Eighty-three (74.1%) of the infections had men. Seventeen patients had their initial antibiotics changed. This
a tooth-related cause (e.g., caries exposed pulp and infected wis- was due to allergy in five cases, due to no improvement in their
dom tooth). The majority of the odontogenic infections developed clinical status in four cases, and due to culture sensitivity test in six
from mandibular posterior teeth (56 cases, 67.5%) (Table 2), 25 of patients. The reasons in the remaining cases were not known. In
these 56 cases involved mandibular wisdom teeth with #38–48 the present study, multiple drug treatment was defined as patients
ratio of 2.57:1. The second most common cause of infection was who changed antibiotics three times or more. In the two cases of
surgery (21.4%), comprising extraction, surgical removal, genio-
plasty, and vestibuloplasty. Other causes (4.5%) were an infected
cyst, a fractured mandible, osteoradionecrosis, and BRONJ. Table 5
Frequency of treatment variables.
The average WBC count was 13,697 ± 9384 cells/mm3 in the
immunocompromised patients and 13,516 ± 5660 cell/mm3 in the Variables % (no. of patients)
noncompromised group. The difference was not statistically sig- Treatment (N = 112)
nificant (p = 0.948, t test). Our analysis revealed that patients with Medication only 9.82 (11)
high WBC counts at admission had a higher number of fascial spaces Combine med + surgery (I&D and/or extraction) 90.18 (101)
involved. However, this was not statistically significant (p = 0.066) Frequency of incision and drainage (N = 92)
One time 90.22 (83)
(Table 3).
Two times 8.70 (8)
Three times 1.08 (1)
Route of incision and drain
3.2. Fascial space involvement 1st time (N = 92)
Extraoral 17.39 (16)
The most commonly involved site was the submandibular space Intraoral 54.35 (50)
(52 cases), followed by the pterygomandibular space (32 cases), Transocket 4.35 (4)
Extraoral + intraoral 23.91 (22)
the buccal space (28 cases), and the submasseteric space (20 cases)
2nd time (N = 8)
(Table 4). Extraoral 37.50 (3)
Intraoral 37.50 (3)
Extraoral + intraoral 25.00 (2)
Table 2 3rd time (N = 1)
Distribution of infected teeth (odontogenic cause). Extraoral 100 (1)
Type of anesthesia
Variables % (no. of patients, N = 83) 1st time (N = 92)
LA 30.43 (28)
Upper anteriors 7.23 (6)
GA 63.04 (58)
Upper posteriors (exclude wisdom teeth) 19.28 (16)
Sedate 6.53 (6)
Upper wisdom teeth 2.41 (2)
2nd time (N = 8)
Lower anteriors 3.61 (3)
LA 37.50 (3)
Lower posteriors (exclude wisdom teeth) 37.35 (31)
GA 62.50 (5)
Lower wisdom teeth 30.12 (25)
3rd time (N = 1)
GA 100 (1)
Type of antibiotic (N = 112)
Table 3 Penicillin 14.29 (16)
Relation between admission WBC count and number of involved spaces. Penicillin + metronidazole 45.54 (51)
Amoxicillin/clavulanate potassium 8.93 (10)
Admission WBC count Mean no. of involved spaces
Clindamycin 10.71 (12)
≤10,000 1.81 ± 1.14 Amoxicillin/clavulanate potassium + metronidazole 8.93 (10)
10,001 to ≤15,000 2.14 ± 1.00 Others 8.04 (9)
>15,000 2.65 ± 1.38 Multiple drug treatment 3.57 (4)
Unknown 2.26 ± 1.15 Change antibiotics 15.18 (17)

There was no statistic significant difference at p value = 0.063 (ANOVA). I&D, incision and drainage.

