ORIGINAL ARTICLE

Who needs admission among infants with acute otitis media?jpc_2249 435..438

Mohamed A Bitar,1,2 Rami Baz2,* and Omar Sabra1,†

Departments of 1Otolaryngology Head and Neck Surgery and 2Pediatrics and Adolescent Medicine, American University of Beirut School of Medicine and

Medical Center, Beirut, Lebanon

Aim: Management of acute otitis media (AOM) in infants younger than 2 months old is controversial. It varies between treatment on an outsidebasis, and hospitalization for intravenous antibiotics and sepsis work-up based on variability of the reported AOM pathogens in this particulargroup. Our aim is to identify clinical indicators that may suggest a need for an invasive medical work-up and/or hospitalization of these youngpatients, and compare their management to that of older infants.Methods: Retrospective chart review. Admitted infants with AOM and a random sample of infants presenting to the emergency room withAOM over a 20-year period. Infants younger than 2 months were designated as ‘young infants’, and those older as ‘older infants’. Demographicdata, relevant history, physical examination, laboratory studies and treatment were reviewed.Results: Twenty-nine admitted infants were included (13 young infants). A sample of 58 outpatients was studied, including two young infants.Compared to older inpatient infants, admitted young infants were less febrile (P < 0.05), had more benign white cell count (P < 0.05) but had moreotorrhea (P < 0.05). These grew gram-negative organisms. Sepsis work-up was negative. Young infants were more likely to be admitted (P < 0.05).Admitted older infants had more otorrhea than outpatients (P < 0.05) or a complication (P < 0.05).Conclusions: Young infants often need admission for intravenous antibiotics, until middle ear culture is out. Sepsis work-up may be necessaryonly in toxic patients. Older infants need admission when severely ill or have a complication.

Key words: bacteriology; infant; neonate; otitis media; sepsis work-up.

Acute otitis media (AOM) is one of the most common infectionsin children. Approximately half of children have an episode ofAOM by the age of 1 year.1 It is most common between the ageof 6 months and 3 years, with a second smaller peak between 4and 7 years. Clear guidelines have been set to treat AOM inchildren, including infants younger than 6 months of age,where treatment with antibiotics is recommended regardless ofthe certainty of the diagnosis.2 However, the management ofAOM in newborns and young infants (1–2 months of age) iscontroversial. While some physicians treat them on an outsidebasis, others insist on hospitalizing them to receive intravenousantibiotics after completing a sepsis work-up, including alumbar puncture. The latter management has been based on a

Correspondence: Associate Professor Mohamed A Bitar, Section of Pedi-atric Otolaryngology, Department of Otolaryngology – Head and NeckSurgery, American University of Beirut School of Medicine and MedicalCenter, PO Box 11-0236/A52, Beirut, Lebanon 1107-2020. Fax: +961 1370793; email: mb36@aub.edu.lb

*Currently is a Pediatrician at the Department of Pediatrics, EmundstonRegional, Hospital, Edmundston, NB – Canada.†Currently in private practice, Khobar, Kingdom Saudi Arabia.There is no potential conflict of interest, real or perceived, and no disclo-sures to be made.Presented as a poster at the annual meeting of the American Society ofPediatric Otolaryngology (ASPO) in Las Vegas, NV (Apr 30–May 2, 2010).

Accepted for publication 25 June 2011.

What is already known on this topic

1 The American Academy of Pediatrics and the AmericanAcademy of family Medicine have already set guidelines in 2004regarding management of uncomplicated AOM in childrenbetween 2 months and 12 years.

2 There is a belief in most of the literature that infants youngerthan 2 months of age (including neonates) have their AOMcaused by unusual organisms, mainly gram-negative bacteria.

3 Rare reports states that AOM in young infants is caused by theusual AOM organisms

What this study adds

1 Reviews a long-term practice in a tertiary academic centreregarding the management of AOM in infants, with emphasis onthose younger than 2 months.

2 It identifies clinical indicators that helps understand why aninfant may need admission for AOM with emphasis on infantsyounger than 2 months of age.

