The Laryngoscope

C 2015 The American Laryngological,

V
Rhinological and Otological Society, Inc.

Clinical Factor for Successful Nonsurgical Treatment of Pediatric

Peritonsillar Abscess

Dong-Kyu Kim, MD; Jung Woo Lee, MD; Yoon Sung Na, MD; Myung Jin Kim, MD;
Jun Ho Lee, MD; Chan Hum Park, MD

Objectives/Hypothesis: Current management for peritonsillar abscess in pediatric patients includes intensive medical
therapy using antibiotics with or without surgical drainage. However, attaining proper surgical drainage is difficult in young
children because they have narrow oral cavities and are less cooperative than adults. The aim of this study was to investigate
which factors are associated with a good response to nonsurgical treatment of a pediatric peritonsillar abscess.
Study Design: A retrospective analysis.
Methods: This study included consecutive children who visited our pediatric clinic for the treatment of peritonsillar
abscess. All patients initially received medical treatment, and additional surgical treatment was provided if the patient
appeared unlikely to recover. Multivariate logistic regression models were constructed to identify factors associated with a
good response to nonsurgical treatment. In addition, a receiver operating characteristic curve was used to identify the age
cutoff for predicting good treatment response.
Results: A total of 88 children were included the study. Patient age, recurrent tonsillitis, and abscess size were signifi-
cantly associated with response to nonsurgical treatment (adjusted odds ratios 5 1.485, 2.403, and 1.325, respectively) after
adjusting for sex, body mass index, season, serum C-reactive protein levels, and tonsil grade. The age cutoff associated with
good response to nonsurgical treatment was 7.5 years (80.0% sensitivity, 51.5% specificity).
Conclusion: Our findings suggest that a younger age, fewer episodes of acute tonsillitis, and smaller abscess size predict
a successful response to nonsurgical treatment in children with peritonsillar abscess.
Key Words: Peritonsillar abscess, deep neck infection, children, antibiotics, drainage.
Level of Evidence: 4.
Laryngoscope, 125:2608–2611, 2015

INTRODUCTION to a peritonsillar abscess due to tissue necrosis and pus

Peritonsillar abscess is a common deep neck infec- collection. Patients with peritonsillar abscess usually
tion in the pediatric population and often results in present with severe throat pain, fever, drooling, trismus,
emergency room visits due to severe pain, fever, and neck swelling, and an altered voice.1
swelling.1–4 Although peritonsillar abscess can occur in The traditional treatment for peritonsillar abscess
all age groups, it primarily affects children and adoles- in adults is surgical drainage with antibiotic therapy.7–10
cents.5,6 This disease occurs as a complication of acute In contrast, a pediatric peritonsillar abscess demands
tonsillitis and involves pus collection in the loose connec- special skills in terms of diagnosis and treatment
tive tissue of the peritonsillar space, which is located because of a child’s lower compliance and smaller anat-
between the capsule of the palatine tonsil and the supe- omy.6–8,11–13 In younger children, the most common pre-
rior constrictor muscle of the pharynx. Acute bacterial sentation is fever rather than sore throat.10,11 Neck
tonsillitis initially presents as cellulitis and progresses swelling and tenderness are also common presentations
of pediatric peritonsillar abscess because the infection
can easily spread into other deep neck spaces in chil-
From the Department of Otorhinolaryngology–Head and Neck
Surgery, Chuncheon Sacred Heart Hospital (D–K.K., J.W.L., Y.S.N., M.J.K., dren.10 Thus, some children may not present with symp-
J.H.L., C.H.P.); and the Nano-Bio Regenerative Medical Institute, toms directly associated with the oropharynx. Moreover,
Hallym University College of Medicine (D–K.K., J.H.L., C.H.P.), Chuncheon, obtaining an accurate history and performing an
Republic of Korea
Editor’s Note: This Manuscript was accepted for publication
adequate physical examination of the oropharynx is usu-
March 23, 2015. ally difficult in sick children.10 If a child is uncoopera-
This work was supported by Hallym University Research Fund tive, the needle aspiration used for diagnosis cannot be
and Cooperative Research Program for Agriculture Science and Technol- easily performed, resulting in potential risk to the air-
ogy Development (PJ011214022015), Rural Development Administration,
Republic of Korea. The authors have no funding, financial relationships, way and adjacent vascular structures.1,7,14
or conflicts of interest to disclose. Whether all children with a peritonsillar abscess
Send correspondence to Chan Hum Park, MD, PhD, Assistant Pro-
fessor, Department of Otorhinolaryngology–Head and Neck Surgery,
need surgical drainage is controversial. Some studies of
Chuncheon Sacred Heart Hospital, Hallym University College of pediatric peritonsillar abscess indicate that antibiotic
Medicine, 153, Kyo-Dong, Chuncheon, 200–704, Republic of Korea. therapy alone is not adequate, although certain abscess
E-mail: hlpch@daum.net
foci may not be drainable.11–14 However, surgical drain-
DOI: 10.1002/lary.25337 age in children carries a high risk of massive bleeding