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
4 P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx

multiple drug treatment, one developed mediastinitis and the other (p = 0.039) (Table 6). Significant increases in hospital stay were
required three times of I&D due to further spread of infection to pre- found between age range groups: between 0–20 years and 21–40
tracheal space. These patients were referred to an otolaryngologist years (p = 0.020), and between 0–20 years and 41–60 years
for further management. (p = 0.005). The patients who had at least four involved spaces had
significantly longer hospital stays compared to those with involve-
3.4. Length of hospital stay and complication: univariate analysis ment of one space (p < 0.001), two spaces (p < 0.001), or three spaces
(p = 0.007).
3.4.1. Length of hospital stay
The present study indicated that the mean duration of admis-
sion was 5.93 ± 5.389 days. Statistical analysis demonstrated 3.4.2. Complications
significant associations between duration of hospital stay and We found that complications developed in 11 patients (9.82%)
age (p < 0.001), compromised immune system (p = 0.027), num- (Table 7). One patient developed an airway obstruction as a pre-
ber of spaces involved (p < 0.001), and location of involved spaces operative complication. In 10 patients, the infections spread to
other spaces, 9 of which required repeat I&D. In the remaining
patient, the abscess drained through a fistula tract and did not
Table 6 need I&D. In those patients whose infections spread to other
Associations between study variables and length of hospital stay (LOS).
spaces, two patients developed serious complications: medias-
Study variable Mean ± SD of LOS p value tinitis and cervicofacial necrotizing fasciitis (Table 7). Statistical
Gender 0.954 analysis demonstrated significant associations between the devel-
Male 5.95 ± 5.41 opment of complication and age (p = 0.015), and the number of
Female 5.89 ± 54 spaces involved (p < 0.001) (Table 8).
Age <0.001a
0–20 years 3.39 ± 1.28
21–40 years 5.02 ± 3.09
41–60 years 9.48 ± 8.08 3.5. Length of hospital stay and complication: multivariate
≥61 years 7.09 ± 5.80 analysis
Cause of infection 0.27
Odontogenic 5.46 ± 5.19 The multivariable linear regression was done by selecting only
Surgery 7.08 ± 6.16
Others 8.20 ± 3.56
independent variables which were statistically significant from the
Onset 0.287 univariate analysis into the model. Patients aged between 41 and
≤3 days 5.20 ± 5.33 60 years were likely to have longer hospitalizations compared to
>3 days 6.33 ± 5.40 those aged less than 20 years (statistical significance p = 0.002).
Immunocompromised condition 0.027a
Having at least four involved spaces positively involved with longer
Present 10.31 ± 6.98
Absent 5.35 ± 4.89 admission days too (p = 0.003) (Table 9).
Number of involved spaces <0.001a The multivariable logistic regression was done to determine
1 space 4.97 ± 6.63 the association of patients’ characteristics and odds of develop-
2 spaces 4.63 ± 2.94 ing complications. The analysis was executed in the same fashion
3 spaces 6.05 ± 3.61
≥4 spaces 11.53 ± 6.19
as the first primary outcome (length of hospital stay) by selecting
Side of involved spaces 0.267 only independent variables which yielded statistical significance in
Right side 5.07 ± 4.01 the chi-square analysis. In this case, the age group and number of
Left side 6.25 ± 4.01 involved spaces served as predictors in the multivariable logistic
Location of involved spaces 0.039a
regression. However, the cutting points were modified to prevent
Deep 6.59 ± 4.88
Superficial 4.28 ± 6.24 failure prediction due to limited number of patients. The cutting
Admission white blood cell count 0.267 point was considered with respect to the statistical significance
≤10,000 1.81 ± 1.14 in the linear regression analysis for admission days. Respondents
10,001 to ≤15,000 2.14 ± 1.00 were coded 1 if their age was more than 40 years, and coded 0 if
>15,000 2.65 ± 1.38
Unknown 2.26 ± 1.15
otherwise. In addition, those having at least four involved spaces
were coded 1 and coded 0 if having less than four spaces. In logistic
a
There was a significant difference between the groups (p values for sex-LOS,
regression, patients aged more than 40 years (p = 0.021) or who had
compromised immune system-LOS, onset-LOS, side of involved spaces-LOS, and
location of involved spaces-LOS were calculated from independent t test whereas at least four involved spaces (p = 0.001) were more likely to develop
the remaining p values were obtained from one-way ANOVA). complications (Table 10).

