Bacterial etiology of acute otitis media andclinical efficacy of amoxicillin—clavulanateversus azithromycin

Mehmet Guven a,*, Yunus Bulut b, Taner Sezer c, Ibrahim Aladag a,Ahmet Eyibilen a, Ilker Etikan d

International Journal of Pediatric Otorhinolaryngology (2006) 70, 915—923

www.elsevier.com/locate/ijporl

aDepartment of Otorhinolaryngology, Faculty of Medicine, Gaziosmanpasa University, Tokat, TurkeybDepartment of Microbiology, Faculty of Medicine, Gaziosmanpasa University, Tokat, TurkeycDepartment of Pediatrics, Faculty of Medicine, Gaziosmanpasa University, Tokat, TurkeydDepartment of Health Sciences, Faculty of Medicine, Gaziosmanpasa University, Tokat, Turkey

Received 20 July 2005; received in revised form 2 October 2005; accepted 6 October 2005

KEYWORDSBacterial etiology;Acute otitis media;Azithromycin;Amoxicillin—clavulanate

Summary

Backgrounds: Acute otitis media (AOM) is one of the most common acute bacterialinfection in childhood and also the most frequent reason for outpatient antibiotictherapy. Little recent information about susceptibility patterns of AOM bacterialpathogens in Turkish children has been reported.Objective: To determine the bacterial etiology of acute otitis media in children andto compare the efficiency of 3 days course of azithromycin with a 10 days course ofamoxicillin—clavulanate.Methods: This prospective, single blind, randomised comparative study was carriedout in 180 children with AOM. Paracentesis was performed for middle ear fluid culturebefore the first dose antibiotic therapy. Children with acute otitis media wererandomised to receive either low dose amoxicillin—clavulanate (45/6.4 mg/kg/dayin two divided doses for 10 days) or low dose azithromycin (10 mg/kg/day for 3 days).Clinical response was assessed on days 2—4, 11—13, 26—28.Results: Bacterial pathogens were isolated from 108 (60%) of 180 children. Strepto-coccus pneumoniae was the most common isolated pathogen (39.7%), followed byHaemophilus influenzae (20.7%), Moraxella catarrhalis (15.5%), Staphylococcus aur-eus (13.8%), Group A beta-hemolytic streptococcus (5.1%), Escherichia coli (3.4%) andEnterococcus faecalis (1.7%). This study demonstrated low resistance rates compared

* Corresponding author. Tel.: +90 356213 31 79; fax: +90 356213 31 79.E-mail address: guvenmehmet28@yahoo.com (M. Guven).

0165-5876/$ — see front matter # 2005 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ijporl.2005.10.004

916 M. Guven et al.

1. Introduction

Acute otitis media (AOM) is one of the most commonacute bacterial infectious diseases in childhood andthe most frequent reason for outpatient antibiotictherapy. Inadequate treatment or frequent reinfec-tions predispose children to persistent middle eareffusions (MEE) and this may lead to difficulties inhearing and verbal communications and also causemany complications such as meningitis, facial palsyand other intracranial complications [1].

The percentages of etiologic agents that causeAOM were found to be bacteria (62%), viruses plusbacteria (45%), and viruses (75%), respectively inprevious studies [2,3]. The most common bacterialorganisms causing AOM are Streptococcus pneumo-niae (20—40%), Haemophilus influenzae (10—30%)and Moraxella catarrhalis (5—15%). Less commonlystaphylococci, streptococci and gram-negative rodsare the other causative agents. It was found that 56%of middle ear aspirates had bacteria and 36% ofthese were S. pneumoniae in a prospective studyin Turkey [4]. Although several attempts have beenmade to generate definitive and broad recommen-dations for the treatment of AOM in children, muchdebate still exists about this topic. Recent consensusrecommendations have suggested that high doseamoxicillin—clavulanate (90/6.4 mg/kg/day) or azi-thromycin (20 mg/kg/day) are the preferredchoices for second-line treatment of children whohave failed first-line therapy for AOM and for thosewho present with a new infection and have receivedantibiotics within the past month [5,6].

Beta-lactam antimicrobials, macrolides and par-enteral application of ceftriaxone are the mainantibiotic choices for children with AOM [7,8]. Peni-cillin resistant S. pneumoniae or beta-lactamaseproducing H. influenzae have been isolated in infec-tions including AOM with an increased frequency.Infections including AOM caused by these organismsmay lead major problems including treatment fail-ures [9]. Our report provides information aboutsusceptibility patterns of bacterial pathogens caus-ing AOM in Turkish children.

In this study, MEE of children with AOM werecollected by paracentesis and etiologic pathogensand resistance patterns were investigated. We alsocompared the efficacy of 10 days amoxicillin—cla-vulanate therapy with 3 days azithromycin therapy.