Laryngoscope 125: November 2015 Kim et al.: Nonsurgical Treatment in Children With Peritonsillar Abscess
2608
 respiratory distress, and poor oral intake. In addition, we
administered a questionnaire on the number of acute tonsillitis
episodes within the last year to all caregivers. For each patient,
the adenotonsillar hypertrophy grade was evaluated by a pedi-
atric otorhinolaryngologist using an endoscope. The size of the
tonsil was evaluated as follows: grade 1, no or minimal tissue
(0% to 25% obstruction); grade 2, a small amount of tissue (26%
to 50% obstruction); grade 3, a large amount of tissue that
might cause obstructive symptoms (51% to 75% obstruction);
and grade 4, obstructing tissue (76% to 100% obstruction).17
White blood cell count and C-reactive protein (CRP) were meas-
ured in the laboratory study, and the size of abscess pockets
was also assessed by radiology. We divided the patients into a
good responder group and a poor responder group based on the
outcome of intensive medical therapy. Patients who received
only medical treatment were classified as good responders,
whereas those who underwent additional surgical treatment
during the admission period were classified as poor responders.
Fig. 1. Distribution of study samples according to their age.

due to the possibility of injuring the internal carotid Statistical Analysis

artery. Thus, some studies suggest that antibiotic therapy Differences in clinical parameters between groups were
can be used alone if there is no evidence of a compro- tested using the Mann-Whitney U test or Student t test, as
mised airway.6,7,15 To date, the factors associated with appropriate. Pearson’s chi-square tests were also used to evalu-
ate associations. In addition, multivariate logistic regression
response to nonsurgical treatment in children with peri-
models were constructed to identify predictors of good response
tonsillar abscess remain unknown. Therefore, the purpose to medical treatment after adjusting for confounding factors. A
of this study was to investigate characteristics associated receiver operating characteristic (ROC) curve was used to deter-
with good outcomes in children with peritonsillar abscess mine the cutoff age for poor response to medical treatment.
who underwent intensive medical therapy by analyzing SPSS statistical software version 18.0 (SPSS Inc, Chicago, IL)
data from our cumulative 10-year experience. was used to conduct the statistical analyses. A P value < 0.05
was considered statistically significant.

MATERIALS AND METHODS

Subjects RESULTS
This study included consecutive subjects who visited the Patient Characteristics
pediatric clinic or emergency department for treatment of peri- A total of 88 children (52 males, 36 females) diag-
tonsillar abscess from January 2002 through December 2012. nosed with peritonsillar abscess were enrolled in this
The disease was diagnosed via computed tomography (CT) scan,
study. Their mean age was 8.5 years (range: 2–12 years).
and abscess was confirmed by rim enhancement.16 Initially,
patients with no symptoms and/or signs of upper airway
The age distribution of the enrolled children is shown in
obstruction and systemic toxicity (hypotension, drowsiness) Figure 1; the most common age was 12 years. Of these
were admitted for close observation with intravenous hydration,
analgesia, and intravenous antibiotic treatment (targeting aero-
TABLE I.
bic and anaerobic pathogens). If the patient appeared unlikely
Characteristic of Children With Peritonsillar Abscess According to
to recover after 24 hours of intensive medical therapy, surgical the Response of Medical Treatment.
treatments such as drainage and tonsillectomy were provided.
Exclusion criteria were as follows: 1) 14 years of age or older; 2) Good Poor
Responder Responder
body mass index (BMI) at or above the 95th percentile; 3) previ- Group (n 5 33) (n 5 55) P value
ous diagnosis of peritonsillar abscess; and 4) craniofacial syn-
dromes, neuromuscular diseases, developmental delay, or other Age, mean (SD) 7.3 (3.1) 9.3 (2.7) 0.003
chronic psychiatric disorders. This retrospective study was Gender (male), n (%) 20 (60.6%) 32 (58.2%) 0.501
approved by the Ethics Committee of the Chuncheon Sacred
BMI, kg/m2 18.3 6 3.6 17.4 6 3.0 0.216
Heart Hospital, Hallym University College of Medicine, Chun-
cheon, Republic of Korea (2014-53). Informed consent was Season of admission, n (%) 0.711
waived by the institutional review board, and patient records Spring 6 (18.2%) 15 (27.3%)
and information were deidentified prior to analysis. Summer 6 (18.2%) 10 (18.2%)
Fall 5 (15.2%) 5 (9.1%)
Winter 16 (48.4%) 25 (45.4%)
Data Collection
C-reactive protein 76.36 6 54.0 72.6 6 59.4 0.558
All clinical information was obtained through a retrospec-
tive medical chart review, including age, sex, BMI, season of Tonsil grade 2.7 6 1.0 3.1 6 0.9 0.055
admission, past history of recurrent acute tonsillitis, presenting Tonsillitis (number/year) 2.6 6 1.3 3.9 6 1.8 0.001
symptoms, physical examination, laboratory studies, radiologi- Abscess size (cm2) 8.42 6 3.0 11.1 6 4.8 0.005
cal findings, and hospital course. Presenting symptoms included
fever, severe odynophagia, trismus, drooling, duration of pain, BMI 5 body mass index; SD 5 standard deviation.