Table 7
Summary of patients who developed complication outcome.

Case Age/sex Cause No. of involved spaces Compromised condition Complication LOS (day)

1 57/F Odontogenic 3 No Re-I&D 8

2 55/M Odontogenic 3 No Re-I&D 7
3 69/F Other (BRONJ) 2 No Re-I&D 10
4 49/M Surgery 4 DM, HT Re-I&D, NF, facial nerve palsy 27
5 36/F Odontogenic 4 No Re-I&D 10
6 66/M Surgery 3 DM, HT Spontaneous drainage 17
7 30/M Odontogenic 6 No Re-I&D, mediastinitis 16
8 62/F Odontogenic 4 No Re-I&D 14
9 51/M Surgery 5 No Re-I&D 18
10 55/F Odontogenic 3 No Re-I&D 7
11 34/M Odontogenic 7 HIV Airway obstruction 13

Re-I&D, re-incision and drainage; LOS, length of stay.

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx 5

Table 8 number of spaces involved, requiring an I&D, or developing com-

Associations between study variables and developing complications.
plication resulted in longer hospital stays.
Study variable Complications (N) p value Multiple layers of cervical fascia encase the contents of the
neck and form the potential head and neck spaces. These fascial
Absent Present
planes constitute important anatomic limitations for the spread of
Gender 1.000
infection. However, once inflammation or infections occur it can
Male 59 6
Female 42 5 spread along these planes. Infections typically present as celluli-
Age 0.015a tis, but can develop into a frank abscess with localized fluctuance,
0–20 years 31 0 erythema, warmth, and tenderness [13,16–18]. Odontogenic infec-
21–40 years 38 3 tions generally pass through three initial stages: inoculation (0–3
41–60 years 24 5
≥61 years 8 3
days), cellulitis (3–7 days), and abscess (over 5 days). The final
Cause of infection 0.690 stage becomes the resolution phase after spontaneous or surgical
Odontogenic 75 8 drainage of abscess cavity occurs [15].
Surgery 21 3 A thorough physical examination should be performed to deter-
Others 5 0
mine if the patient’s illness requires urgent care. The signs of serious
Onset 0.776
≤3 days 37 3 infection are elevation of the floor of the mouth, a change in voice
>3 days 64 8 quality, stridor, and change of consciousness [19]. Proper diagno-
Immunocompromised condition 0.117 sis and treatment, proper airway management, effective antibiotic
Present 10 3 therapy, and surgical intervention are important factors in the man-
Absent 91 8
agement of deep neck infections.
Number of involved spaces <0.001a
1 space 35 0 In the present study, wisdom teeth were found to be one of the
2 spaces 40 1 most common causes of infection. Educating patients about the risk
3 spaces 17 4 of infection from a wisdom teeth and the value of wisdom tooth
≥4 spaces 9 6
removal could reduce the incidence of space infection. We deter-
Side of involved spaces 0.051
Right side 39 3 mined that the most commonly involved space was submandibular
Left side 59 6 space (46.4%), which was similar to the findings of other studies
Both 2 2 [14,20]. We also found that most patients who developed more
Midline 1 0 than two involved spaces had submandibular space involvement.
Location of involved spaces 0.174
The submandibular space connects to several other spaces. The
Deep 70 10
Superficial 31 1 submandibular space is the most likely pathway for odontogenic
Admission white blood cell count 0.712 infection that passes through the thin lingual plate of mandible [15].
≤10,000 24 3 Early diagnosis and preventive treatment of this infected space can