2. Methods

This prospective, single blind, randomised compara-tive study was conducted in outpatient clinics ofPediatrics and Otorhinolaryngology Departments atGaziosmanpasa University Faculty of Medicinebetween June 2002 and April 2004. The EthicalCommittee of Gaziosmanpasa University approvedthe study.

Children with complaints suggesting AOM (fever,irritability, and earache) were first examined bypediatrician, and then referred to Otorhinolaryngol-ogy specialist for otoscopic examination. Middle earfluid was revealed by tympanometry or pneumaticotoscopy. Patients who had middle ear fluid, two ormore local signs such as erythema, fullness or bul-ging of the tympanic membrane, loss of tympanicmembrane landmarks and acute perforation withpurulent otorrhea were included to our study. Thefollowing criteria were applied for the patients ofthe study:

1. A

ge ranging from 6 months to 12 years. 2. N o antibiotic treatment given within 2 weeks. 3. N o diagnosis of chronic otitis media or purulent

otorrhea for more than 24 h.

4. A bsence of allergy in story to any of the drugs

used in study.

5. A bsence of serious underlying disease that may

impair response to treatment (immunodefi-ciency, renal or hepatic insufficiency).

6. W

ritten consent from the parents.

Paracentesis was performed to all children withan intact tympanic membrane before the first doseof antibiotiotic treatment and it was performed onthe more symptomatic side if there was a bilateral

to studies of different countries. Although clinical response rates were better inpatients treated with amoxicillin—clavulanate, this was not statistically significant[86.6% (78 of 90)] versus [95.2% (80 of 84)]. Success rates of amoxicillin—clavulanatewere high for both S. pneumoniae and H. influenzae. Difference between successrates was not statistically significant (P = 0.144 and 0.352).Conclusions: Bacteria were isolated in 60% of AOM cases. The clinical efficiency ofamoxicillin—clavulanate was found to be equal compared to azithromycin in childrenwith acute otitis media.# 2005 Elsevier Ireland Ltd. All rights reserved.

Bacterial etiology of acute otitis media and clinical efficacy 917

infection. Middle ear fluid samples were collectedby paracentesis and by direct swap sample if thetympanic membrane ruptured. The collected speci-men was placed in a transport medium by a sterilecollector for the microbiologic examination.

Patients were randomised to receive either oralsuspension of azithromycin (Group 1: azithromycin10 mg/kg/24 h, peroral, once daily for 3 days) oramoxicillin—clavulunate (Group 2: amoxicillin45 mg/kg/24 h + clavulunate 6.4 mg/kg/24 h, per-oral, in two divided doses for 10 days). Three controlotoscopic examinations were done on days between2, 4 (first control visit), 11, 13 (second control visit)and 26, 28 (third control visit), respectively.Patients with persisting MEE were followed-up for3 months.

The criteria for success were defined as follows:

(1) C

linical cure: complete resolution of sign andsymptoms due to AOM.

(2) C

linical failure: failure to clear sign and symp-toms.

(3) Im

provement: incomplete resolution of AOMsigns and symptoms (persistence of serous mate-rial in the middle ear).

(4) R

elapse: after an initial period of improvement,recurrence of clinical and otoscopic findings onsecond control.

(5) R

einfection: recurrence of clinical and oto-scopic findings in a patient during the 30 daysfollow-up period in whom cure or improvementhad been detected on second control.

2.1. Microbiology

The specimens were cultured on 5% sheep bloodagar, eosine methylene blue agar, chocolate agar,sabouraud dextrose agar and thioglyconate agar(Oxoid) in aerobic conditions. All were incubatedat 35—37 8C for 24—48 h. Preparations from speci-mens were directly examined by gram staining. S.pneumoniaewas identified on the basis of sensitivityto optichin and alpha-hemolysis, cathalase andgram-positive diplococci.

Streptococcus pyogenes were defined on thebasis of basitracine sensitivity, PYR positivity, andbeta-hemolysis on blood agar and were grouped bylatex agglutination test (Plasmatec-Axis-ShieldDiagnostics, Dundee, UK). Haemophilus specieswere identified on the basis of growth in chocolateagar, requirement for factors X and V observation ofgram-negative cocco-bacilli in gram stain. Sero-types of H. influenzaewere defined by H. influenzaetype b antibody (Difco). Gram-negative cocci, oxi-dase-positive, DNAase-positive were identified asM.

catarrhalis. Gram-positive cocci, cathalase posi-tive, coagulase positive were identified as Staphy-lococcus aureus. Other organisms were classifiedaccording to classical methods and the diagnoseswere confirmed by sceptor. H. influenzae and M.catarrhalis were tested for beta-lactamase produc-tion by the chromogenic nitrocefin test.