Laryngoscope 125: November 2015 Kim et al.: Nonsurgical Treatment in Children With Peritonsillar Abscess
2609
 was 0.687 (95% confidence interval [CI]: 0.572–0.801;
P < 0.003) (Fig. 2).
Logistic regression models demonstrated that age
(adjusted odds ratio [OR] 5 1.485), number of recurrent
acute tonsillitis episodes (adjusted OR 5 2.403), and
abscess size (adjusted OR 5 1.325) were significantly
associated with response to intensive medical therapy.
The strongest association was found for the number of
recurrent acute tonsillitis episodes (adjusted OR 5 2.403,
95% CI: 1.329–4.347) (Table II).

DISCUSSION
The incidence of peritonsillar abscess in children is
approximately 14 to 30 cases per 100,000 children per
year.18 Although pediatric peritonsillar abscess is more
prevalent in adolescents than younger children, its diag-
Fig. 2. Receiver operating characteristic curve of the child’s age nosis and treatment present challenges for pediatricians.
according to the response of medical treatment in pediatric peri- In particular, peritonsillar abscess in young children
tonsillar abscess. demands special skills in terms of diagnosis and treat-
ment because of the patient’s low compliance and small
children, 33 (37.5%) responded well to intensive medical anatomy.11–13 To date, numerous peritonsillar abscess
therapy. General characteristics of good and poor respond- studies have been reported in the pediatric population.
ers to intensive medical therapy are presented in Table I. However, these studies have presented only descriptive
data, including age, symptom, seasons, bacteriology, and
admission course. In contrast, our study focused on
Factors Associated With Response to Medical treatment decisions in the pediatric population (medical
Treatment versus surgical) to allow for the recommendation of the
There were no significant differences in gender appropriate treatment. To the best of our knowledge,
(P 5 0.501) and BMI (P 5 0.216) between the good and poor this is the first large population study concerning treat-
responders. Similarly, we did not identify significant differ- ment decisions in children with peritonsillar abscess. We
ences in the season of admission (P 5 0.711), serum CRP showed that good response to nonsurgical treatment was
values (P 5 0.558), and tonsil grade (P 5 0.055) between significantly associated with age, tonsil grade, and his-
good and poor responders. Meanwhile, the history of recur- tory of recurrent acute tonsillitis episodes.
rent acute tonsillitis (number/year) (P 5 0.001) and size of There has been some debate about the role of tonsil-
abscess pockets (P 5 0.005) were significantly higher in the lectomy in the treatment of peritonsillar abscess.19–21
poor responders compared to the good responders. Several authors have challenged the idea that this condi-
We used the ROC curve analysis to evaluate the tion is an absolute indication for a tonsillectomy,19,21–23
clinical determining value of prediction of good response although tonsillectomy provides sufficient surgical drain-
to intensive medical therapy. The ROC curve analysis age and decreases the risk of future pharyngeal infec-
demonstrated that the cutoff age for the poor responder tions. With regard to pediatric peritonsillar abscess, an
group was 7.5 years (80.0% sensitivity, 51.5% specific- immediate (Quinsy) tonsillectomy has been advocated in
ity). The area under the ROC curve for good responders several studies in spite of the persisting controversy.8,12,14