10,001 to ≤15,000 34 3
reduce the spread of infection, hospital stay, and hospitalization
>15,000 26 5
Unknown 17 0
expense.
a
Statistical analysis indicated that hospital admission after more
Statistically significant between the groups (chi-squared).
than 3 days from the onset of infection was not associated with the
number of fascial spaces involved with the infection (p = 0.312). This
Table 9 finding corresponded to that of Flynn et al. [21], who found that the
Result of linear regression of study variables and length of hospital stay (LOS). number of days from the onset of infection before admission did not
Study variables Coefficients p value 95% CI correlate with the initial severity score. The severity score value is
a means for assessing and quantifying the severity of cases based
Age (vs. ≤20 years)
21–40 years 0.281 0.808 −2.006 to 2.569
on the location of the infected spaces [2,15,21,22]. In the present
41–60 years 4.297 0.002* 1.668–6.927 study, the number of fascial spaces involved in immunocompro-
≥61 years 2.606 0.123 −0.719 to 5.932 mised patients was statistically higher compared to the normal
Immunocompromised 1.446 0.35 −1.605 to 4.497 population (p = 0.002). Similar to other studies, suffering from dia-
Space involved (vs. 1 space)
betes mellitus was associated with deep neck infection and wide
2 spaces −0.984 0.381 −3.203 to 1.235
3 spaces −0.622 0.646 −3.300 to 2.057 spread inflammation [14,23]. Therefore, special attention should be
≥4 spaces 4.741 0.003* 1.670–7.813 paid to patients with an immunocompromised status.
Deep spaces (vs. superficial) 1.064 0.32 −1.049 to 3.176 Previous studies compared diabetic and nondiabetic patients
R2 = 0.328. with deep neck infections, finding significant increases in age, dura-
*
Statistically significant (p < 0.05). tion of hospital stay, more frequent complications, and number of
operations and involved space in the diabetic group [10,11]. Short-
and long-term hyperglycemia alters a patient’s immune function,
Table 10
Result of logistic regression of study variables and complications. including defects in white blood cell migration and chemotaxis,
neutrophil bactericidal function, and vascular blood flow to small
Study variables Odds ratio p value 95% CI
vessel tissue beds [15,24]. Therefore, diabetic patients are suscep-
Age >40 years (vs. age ≤40 years) 6.006 0.021* 1.312–27.477 tible to more serious infections, such as necrotizing fasciitis and
≥4 spaces involved (vs. <4 spaces) 12.76 0.001* 2.905–56.054 fungal infections. The control of blood sugar levels is important in
*
Statistically significant (p < 0.05). the management of diabetic patients.
In our study, severe complications encountered included airway
obstruction, necrotizing fasciitis (NF), and descending mediastini-
4. Discussion tis. One patient had an airway obstruction and received emergency
intubation at the ward. This patient had a diagnosis of Ludwig’s
The present study investigated the factors involved with max- angina, buccal, pterygomandibular, and lateral pharyngeal space
illofacial space infection treatment outcomes. We found that infections. Airway maintenance is of paramount importance in
numerous patient characteristics including age, immune status, high-risk groups. Another study reported NF as a complication of

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
6 P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx

the removal of wisdom tooth [3]. In our study, NF developed in a and noncompromised group was not statistically different, which
patient after the extraction of tooth #26, which had an exposed is similar to the study of Zheng et al. [14]. Based on this, WBC count
pulp. There were no signs of infection at the initial physical exam- at admission might not to be a good predictor of severe infections,
ination. Incision and drainage were performed at the canine space, or identify an immunocompromised patient.
but the infection then spread to the buccal, orbital, and superficial Studies have shown that immunocompromised patients were
temporal spaces. Subsequently, NF developed at the buccal area, statistically associated with longer hospital stays and devel-
and facial nerve palsy occurred. This NF patient also had uncon- oping complications [8,14,34]. However, in the present study
trolled DM, which affected the immune status of the patient. immunocompromised patients were not statistically associated
Descending necrotizing mediastinitis usually occurs as a com- with developing complications. This was possibly due to the lower
plication of odontogenic infections or peritonsillar abscess [25,26], number of immunocompromised patients compared to healthy
especially in the diabetic, immunocompromised, or debilitated patients in our study. In the current study, age and multiple spaces
patients [26,27]. In the current study, the patient who developed involvement were statistically associated with developing compli-
descending mediastinitis was middle aged with no underlying dis- cations. Increasing age tended to have a positive correlation with
ease. The cause of the infection was from a mesioangular impacted longer admission days. Of note was that the significant finding was
#38 with a caries exposed pulp. The infection involved the sub- only for late adulthood (aged 41–60 years) patients but not for the
mandibular, lateral pharyngeal, and visceral spaces. The infection elderly (aged at least 61 years). The nonstatistical significance in the
later spread to the mediastinum and caused a pericardial effusion. eldest age group was due to the statistical matter rather than the
Thoracotomy and substernal mediastinitis drainage were per- clinical rationale since its population size was comparatively small.
formed by a cardiovascular thoracic surgeon. The patient recovered Increased age results in decreased immune system function; there-
slowly and was transferred to a different hospital for rehabilitation. fore, the complication rate can increase. Older patients (>65 years)
The visceral fascia is the connective tissue coat surrounding the tended to suffer multiple associated conditions, such as pneumonia
esophagus, thyroid gland, and trachea. The visceral fascia consists or malnutrition [33]. A previous study showed that patients with
of the anterior pretracheal and posterior buccopharyngeal fascias more than two involved spaces were likely to develop complicated
[18]. The pretracheal fascia fuses with the pericardial and pari- deep neck infections [31].
etal pleural in the anterior mediastinum. We believe that in our
descending mediastinitis case, the infection spread anteriorly to 5. Conclusion
the thyroid gland and then accessed the anterior mediastinum
via the pretracheal space. Studies have shown that patients who Increasing age and involvement at least four spaces manifested
develop mediastinitis had a high mortality rate of 40–50%, usually strong association with longer hospital stay and risk of develop-
from overwhelming sepsis [8,18]. With improvements in antibi- ing complications. Immunocompromised status was regarded as
otics, imaging, and surgical management, the mortality rate has another risk factor for prolonging admission days though with a
declined to 20–40% [26]. less statistical significance. The better understanding about the
The present study found that the most commonly used antibi- related factors of the treatment outcomes of maxillofacial fascia
otics were penicillin and metronidazole, which is similar to the space infection might substantially benefit the healthcare providers
study of Bakir et al. [20]. High-dose intravenous penicillin is still to prevent serious consequences.
the empiric drug of choice, because most oral flora remain sus-