3 It suggests guidelines based on our and other available studiesin the literature discussing that issue.

doi:10.1111/j.1440-1754.2011.02249.x

bs_bs_banner

Journal of Paediatrics and Child Health 48 (2012) 435–438© 2011 The AuthorsJournal of Paediatrics and Child Health © 2011 Paediatrics and Child Health Division (Royal Australasian College of Physicians)

435

common belief that newborns and young infants with AOM arealways infected by bacteria, other than those encountered inolder infants and children. However, a recent study by Turneret al. showed that infants younger than 2 months are infectedwith the usual AOM pathogens.3

The absence of clear guidelines on how to approach thisparticular group of patients triggered us to look for specificclinical indictors to identify the infants who might need hospi-talization and a sepsis work-up.

Materials and Methods

We underwent a retrospective chart review of hospitalizedinfants with AOM as the primary diagnosis on admission, overa 20-year period, in a tertiary academic medical centre. Arandom sample of infants presenting to the emergency roomwith AOM over the same period was included. The number ofpatients in this sample was based on a ratio of 2:1 with theadmitted patients. Infants younger than 2 months (from birth to2 months, inclusive) were designated as ‘young infants’, thoseolder (from 2 to 12 months) as ‘older infants’. Demographicdata, relevant history, physical examination, laboratory studiesand treatment offered were reviewed. The mean values of out-patient and inpatient clinical indicators, as well as the charac-teristics of young versus older infants, were compared usingANOVA and a Chi-square test. Probability values below 0.05 wereregarded as significant.

Results

Twenty-nine admitted infants were included, 13 were younginfants (Table 1) and 16 were older (Table 2). A sample of 58outpatients (Table 3), were randomly selected from the totalnumber of patients who were seen and treated in the emer-gency room for AOM during the same period; only two (3.45%)were young infants. Compared to older inpatient infants, admit-

ted young infants were less frequently febrile (P < 0.05), hadmore frequently normal white blood cell (WBC) count (P <0.05) and less complications (P < 0.05), but more frequently hadotorrhea (P < 0.05) on admission (Fig. 1). Middle ear cultureswere available in 61.5% of the young inpatient infants. Thesegrew mainly gram-negative organisms, mostly Pseudomonasaeruginosa; while the older inpatient infants (cultures availablein only 31.2%) grew the usual AOM organisms. The cultureswere either taken from fresh ear discharge through middle earsuctioning (through an existing tympanic membrane perfora-tion) or intra-operatively when surgery was performed(Tables 1,2). Blood, urine and cerebral spine fluid cultures wereall negative. All inpatients had urine culture done; 84.6% ofyoung infants and 81.2% of older inpatients infants had bloodculture done; 53.8% of young infants and 25% of older inpa-tient infants had a lumbar puncture done. The young infantswere more likely to be admitted (44.8% of total inpatients vs.3.6% of total outpatients, P < 0.05). Admitted older infants hadmore otorrhea than non-admitted ones (31.2% vs. 3.7%, P <0.05) or a complication (43.1% vs. 0%, P < 0.05) (Table 4).

The patients treated on outside basis received oral antibioticsagreeing with the known guidelines,2 in 94.8% of the cases.

Table 1 Characteristics of admitted young infants

No Age (days) Gender Febrile WBC Organism Source

1 14 M No N Negative Otorrhea

2 17 M No N P. aeroginosa Otorrhea

3 20 F No H NA

4 30 M No H NA

5 35 F No N S. aureus Otorrhea

6 40 M No H E. cloacae, K. oxytoca Otorrhea

7 40 F No N P. aeroginosa Otorrhea

8 42 F No N NA

9 45 M Yes N NA

10 56 F No H P. aeroginosa Otorrhea

11 60 M No H P. aeroginosa Otorrhea

12 60 M No H P. aeroginosa Otorrhea

13 60 M Yes H NA

E, Escherichia; F, female; H, high count; K, Klebsiella; M, male; N, normal count; NA, not available; P, Pseudomonas; S, Staphylococcus; WBC, white blood

count.

0

10

20

30

40

50

60

70

80

Fever High WBC Complications Otorrhea

Pe

rce

nta

ge

Younginfants

Olderinfants

Fig. 1 Severity of illness among admitted young infants vs. older infants.

Otitis media in infants MA Bitar et al.