TABLE II.
Logistic Regression Analyses to Identify Factors Associated With Good Response to Medical Treatment in Children With Peritonsillar
Abscess.
Univariate Analysis Multivariate Analysis

Variables OR 95% CI P value OR 95% CI P value

Age 1.249 1.070–1.459 0.005 1.485 1.154–1.911 0.002

Gender 0.904 0.375–2.181 0.823 1.195 0.192–7.448 0.849
BMI 0.919 0.804–1.051 0.216 0.888 0.732–1.077 0.227
Season 0.872 0.615–1.237 0.442 1.120 0.582–2.152 0.735
CRP 0.998 0.991–1.006 0.764 0.989 0.976–1.001 0.079
Tonsil grade 1.565 0.985–2.488 0.058 2.002 0.912–4.396 0.084
Tonsillitis history 1.742 1.238–2.451 0.001 2.403 1.329–4.347 0.004
Abscess size 1.172 1.044–1.316 0.007 1.325 1.079–1.627 0.007

BMI 5 body mass index; CI 5 confidence interval; CRP 5 C-reactive protein; OR 5 odds ratio.

Laryngoscope 125: November 2015 Kim et al.: Nonsurgical Treatment in Children With Peritonsillar Abscess
2610
These studies reported that immediate tonsillectomy is a with analgesia and hydration may be considered as a
safe and effective treatment option for children with peri- first-line treatment for young children presenting with a
tonsillar abscess compared to interval (or delayed) tonsil- small abscess.
lectomy. One large-scale study recommended that the
decision to perform an immediate tonsillectomy should be BIBLIOGRAPHY
based on a history of peritonsillar abscess and recurrent
1. Albertz N, Nazar G. Peritonsillar abscess: treatment with immediate ton-
tonsillitis.13 In this study, we found that the number of sillectomy—10 years of experience. Acta Otolaryngol 2012;132:1102–
tonsillitis episode and the size of abscess pocket were 1107.
2. Huang TT, Liu TC, Chen PR, Tseng FY, Yeh TH, Chen YS. Deep neck
inversely correlated with response to nonsurgical treat- infection: analysis of 185 cases. Head Neck 2004;26:854–860.
ment. Therefore, our findings indicate that tonsillectomy 3. Herzon FS, Nicklaus P. Pediatric peritonsillar abscess: management guide-
lines. Curr Probl Pediatr 1996;26:270–278.
may be required under certain conditions. 4. Novis SJ, Pritchett CV, Thorne MC, Sun GH. Pediatric deep space neck
Our ROC analysis indicated that the cutoff age for infections in U.S. children, 2000–2009. Int J Pediatr Otorhinolaryngol
2014;78:832–836.
poor response to nonsurgical treatment in children with 5. Risberg S, Engfeldt P, Hugosson S. Incidence of peritonsillar abscess and
peritonsillar abscess was 7.5 years. A previous study of relationship to age and gender: retrospective study. Scand J Infect Dis
2008;40:792–796.
pediatric peritonsillar abscess found that clinical presenta- 6. Hsiao HJ, Huang YC, Hsia SH, Wu CT, Lin JJ. Clinical features of peri-
tion differed between children 12 years and < 12 years of tonsillar abscess in children. Pediatr Neonatol 2012;53:366–370.
age.6 Another study reported that children 6 years old 7. Blotter JW, Yin L, Glynn M, Wiet GJ. Otolaryngology consultation for peri-
tonsillar abscess in the pediatric population. Laryngoscope 2000;110:
were more likely to respond to medical treatment than 1698–1701.
older children.7 Finally, one previous study offered a treat- 8. Richardson KA, Birck H. Peritonsillar abscess in the pediatric population.
Otolaryngol Head Neck Surg 1981;89:907–909.
ment algorithm for managing children with peritonsillar 9. Johnson RF, Stewart MG, Wright CC. An evidence-based review of the
abscess while taking their age into account.13 Therefore, treatment of peritonsillar abscess. Otolaryngol Head Neck Surg 2003;
128:332–343.
our findings suggest that pediatricians should consider 10. Friedman NR, Mitchell RB, Pereira KD, Younis RT, Lazar RH. Peritonsil-
nonsurgical treatment for younger children. lar abscess in early childhood. Presentation and management. Arch Oto-
laryngol Head Neck Surg 1997;123:630–632.