The penicillin resistance in S. pneumoniae wasevaluated by the use of E-test (Oxoid-ET 0268,Sweden). In addition, the level of resistance of otherisolated strains to antimicrobials (including azithro-mycin and amoxicillin—clavulanate) were deter-mined by the use of oxacilline tests and diskdiffusion method as defined by NCCLS [10].

S. pneumoniae was susceptible to penicillin witha minimum inhibitory concentration (MIC) of0.06 mg/ml or less, intermediate resistant of0.25—2 mg/ml, resistant of �4 mg/ml. Antimicro-bial sensitivity of S. pneumoniae, S. pyogenes andEnterecocci strains were evaluated in 5% sheepblood Muller—Hinton agar; antimicrobial sensitivityof staphylococci, Haemophilus strains, M. catarrha-lis and Escherichia coli strains were evaluated inMuller—Hinton agar. Haemophilus strains with azone diameter azithromycin resistant, respectively;if the zone diameter was >18 and >19 mm foramoxicillin—clavulanate and azithromycin, respec-tively, the strains were considered sensitive asdefined by NCCLS [10].

Other strains with a zone diameter of <13 and<14 mm for amoxicillin—clavulanate and azithro-mycin, respectively were considered resistant,and organisms with a zone diameter >18 and>18 mm, respectively, were considered sensitiveas defined by NCCLS [10].

2.2. Statistical methods

In both group data, descriptive statistics were usedto analyse baseline demographic data. Bacteriologicand clinical responses were compared with a x2 testor Fisher’s exact test. The statistical significancewas set at P < 0.05.

3. Results

A total of 180 patients were enrolled and randomlyassigned to receive either azithromycin (n = 94) oramoxicillin—clavulanate (n = 86). After the initialexamination and MEE sample collection fourpatients did not attend to the first control visitand were excluded from the study. Two patientsdid not show up for the second control visit (days11 and 13). Totally six patients were excluded from

918 M. Guven et al.

Table 1 Demographic and clinical characteristics of the whole study group

Azithromycin group(n = 94)

Amoxicillin—clavulanategroup (n = 86)

Total(n = 180)

Gender (F/M) 24/70 18/68 42/138Age (months, mean � S.D.) 35.40 � 22.51 38.37 � 21.43 36.82 � 21.92Weight (kg, mean � S.D.) 16.70 � 5.66 17.67 � 5.38 17.20 � 5.52Height (cm, mean � S.D.) 86.06 � 16.81 87.23 � 19.95 86.62 � 18.28Culture positive cases (n) 52/94 56/86 108/180

AOM lateralizationRight 36/94 36/86 72Left 40/94 42/86 82Bilateral 18/94 8/86 26

Previous episodes of AOM 11/94 13/86 24

ComplaintsEar ache, n 64/94 42/86 106Fever, n 82/94 72/86 154Irritability, n 48/94 32/86 80

Leukocyte count mean/mm3 10580 9941 10275

study. Four of the excluded patients were in azi-thromycin group whereas two were in the amoxi-cillin—clavulanate group. One hundred and thirty-eight (77%) patients were male, 42 (23%) patientswere female; mean age was 36.82 � 21months (10—100 months). There were no significant differenceswith regard to age, gender, weight, height, time toresolution of symptoms between the study groupand group in which no etiologic pathogens wereidentified.

Right, left andbilateral ear infectionwasdetectedin 72 (40%) patients, 82 (45.5%) patients and 26(14.4%) patients, respectively (Table 1). Bacterialpathogens were isolated from the MEE in 108 (60%)of 180 children at initial visit. Fever, earache, andirritability were present in 106 (58.8%), 154 (85.5%),80 (44.4%), respectively at first visit.

The percentages of patients with clinical cure orimprovement were 36.3 and 63.7%, respectively atfirst control visit (Table 2). No treatment failure wasseen at the patients. At second control visit, thepercentages of clinical cure and improvement were79.3 and 20.7%, respectively. There was not anyrelapsing AOM. At the third control examinationthe percentages of clinical cure, improvement

Table 2 Clinical response in evaluable patients at first con

Groups First control (total = 176 cases)

Clinicalcure, n (%)

Improved,n (%)

Fan (

Group 1, n (%) 32 (35.6) 58 (64.4) —Group 2, n (%) 32 (37.2) 54 (62.8) —

Total, n (%) 72 (36.3) 112 (63.7)

and reinfection rates were 82.7, 8.3, and 9%,respectively (Table 3).