Journal of Paediatrics and Child Health 48 (2012) 435–438© 2011 The Authors

Journal of Paediatrics and Child Health © 2011 Paediatrics and Child Health Division (Royal Australasian College of Physicians)

436

Surprisingly, the inpatients were mostly treated using similarregimens but through an intravenous route (53.8% of younginfants and 75% of older infants). The remaining inpatientsreceived treatment aiming at covering the unusual AOM organ-isms (i.e. gram-negative bacteria) (Fig. 2).

Discussion

The incidence of AOM as a primary cause of febrile illnessamong young infants is not that common in our population. Wehave found only 13 young infants (4 being neonates) admittedfor AOM over a 20-year period. The estimated incidence amongyoung infants treated as outpatients was 3.64%. This is closeto Marom et al.’s reported incidence of 9.3%.4 However, this lowincidence does not apply to all populations. In a study byBoswell and Nienhuys,5 95% of the studied Aboriginal infantsdeveloped AOM or otitis media with effusion during the first8 weeks of life. The incidence was lower (30%) in non-Aboriginal Australian infants.

The need for sepsis work-up and hospitalization of younginfants suspected to have AOM is not quite clear. It is often done

for fear of development of complications secondary to AOM inthis age group. Nozicka et al. studied 40 infants 2–8 week old, ofwhom 15 (38%) were febrile. They found that all afebrileinfants had negative blood, urine and cerebrospinal fluid cul-tures; while only two febrile infants had positive sepsis work-up.6 We have noticed similar results among our admitted younginfants where only 15.4% were febrile. However, all sepsiswork-up was negative.

The development of complications in young infants secondaryto AOM is always a possibility. Marom et al. reported a 28%incidence of serious bacterial infection among febrile neonateswith AOM.4 Curns et al. believed that the concern for sepsis,meningitis or lower respiratory tract disease is the principlereason for admission of infants with AOM.1 This practice wasobserved in the population we studied, where the real reasonbehind admitting the young infants was the age itself, the fear of

Table 2 Characteristics of admitted older infants

No Age (months) Gender Febrile WBC Organism Source

1 2.5 F No H S. aureus, E. coli Otorrhea

2 3 M No H H. influenza Otorrhea

3 4 M No N NA

4 6 M Yes H a-haemolytic streptococcus Myringotomy

5 7 F Yes N NA

6 8 F Yes H NA

7 8 M Yes H NA

8 8 F Yes H NA

9 10 M No H NA

10 10 F No H Negative Myringotomy

11 10 F Yes H NA

12 11 M Yes H NA

13 11 M Yes N NA

14 11 M No H b-haemolytic streptococcus Myringotomy

15 12 M No H NA

16 12 F Yes N NA

E, Escherichia; F, female; H., Haemophilis; H, high; M, male; N, normal; NA, not available; S, Staphylococcus; WBC, white blood count.

Table 3 Characteristics of included outpatients

All outpatients Young infants Older infants

No 58 2 56

Gender: M/F 33/25 1/1 32/24

(1.3/1) (1.3/1)

Age range (weeks) 4–48 4–8 12–48

Febrile (%) 49 (84.48%) 2 (100%) 47 (83.93%)

Otorrhea (%) 2 (3.45%) 0 2 (3.57%)

0

10

20

30

40

50

60

Amox

icillin

Amox

iclav

1

Ceftri

axone

Ceftria

xone

com

b 2

Cefuro

xime

Ampic

illin

com

b 3

Cefpr

ozil

Cefota

xim co

mb 4

Cefpo

doxim

Cefixi

me

Azithr

omyc

in

Pe

rce

nta

ge

Outpatients

Young inpatientinfants

Older inpatientinfants

Fig. 2 Antibiotics regimens used in all studied patients. (1, Amoxicillin/

Clavulinic acid; 2, Ceftriaxone was combined with Oxacillin, Vancomycin, or

Erythromycin; 3, Ampicillin was combined with Gentamycin, Ceftriaxone, or

Cefotaxime; 4, Cefotaxime was combined with Oxacillin, or Vancomycin).