Our study has several limitations. First, a routine 11. Segal N, El-Saied S, Puterman M. Peritonsillar abscess in children in the
CT scan was used to confirm diagnosis. Some studies southern district of Israel. Int J Pediatr Otorhinolaryngol 2009;73:1148–
1150.
have asserted that using CT scans to diagnose peritonsil- 12. Holt GR, Tinsley PP Jr. Peritonsillar abscesses in children. Laryngoscope
lar abscess is relatively inaccurate compared to surgical 1981;91:1226–1230.
findings.7,16 However, CT scans are a useful tool if the 13. Schraff S, McGinn JD, Derkay CS. Peritonsillar abscess in children: a 10-
year review of diagnosis and management. Int J Pediatr Otorhinolar-
diagnosis is uncertain or a full clinical examination can- yngol 2001;57:213–218.
not take place.6,16,24 Second, the duration of intensive 14. Simon LM, Matijasec JW, Perry AP, Kakade A, Walvekar RR, Kluka EA.
Pediatric peritonsillar abscess: Quinsy ie versus interval tonsillectomy.
medical therapy administered in this study may have Int J Pediatr Otorhinolaryngol 2013;77:1355–1358.
been insufficient to improve a patient’s condition. How- 15. Plaza Mayor G, Martinez-San Millan J, Martinez-Vidal A. Is conservative
treatment of deep neck space infections appropriate? Head Neck 2001;
ever, we could not wait longer than 24 hours because a 23:126–133.
local infection can suddenly become a systemic infection 16. Capps EF, Kinsella JJ, Gupta M, Bhatki AM, Opatowsky MJ. Emergency
imaging assessment of acute, nontraumatic conditions of the head and
in children. Third, our study lacked information on some neck. Radiographic 2010;30:1335–1352.
important factors such as the socioeconomic status and 17. Brodsky L. Modern assessment of tonsils and adenoids. Pediatr Clin North
Am 1989;36:1551–1569.
the exact time of symptom onset, which may influence 18. Millar KR, Johnson DW, Drummond D, Kellner JD. Suspected peritonsil-
the immunologic status in children. In spite of these lim- lar abscess in children. Pediatr Emerg Care 2007;23:431–438.
itations, the advantages of the present study are its 19. Apostolopoulos NJ, Nikolopoulos TP, Bairamis TN. Peritonsillar abscess in
children. Is incision and drainage an effective management? Int J
large sample size and its use of the same treatment pro- Pediatr Otorhinolaryngol 1995;31:129–135.
tocol for each case of peritonsillar abscess. 20. Paradise JL, Bluestone CD, Colborn DK, Bernard BS, Rockette HE, Kurs-
Lasky M. Tonsillectomy and adenotonsillectomy for recurrent throat
infection in moderately affected children. Pediatrics 2002;110:7–15.
21. Wolf M, Kronenberg J, Kessler A, Modan M, Leventon G. Peritonsillar
CONCLUSION abscess in children and its indication for tonsillectomy. Int J Pediatr
We found that younger age, fewer episodes of acute Otorhinolaryngol 1988;16:113–117.
22. Kronenberg J, Wolf M, Leventon G. Peritonsillar abscess: recurrence
tonsillitis, and small abscess pockets are significant pre- rate and the indication for tonsillectomy. Am J Otolaryngol 1987;8:
dictors of good response to nonsurgical treatment in 82–84.
23. Wolf M, Even-Chen I, Talmi YP, Kronenberg J. The indication for tonsil-
pediatric peritonsillar abscess patients. Our study also lectomy in children following peritonsillar abscess. Int J Pediatr Otorhi-
suggests that the response of nonsurgical treatment sig- nolaryngol 1995;31:43–46.
24. Baker KA, Stuart J, Sykes KJ, Sinclair KA, Wei JL. Use of computed
nificantly deteriorates after 7.5 years of age. Therefore, tomography in the emergency department for the diagnosis of pediatric
nonsurgical treatment such as intravenous antibiotics peritonsillar abscess. Pediatr Emerg Care 2012;28:962–965.

Laryngoscope 125: November 2015 Kim et al.: Nonsurgical Treatment in Children With Peritonsillar Abscess
2611