Bacteria were not isolated in 72 (40%) of 180patients. Of these 42 (58.3%) were in Group 1 and30 (41.7%) were in Group 2. At the third control visit,clinical cure rate was 100% in Group 1 for patientswith sterile culture whereas clinical cure andimprovement rates were 80 and 20% in Group 2,respectively (Table 4). Of these, neither relapsesnor recurrence was detected after the third controlvisit. There was not a statistical difference betweenthe groups with sterile culture according to clinicalsuccess rates after the third control visit (P = 0.167).

Totally 116 species were isolated from 108 (60%)of 180 patients (Fig. 1). S. pneumoniaewas the mostfrequently isolated organism (n = 46; 39.7%) fol-lowed by H. influenzae (n = 24; 20.7%), M. catar-rhalis (n = 18; 15.5%), S. aureus (n = 16; 13.8%),Group A beta-hemolytic streptococcus (n = 6;5.1%), E. coli (n = 4; 3.4%) and Enterococcus faecalis(n = 2; 1.7%). Multipl pathogens were isolated ineight patients. Four of 24 H. influenzae strains weretype b (16.7%), and had beta-lactamase activity.Fifty-five percent of the M. catarrhalis strains (10of 18) were beta-lactamase positive.

trol and second control visit

Second control (total = 174 cases)

ilure,%)

Clinicalcure, n (%)

Improved,n (%)

Failure,n (%)

70 (77.7) 20 (22.3) —68 (80.9) 16 (19.1) —

138 (79.3) 36 (20.7) —

Bacterial etiology of acute otitis media and clinical efficacy 919

Table 3 Clinical response in evaluable patients at days 26—28 (third control) evaluation

Groups Failure,n (%)

Improved,n (%)

Clinical cure,n (%)

Relaps,n (%)

Reinfection,n (%)

Total,n (%)

Group 1, n (%) — 8 (8.9) 70 (77.7) — 12 (13.3) 90 (51.7)Group 2, n (%) — 6 (7.2) 74 (88) — 4 (4.7) 84 (48.3)

Total, n (%) — 14 (8.3) 144 (82.7) — 16 (9.1) 174 (100)

Forty-two percent of the cases (n = 76) wereyounger than 24 months, of whom bacteria wasisolated in 42 (55%). Bacteria was isolated in 66(63%) of the patients older than 24 months(n = 104). Although the bacterial isolation rateswere higher in the latter group, the differencewas not statistically significant (P = 0.324). Althoughbacterial isolation rates were higher in this group,the difference was not statistically significant(P = 0.324). Most frequently isolated organisms inthe �24 months age group were as follows: S.pneumoniae (n = 16; 38%), H. influenzae (n = 10;24%), S. aureus (n = 8; 19%).

3.1. Comparisons between groups

There were no significant differences between thetwo groups according to gender, age, weight, height,frequency of ear ache, irritability and fever. Bacteriawere isolated from 52 (55.3%) patients in Group 1 and56 (65%) patients in Group 2. Table 2 shows clinicalresponse rates in first and second control visit.

Table 4 Evaluation of both groups clinical response accord

Failure,n (%)

Improved,n (%)

Clinn (%

Group 1 (azithromycin)Culture negative — — 38S. pneumoniae — — 20H. influenzae — 6 (75) 2M. catarrhalis — —Multipl pathogensa — 2Other pathogens — 2 (20) 8

Total, n (%) — 8 (8.9) 70

Group 2 (amoxicillin—clavulanate)Culture negative — 6 (20) 24S. pneumoniae — — 10H. influenzae — — 12M. catarrhalis — — 12Multipl pathogensb — — 2Other pathogens — — 14

Total, n (%) — 6 (7.2) 74

Excluded cases were four in first group (culture negative) and twoa Multipl pathogens were S. pneumoniae and H. influenzae in twb Multiple pathogens were S. pneumoniae and M. catarrhalis in t

At the third control visit (days 26—28), a total of174 patients were considered evaluable (Table 3). Asatisfactory clinical response was achieved in 95.2%(80 of 84) of amoxicillin—clavulanate patients, with88% clinical cure and 7.2% improvement rates.These responses were comparable to the distribu-tion of clinical response in the 90 evaluable azi-thromycin patients, of whom 77.7% were clinicallycured and 8.8% improved. Although the improve-ment percentages were better in Group 2, thedifference between two groups was not statisticallysignificant (P = 0.087).