MA Bitar et al. Otitis media in infants

Journal of Paediatrics and Child Health 48 (2012) 435–438© 2011 The AuthorsJournal of Paediatrics and Child Health © 2011 Paediatrics and Child Health Division (Royal Australasian College of Physicians)

437

dealing with unusual AOM pathogens and the possibility ofdeveloping sepsis or other complications. This is in contrast withthe reason for admission of the older infants, which was eitherthe presence of a complication, a persistent fever or a high WBCcount. On the other hand, Turner et al. showed that the pres-ence of AOM in young infants does not predispose them toserious bacterial infections especially when afebrile.3 Seriousbacterial infections were diagnosed in only 6 of 137 (4%)patients.

The presence of unusual organisms causing AOM is anotherworrisome factor to be considered when deciding on the man-agement of a young infant with AOM. Though not all studiedinfants had an available middle ear fluid culture result, we havefound that AOM in young infants is mostly caused by gram-negative organisms, in contrast to older infants and the findingsof Turner et al., where enteric gram-negative bacilli representedonly 7% of the recovered organisms.3

We have noticed over that long period (20 years) that thecriteria to admit a young infant with AOM did not change (e.g.age, fever, elevated WBC); however, the frequency of perform-ing a lumbar puncture has decreased (unless meningeal signsare suspected).

The choice of the initial antibiotic regimen varied accordingto the treating physician’s high suspicion of the presence ofunusual AOM pathogens. There was no clear clinical indicatorthat made the physician skew towards one treatment option orthe other.

We acknowledge the presence of limitations in this study. It isa retrospective review, which means that all required data maynot be available in the charts. The number of cases is small, butthat was unavoidable due to the rarity of AOM among younginfants in our population. The non-uniformity of approach tothis problem made some results unavailable for all patients, forexample middle ear fluid, blood and cerebral spine fluidcultures.

Based on the results of our study and that of other similarstudies in the literature,1,3,4,6 we suggest the followings:1 To admit all young infants with confirmed AOM.2 To get a middle ear culture before starting intravenous anti-

biotics. The culture is either taken from fresh otorrhea ifpresent or via tympanocentesis if otorrhea is absent.

3 To perform a sepsis workup only in febrile young infants orin any infant who looks toxic.

4 To adjust antibiotic therapy when the middle ear culture isout, including switching to oral antibiotics and continuingtreatment on outside basis when appropriate.

References

1 Curns AT, Holman RC, Shay DK et al. Outpatient and hospital visitsassociated with otitis media among American Indian and Alaska nativechildren younger than 5 years. Pediatrics 2002; 109: E41–1.

2 American Academy of Pediatrics – Subcommittee on Management ofAcute Otitis Media. Diagnosis and management of acute otitis media.Pediatrics 2004; 113: 1451–65.

3 Turner D, Leibovitz E, Aran A et al. Acute otitis media in infantsyounger than two months of age: microbiology, clinical presentationand therapeutic approach. Pediatr. Infect. Dis. J. 2002; 21:669–74.

4 Marom R, Sakran W, Antonelli J et al. Quick identification of febrileneonates with low risk for serious bacterial infection: an observationalstudy. Arch. Dis. Child. Fetal Neonatal Ed. 2007; 92: F15–18.

5 Boswell JB, Nienhuys TG. Onset of otitis media in the first eight weeksof life in aboriginal and non-aboriginal Australian infants. Ann. Otol.Rhinol. Laryngol. 1995; 104: 542–9.

6 Nozicka CA, Hanly JG, Beste DJ, Conley SF, Hennes HM. Otitis media ininfants aged 0–8 weeks: frequency of associated serious bacterialdisease. Pediatr. Emerg. Care 1999; 15: 252–4.

Table 4 Complications among inpatients

Complication before admission No of patients Age range (weeks) Surgery

Young infants Facial palsy 1 2 Myringotomy + tube insertion

Older infants 12–44

Febrile seizure 4

Dehydration 1

Pneumonia 1

Mastoiditis 1 Myringotomy + tube insertion

Mastoiditis + subperiosteal abscess 1 Myringotomy + tube insertion + Incision & Drainage

Otitis media in infants MA Bitar et al.

Journal of Paediatrics and Child Health 48 (2012) 435–438© 2011 The Authors

Journal of Paediatrics and Child Health © 2011 Paediatrics and Child Health Division (Royal Australasian College of Physicians)

438