ceptible to penicillin [28,29]. However, some oral pathogens can
References
synthesize ␤-lactamase, which can hydrolyze the ␤-lactam ring of
penicillin, rendering them penicillin resistant. Approximately 25% [1] Caccamese Jr JF, Coletti DP. Deep neck infections: clinical considerations in
of the strains of the Prevotella and Porphyromonas are able to syn- aggressive disease. Oral Maxillofac Surg Clin North Am 2008;20:367–80.
thesis this enzyme. ␤-Lactamase can also be found in some strains [2] Flynn TR, Shanti RM, Levi MH, Adamo AK, Kraut RA, Trieger N. Severe
odontogenic infections. Part 1: Prospective report. J Oral Maxillofac Surg
of Fusobacterium and Streptococcus. The prevalence of antibiotic 2006;64:1093–103.
resistance is increasing. The use of penicillin plus a ␤-lactamase [3] Ricalde P, Engroff SL, Jansisyanont P, Ord RA. Paediatric necrotizing fasciitis
inhibitor or a penicillin plus metronidazole is a reasonable treat- complicating third molar extraction: report of a case. Int J Oral Maxillofac Surg
2004;33:411–4.
ment for serious odontogenic infections [9,18,19,30]. [4] Brunworth J, Shibuya TY. Craniocervical necrotizing fasciitis resulting form
Other studies have shown that increased length of stay for dentoalveolar infection. Oral Maxillofac Surg Clin North Am 2011;23:425–32.
patients with odontogenic infections was predicted by the number [5] Sahoo NK, Tomar K. Necrotizing fasciitis of the cervico-facial region due to
odontogenic infection. J Oral Maxillofac Surg Med Pathol 2014;26:39–44.
of spaces involved, the severity of the infection, the occurrence of [6] Schutz P, Joshi RM, Ibrahim HHH. Odontogenic necrotizing fasciitis of the neck
complications, such as penicillin therapeutic failure, and the need and upper chest wall. J Oral Maxillofac Surg Med Pathol 2012;24:32–5.
for reoperation [22,31]. These findings correspond with those of [7] Daramola OO, Flanagan CE, Meisel RH, Odland RM. Diagnosis and treatment of
deep neck space abscesses. Otolaryngol Head Neck Surg 2009;141:123–30.
our study, where patients with more than three involved spaces,
[8] Chen MK, Wen YS, Chang CC, Huang MT, Hsiao HC. Predisposing factors of
who underwent frequent I&D, or who developed complication had life-threatening deep neck infection: logistic regression analysis of 214 cases.
longer hospital stays. Moreover, in the present study, older or J Otolaryngol 1998;27:141–4.
[9] Ellis E, Hupp J, Tucker M. Contemporary oral and maxillofacial surgery. 6th ed.
immune compromised patients also had longer hospital stays in
St. Louis: Mosby; 2013.
the univariate analysis. There was a positive correlation between [10] Chen MK, Wen YS, Chang CC, Lee HS, Huang MT, Hsiao HC. Deep neck infections
the length of admission and the age of the patients, a finding sim- in diabetic patients. Am J Otolaryngol 2000;21:169–73.
ilar to that of the study of Wang et al. [32]. Santos Gorjon et al. [11] Huang TT, Tseng FY, Liu TC, Hsu CJ, Chen YS. Deep neck infection in dia-
betic patients. Comparison of clinical picture and outcomes with nondiabetic
[33] reported that the patient’s underlying medical condition was patients. Otolaryngol Head Neck Surg 2005;132:943–7.
a good predictor for the length of hospital stay. In immunocompro- [12] Boscolo-Rizzo P, Da Mosto MC. Submandibular space infection: a potentially
mised patients, the hospital stay was more than 5 days on average lethal infection. Int J Infect Dis 2009;13:327–33.
[13] Vieira F, Allen SM, Stocks RM, Thompson JW. Deep neck infection. Otolaryngol
due to the comorbidity and multidisciplinary treatment received Clin North Am 2008;41:459–83.
[33]. In several studies, WBC count at admission was a significant [14] Zheng L, Yang C, Kim E, Zhang W, Cai X, Jiang B, et al. The clinical features
predictor of the length of hospital stay [21,32,35,36]. In contrast, of severe multi-space infections of the head and neck in patients with dia-
betes mellitus compared to non-diabetic patients. Br J Oral Maxillofac Surg
in our study, WBC count at admission was not statistically associ- 2012;50:757–61.
ated with hospital stay or developing complications. Furthermore, [15] Miloro M, Ghali G, Larsen P, Waite P. Peterson’s principle of oral and maxillo-
we found that WBC count in the immunocompromised patients facial surgery. 3rd ed. Shelton: PMPH; 2012.