The clinical response rates in both groups accord-ing to culture results are shown at Table 4. Althoughthe clinical cure rates in Group 2 were higher, nosignificant differences were noted between groupsfor clinical response rates for patients infected withS. pneumoniae singly (P = 0.144). Of the eight eva-luable azithromycin patients infected with H. influ-enzae singly, 25% were clinically cured and 75% wereimproved. All of the eight patients with singly H.influenzae infection in the amoxicillin—clavulanate

ing to culture results

ical cure,)

Relaps,n (%)

Reinfection,n (%)

Total,n (%)

(100) — — 38 (42.2)(66.7) — 10 (33.3) 30 (33.3)(33.3) — — 8 (8.9)

— —(50) — 2 (50) 4 (4.4)(80) — — 10 (11.1)

(77.8) — 12 (13.3) 90 (100)

(80) — — 30 (35.7)(100) — — 10 (11.9)(100) — — 12 (14.3)(83.3) — 2 (17.7) 14 (16.7)(50) — 2 (50) 4 (4.8)(100) — — 14 (16.7)

(88) — 4 (4.8) 84 (100)

in second group (other pathogen).o cases, and S. pneumoniae and M. catarrhalis in two cases.wo cases, and H. influenzae and S. pyogenes.

920 M. Guven et al.

Fig. 1 Culture results from tympanocentesis (n = 106). S. pneumoniae (n = 46), H. influenzae alone (n = 24), M.catarrhalis (n = 18), S. aureus (n = 16), S. pyogenes (n = 6), E. coli (n = 4), E. faecalis (n = 2).

group were clinically cured. High clinical cure ratesin Group 2 were found to be statistically significant(P = 0.024). Reinfection rates of the groups were12.8% (12 of 90) for azithromycin group and 4.7% (4of 84) for amoxicillin—clavulanate group (Table 4).Although reinfection rates of azithromycin groupwere higher than amoxicillin—clavulanate group,the difference was not statistically significant(P = 0.325).

3.2. Serous material after otitis

At the second control visit (days 11—13), persistenceof MEE was noticed in 36 patients (20 cases in Group1 and 16 cases in Group 2). It decreased to 14patients (eight in Group 1 and six in Group 2; 7/87; 8%) at the third control visit. There was nosignificant difference between two groups concern-ing serous material after otitis media (P > 0.05).MEE resolved in 10 of 14 patients at 3 months follow-up. Ventilation tubes were inserted in remainingfour patients. In eight of 14 cases, several bacteriawere isolated as follows: H. influenzae in six

Table 5

Bacteria, n (%) Bacteria resistant to ant

AZT AMC P

S. pneumoniae (46; 47.8) 10 — 22a

H. influenzae (24; 16.7) 3 — 16M. catarrhalis (18; 15.5) 2 — 10S. aureus (16; 13.8) 2 4 8S. pyogenes (6; 5.2) — — —E. coli (4; 3.4) 2 — —E. faecalis (2; 1.7) — — 2

Total resistance (%) (106; 100) 18.9 3.4 50

AZT, azithromycin; AMC, amoxicillin—clavulanate; P, penicillin; AMSCRO, ceftriaxone; CXM, cefuroxime; b-Lac, beta-lactamase activita Eight cases of intermediate resistance.

patients and E. coli in two patients. No type bwas detected in H. influenzae cases.

3.3. Side effects

Side effects related to treatment were recorded infour (4.4%) of 90 azithromycin patients (diarrhoea)and in four (4.7%) of 84 amoxicillin—clavulanatepatients (abdominal pain and skin rash). All sideeffects were mild in severity and both drugs weretolerated well.

3.4. Antibiotic resistance

Fifty-five percent of the M. catarrhalis isolates wasbeta-lactamase positive. All M. catarrhalis strainswere susceptible to amoxicillin—clavulanate; incontrast, one of the M. catarrhalis strain was azi-thromycin—resistant. Seventeen percent of the H.influenzae isolates (4 of 24) had beta-lactamaseactivity. All H. influenzae isolates were also suscep-tible to amoxicillin—clavulanate, but four strainswere resistant to azithromycin (Table 5).

ibiotics (n = 116)

AMS TMX CEF CRO CXM b-Lac

22 20 12 4 48 8 8 — — 4

10 4 4 2 2 108 4 4 4 4 —— — — — — —2 2 2 — —2 — 1 — —

44.8 36.3 27.5 8.6 8.6 12

, amoxicillin; TMX, trimetoprim/sulfometoxasol; CEF, cefaclor;y.

Bacterial etiology of acute otitis media and clinical efficacy 921

Twenty-two percent of S. pneumoniae isolates(10 of 46) were resistant to azithromycin; in con-trast, all of the S. pneumoniae isolates were sus-ceptible to amoxicillin—clavulanate. Four of the S.aureus strains with oxacillin resistance (MRSA) werealso resistant to both azithromycin and amoxicillin—clavulanate. Of the four amoxicillin—clavulanatesusceptible E. coli isolates, 2 (50%) were azithro-mycin resistant. All S. pyogenes and E. faecalisstrains isolated were susceptible to azithromycinand amoxicillin—clavulanate. Overall, 18.9% ofthe strains were azithromycin resistant and 3.4%were amoxicillin—clavulanate resistant. Ten (45%)of the pneumococcal strains and two (50%) of thestaphylococcal strains with penicillin resistancewere isolated from children �24 months of age.Four (40%) of the M. catarrhalis strains with beta-lactamase activity were detected in this group.Frequency of penicillin resistant bacteria was 50%for the whole group and 57% in the age group >24months (P = 0.424).

4. Discussion

The microbiological causes of AOM have been docu-mented on the basis of culture results of MEEobtained by tympanocentesis. It has been reportedthat bacteria and viruses are the main causes ofacute otitis media [2,4]. The most likely pathogensshow some degree of geographical variation. Resis-tance patterns change every year in a dynamicmanner as antibiotics are used more frequently[11].

S. pneumoniae, H. influenzae, and M. catarrhalishave been reported to be the most frequent etio-logic agents of AOM, respectively [12—14]. However,H. influenzae has been reported to be the mostcommon isolated bacteria in AOM [5,15]. Pichicheroand Casey [7] reported that in countries, where thenew pneumococcal conjugate vaccine and high doseamoxicillin as a first-line antibiotic treatment hadbeen widely used incidence of H. influenzae infec-tions had increased. In our study, the percentage oftypes of bacteria causing AOM is similar when com-pared with the literature for S. pneumoniae, H.influenzae, M. catarrhalis. Contrast to literatureS. aureus is the fourth etiologic agent causingAOM in our study (13.8%). In a similar study in TurkeyS. pneumoniae, H. influenzae and S. aureus (18%)were found to be most common isolated bacteria,respectively [4]. In developed countries, thereported incidence for S. aureus is somewhat lowerthan those in developing countries [11,15]. Thisdifference is also present between different pro-

vinces of Turkey. Althoughwe could not find any dataabout the incidence of E. coli at AOM in the litera-ture, we isolated E. coli in 4 (3.7%) patients. Onemust keep in mind that the gram-negative bacteriaswhich are rarely isolated in the middle ear fluidcultures could possibly be contaminated from theexternal auditory canal skin.

S. pneumoniae is the leading pathogen in AOMand even though there are some geographical dif-ferences, the reported penicillin resistance is about40% [16]. Penicillin resistance for S. pneumoniaewas reported to be 36.4% in a similar study con-ducted in Turkey [4], but the bacterial resistancehad not been evaluated with E-test and bactericidalMIC concentrations. In our study, 47.8% of pneumo-cocci were penicillin resistant. Previously, in twostudies conducted in Turkey 3.5—31% moderate andhigh resistance to penicillin have been reported in S.pneumoniae strains [17,18]. The amoxicillin resis-tance is reported between 62 and 89% and patientsyounger than 24 months of age account for themajority of the penicillin resistant cases [16,18].In developed countries as a consequence of increas-ing levels of resistance in bacterial AOM pathogens,the recommended doses for amoxicillin—clavula-nate and azithromycin are 90 mg/kg/day and20 mg/kg once a day, respectively, for high riskchildren (recurrent or persistent AOM) [5,6,22]. Inour study, most of the children did not have anyhistory of recurrent or persistent AOM. Most of thechildren in our study were not in high risk group ofAOM. In addition to low resistance rates in Turkey,recommended use of high dose amoxicillin—clavu-lanate or azithromycin for developed countries isstill considered unnecessary.

The incidence of beta-lactamase producing H.influenzae andM. catarrhalis has increased in recentyears. These beta-lactamase producing organismsconfer resistance to amoxicillin, second-generationcephalosporins, and even to trimethoprim—sulfa-methoxazole [19]. We detected that 16.7% of theH. influenzae isolates had beta-lactamase activity,and all of them were H. influenzae type b. Fifty-fivepercent of the M. catarrhalis isolates was also beta-lactamase positive. When the whole isolated patho-gens were taken together the resistance to amoxi-cillin and penicillin were 44.8 and 50%, respectively.

Block [20] has reported that azithromycin hadsome unique pharmacological features and despiteinadequate bactericidal serum levels, the concen-trations in middle ear tissues and leukocytes provideefficacy; so this macrolide agent could be one of theantibiotic choice in AOM. Dagan et al. [21] reportedthat amoxicillin—clavulanate demonstrated super-ior bacteriological efficacy against H. influenzae atdays 4—6 and superior clinical efficacy on days 12—

922 M. Guven et al.

14. However, the latter difference was no longerstatistically significant at days 22—28 [21,22].

We found higher clinical cure rate with amoxi-cillin—clavulanate when compared with azithromy-cin. There was also a trend favoring amoxicillin—clavulanate for clinical cure of S. pneumoniae andH. influenzae. The clinical success rates for H.influenzae were similar in both treatment groups.We conclude that amoxicillin—clavulanate and azi-thromycin provides satisfactory clinical efficacy inchildren with AOM, including those with infectionscaused by PRSP and beta-lactamase producing H.influenzae similar to previous studies [23—25]. Inour study the clinical success rates were found to beequal for both amoxicillin—clavulanate and azithro-mycin in contrast to previous study conducted byDagan et al. [21]. Differences between Dagan’s andour’s were as follows: in Dagan’s study culture,negative patients were excluded from the study,double tympanocentesis (repeated tympanocent-esis before medication and at second visit) wasperformed and H. influenzae was the most commonisolated pathogen. In our study S. pneumoniae waspredominant pathogen and mean patient age washigher than Dagan’s study.