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009
G Model
JOMSMP-294; No. of Pages 7 ARTICLE IN PRESS
P. Jansisyanont et al. / Journal of Oral and Maxillofacial Surgery, Medicine, and Pathology xxx (2014) xxx–xxx 7

[16] Osborn TM, Assael LA, Bell RB. Deep space neck infection: principle of surgical [27] Sancho LM, Minamoto H, Fernandez A, Sennes LU, Jatene FB. Descending necro-
management. Oral Maxillofac Surg Clin North Am 2008;20:353–65. tizing mediastinitis: a retrospective surgical experience. Eur J Cardiothorac
[17] Yellon RF. Infections of the fascial spaces of the head and neck in children. Surg 1999;16:200–5.
Semin Pediatr Infect Dis 1998;9:60–9. [28] Rega AJ, Aziz SR, Ziccardi VB. Microbiology and antibiotic sensitivities of
[18] Marra S, Hotaling AJ. Deep neck infection. Am J Otolaryngol 1996;17: head and neck space infections of odontogenic origin. J Oral Maxillofac Surg
287–98. 2006;64:1377–80.
[19] Topazian RG, Goldberg MH, Hupp J. Oral and maxillofacial infections. 4th ed. [29] Jansisyanont P, Sessirisombat S, Sastravaha P, Bamroong P. Antibiotic pro-
Philadelphia: Saunders; 2002. phylaxis for orthognathic surgery: a prospective, comparative, randomized
[20] Bakir S, Tanriverdi MH, Gun R, Yorgancilar AE, Yildirim M, Tekbas G, et al. Deep study between amoxicillin–clavulanic acid and penicillin. J Med Assoc Thai
neck space infections: a retrospective review of 173 cases. Am J Otolaryngol 2008;91:1726–31.
2012;33:56–63. [30] Brook I, Frazier EH, Gher ME. Aerobic and anaerobic microbiology of periapical
[21] Flynn TR, Wiltz M, Adamo AK, Levy M, McKitrick J, Freeman K, et al. Predicting abscess. Oral Microbiol Immunol 1991;6:123–5.
length of hospital stay and penicillin failure in severe odontogenic infections. [31] Lee JK, Kim HD, Lim SC. Predisposing factors of complicated deep neck infection:
Int J Oral Maxillofac Surg 1999;28:48. an analysis of 158 cases. Yonsei Med J 2007;48:55–62.
[22] Flynn TR, Shanti RM, Hayes C. Severe odontogenic infections, Part 2: Perspective [32] Wang J, Ahani A, Pogrel MA. A five-year retrospective study of odontogenic
outcomes study. J Oral Maxillofac Surg 2006;64:1104–13. maxillofacial infections in a large urban public hospital. Int J Oral Maxillofac
[23] Hasegawa J, Hidaka H, Tateda M, Kudo T, Sagai S, Miyazaki M, et al. An analy- Surg 2005;34:646–9.
sis of clinical risk factors of deep neck infection. Auris Nasus Larynx 2011;38: [33] Santos Gorjon P, Blanco Perez P, Morales Martin AC, Del Pozo de Dios JC, Estevez
101–7. Alonso S, Calle de la Cabanillas MI. Deep neck infection: review of 286 cases.
[24] Delamaire M, Maugendre D, Moreno M, Le Goff MC, Allannic H, Genetet Acta Otorrinolaringol Esp 2012;63:31–41.
B. Impaired leukocyte functions in diabetic patients. Diabetic Med [34] Huang TT, Liu TC, Chen PR, Tseng FY, Yen TH, Chen YS. Deep neck infection:
1997;14:29–34. analysis of 185 cases. Head Neck 2004;26:854–60.
[25] Kinzer S, Pfeiffer J, Becker S, Ridder GJ. Severe deep neck space infections and [35] Dodson TB, Barton JA, Kaban LB. Predictors of outcome in children hospitalized
mediastinitis of odontogenic origin: clinical relevance and implications for with maxillofacial infections: a linear logistic model. J Oral Maxillofac Surg
diagnosis and treatment. Acta Otolaryngol 2009;129:62–70. 1991;49:838–42.
[26] Roccia F, Pecorari GC, Oliaro A, Passet E, Rossi P, Nadalin J, et al. Ten years of [36] Seppanen L, Lemberg KK, Lauhio A, Lindqvist C, Rautemaa R. Is dental treatment
descending necrotizing mediastinitis: management of 23 cases. J Oral Maxillo- of an infected tooth a risk factor for locally invasive spread of infection? J Oral
fac Surg 2007;65:1716–24. Maxillofac Surg 2011;69:986–93.

Please cite this article in press as: Jansisyanont P, et al. Factors related to the treatment outcome of maxillofacial fascia space infection.
J Oral Maxillofac Surg Med Pathol (2014), http://dx.doi.org/10.1016/j.ajoms.2014.04.009