Clinical efficacy of azithromycin and beta-lactamantibiotics for AOM treatment was demonstrated incase-controlled studies [4,23—26]. In our study, 10days amoxicillin—clavulanate therapy was shown tobe equal effective with 3 days azithromycin therapy[86.6% (78 of 90) versus 95.2% (80 of 84)]. All thepenicillin resistant pneumococci strains also hadazithromycin resistance. There were sixteen rein-fection cases totally (12 in azithromycin and 4 inamoxicillin—clavulanate group). In Group 1, amongthe patients with resistant S. pneumoniae infection10 patients reinfected. Difference of reinfectionrates between groups was not statistically signifi-cant. This may be explained by intermediate level ofresistance or discordance between in vitro tests andclinical efficacy.

Azithromycin and amoxicillin—clavulanate wereboth tolerated well and had equal incidence oftreatment-related adverse effects. Our results weresimilar to findings of two previous randomised stu-dies [27,28], but not in the other five [21,23—25,29].Number of cases in our study were lower than pre-vious case-controlled studies.

5. Conclusion

In conclusion, 60% of AOM cases in Turkish childrenhad bacteriologic etiology and S. pneumoniae wasmost frequently isolated pathogen, followed by H.

influenzae, S. aureus, M. catarrhalis, S. pyogenes,E. coli and E. faecalis. Of all H. influenzae strains,16.7% were type b. Resistance rates of penicilin andamoxicillin for all of the pathogens were noted as 50and 44.8%, respectively. Ten days amoxicillin—cla-vulanate treatment is equal effective with 3 daysazithromycin treatment.

Acknowledgement

Supported by a grant (Project number: 03-GEKTIP-22) from the Research Fund of Gaziosmanpasa Uni-versity.

References

[1] G.A. Zielhius, A.A. Gerritsen, W.H. Gorissen, L.J. Dekker,M.M. Rovers, G.J. van der Wilt, et al. Hearing deficits atschool age: the predictive value of otitis media in infants,Int. J. Pediatr. Otolaryngol. 44 (1998) 227—234.

[2] O. Ruuskanen, T. Heikinen, Otitis media: etiology and diag-nosis, Pediatr. Infect. Dis. J. 13 (1994) 23—26.

[3] T. Heikkinen, M. Thint, T. Chonmaitree, Prevalance of var-ious respiratory viruses in the middle ear during acute otitismedia, N. Engl. J. Med. 340 (1999) 260—264.

[4] F. Oguz, E. Unuvar, Y. Suoglu, B. Erdamar, G. Dundar, S.Katircioglu, et al. Etiology of acute otitis media in childhoodand evaluation of two different protocols of antibiotic ther-apy: 10 days cefaclor vs. 3 days azitromycin, Int. J. Pediatr.Otorhinolaryngol. 67 (2003) 43—51.

[5] R. Dagan, A. Hoberman, C. Johnson, et al. Bacteriologic andclinical efficacy of high dose amoxicillin/clavulanate inchildren with acute otitis media, Pediatr. Infect. Dis. J.20 (2001) 829—837.

[6] A. Arrieta, A. Arguedas, P. Fernandez, S.L. Block, P. Emper-anza, High-dose azithromycin versus high-dose amoxicillin—clavulanate for treatment of children with recurrent orpersistent acute otitis media, J. Antimicrob. Chemother.47 (2003) 3179—3186.

[7] M.E. Pıchıchero, J.R. Casey, Acute otitis media diseasemanagement, Mınerva Pediatrica 55 (2003) 415—438.

[8] A. Korzyrskyj, E. Hildes-Ripstein, Lougstaffe SEA, J.L. Win-cott, D.S. Sitar, T.P. Klassen, et al. Treatment of acute otitismedia with a shortened course of antibiotics, JAMA 279(1998) 1736—1742.

[9] R. Dagan, O. Abramson, E. Leibowitz, R. Lang, S. Goshen, D.Greenberg, Impaired bacteriologic response to oral cepha-losporins in acute otitis media caused by pneumococci withintermediate resistance to penicillin, Pediatr. Infect. Dis. J.15 (1996) 980—985.

[10] National Committee for Clinical Laboratory Standards, Per-formance standards for antimicrobial susceptibility tenting,Eleventh Informational Supplement, M2-A7 (M100-S11),National Committee for Clinical Laboratory Standards, MaryJane Ferraro, PA, 2001.

[11] R. Dagan, Can the choice of antibiotics for therapy of acuteotitis media be logical? Eur. J. Clin. Microbiol. Infect. Dis. 17(1998) 1—5.

[12] D. Majee, T. Harris, Acute otitis media in children, Br. Med.J. 315 (1997) 680—689.

Bacterial etiology of acute otitis media and clinical efficacy 923

[13] T. Kilpi, E. Herva, T. Kaijalainen, R. Syranen, A.K. Takala,Bacteriology of acute otitis media in a cohort of Finnishchildren followed for the first two years of life, Pediatr.Infect. Dis. J. 20 (2001) 654—662.

[14] L. Piglansky, E. Leibowitz, S. Raiz, D. Greenberg, J. Pres, A.Leiberman, et al. Bacteriologic and clinical efficacy of highdose amoxicillin for therapy of acute otitis media in chil-dren, Pediatr. Infect. Dis. J. 22 (2003) 405—413.

[15] A. Arguedas, C. Loaiza, A. Perez, F. Vargas, M. Herrera, G.Rodriguez, et al. Microbiology of acute otitis media in CostaRikan children, Pediatr. Infect. Dis. J. 17 (1998) 680—689.

[16] M.R. Jacobs, R. Dagan, P.C. Appelbaum, D. Burch, Preva-lance of antimicrobial-resistant pathogens in middle earfluid: multinational study of 917 children with acute otitismedia, Antimicrob. Agent Chemother. 42 (1998) 589—595.

[17] B. Sener, A. Gunalp, Trends in antimicrobial resistance ofStreptococcus pneumoniae in children in a Turkish hospital,J. Antimicrob. Chemother. 42 (1998) 381—384.

[18] D. Gur, F. Tunckanat, B. Sener, G. Kanra, H.E. Akalin, Peni-cillin resistance in Streptococcus pneumoniae in Turkey, Eur.J. Clin. Microbiol. Infect. Dis. 13 (1994) 440—441.

[19] S.L. Block, Strategies for dealing with amoxicillin failure inacute otitis media, Arch. Fam. Med. 8 (1999) 68—76.

[20] S.L. Block, Causative pathogens, antibiotic resistance andtherapeutic considerations in acute otitis media, Pediatr.Infect. Dis. J. 16 (1997) 449—456.

[21] R. Dagan, C.E. Johnson, S. McLinn, N. Abugdali, J. Feris, E.Leibowitz, et al. Bacteriologic and clinical efficacy ofamoxicillin/clavulanate vs. azithromycin in acute otitismedia, Pediatr. Infect. Dis. J. 19 (2000) 95—104.

[22] R. Dagan, G.H. McCracken, Flaws in design and conduct ofclinical trials in acute otitis media, Pediatr. Infect. Dis. J. 21(2002) 894—902.

[23] C.M. Khurana, A multicenter, randomised, open label com-parison of azithromycin and amoxicillin-clavulanate in acuteotitis media among children attending day care or school,Pediatr. Infect. Dis. J. 15 (1996) 24—29.

[24] G. Aronovitz, A multicenter, open label trial of azithromycinvs. amoxicillin clavulanate for the management of acuteotitis media in children, Pediatr. Infect. Dis. J. 15 (1996)15—19.

[25] S. Mclinn, A multicenter, double blind comparison of azi-thromycin and amoxicillin clavulanate for the treatment ofacute otitis media in children, Pediatr. Infect. Dis. J. 15(1996) 20—23.

[26] E. Mosh, A. Rodriquez-Solares, A. Rivas, Z. El Hoshy, Acomparative study of azithromycin and amoxicillin in pae-diatric patients with acute otitis media, J. Antimicrob.Chemother. 31 (1993) 73—79.

[27] R.R. Daniel, Comparison of azithromycin and co-amoxilav inthe treatment of otitis media in children, J. Antimicrob.Chemother. 31 (1993) 65—71.

[28] N. Princibi, Multicentre comparative study of efficacy andsafety of azithromycin comparedwith amoxicillin—clavulanicacid in the treatment of paediatric patientswith otitismedia,Eur. J. Clin. Microbiol. Infect. Dis. 14 (1995) 669—676.

[29] U.B. Schaad, Multicentre evaluation of azithromycin incomparison with co-amoxiclav for the treatment of acuteotitis media in children, J. Antimicrob. Chemother. 31(1993) 81—88.