International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–1216

Review article

Acute otitis media: From diagnosis to prevention. Summary of the Italianguideline

Paola Marchisio a,1,*, Luisa Bellussi b,1, Giuseppe Di Mauro c,1, Mattia Doria d,1, Giovanni Felisati e,1,Riccardo Longhi f,1, Andrea Novelli g,1, Annamaria Speciale h,1, Nicola Mansi i,1, Nicola Principi a,1

a Department of Maternal and Pediatric Sciences, University of Milan and Foundation, IRCCS Ca Granda Ospedale Maggiore Policlinico, Italyb Italian Society of Pediatric Otolaryngology, Siena, Italyc Italian Society for Preventive and Social Pediatrics, Caserta, Italyd Primary Care Pediatrician, Venice, Italye Department of Otolaryngology, University of Milan, Italyf Italian Society of Pediatrics, Como, Italyg Department of Pharmacology, University of Florence, Italyh Department of Microbiology, University of Catania, Italyi Italian Society of Pediatric Otolaryngology, Naples, Italy

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210

1.1. Reasons for an Italian guideline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210

1.2. Objective of the guideline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210

1.3. Development and implementation of the guideline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210

2. Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1211

2.1. Certain diagnosis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1211

2.2. Diagnostic instruments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1211

2.3. Ear wax removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1211

3. Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212

3.1. Earache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212

3.2. Selection of patients to be immediately treated with antibiotics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212

3.3. Choice of antimicrobial drug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212

A R T I C L E I N F O

Article history:

Received 6 July 2010

Received in revised form 17 August 2010

Accepted 18 August 2010

Available online 16 September 2010

Key words:

Acute otitis media

Otitis media

Antibiotic treatment

Watchful waiting

Vaccines

Prevention

A B S T R A C T

Acute otitis media (AOM) is the most common disease occurring in infants and children and has major

medical, social and economic effects. If we consider the Italian pediatric population and the incidence

rates in different age ranges it can be calculated that almost one million cases of AOM are diagnosed in

Italy every year. Various attempts have been made internationally to clarify the most appropriate ways

in which AOM should be managed. In Italy, this has been done at local or regional level but there have so

far been no national initiatives. The objective of this guideline is to provide recommendations to

pediatricians, general practitioners and otolaryngologists involved in the clinical management of acute

otitis media in healthy children aged 2 months to 12 years. After a systematic review and grading of

evidences from the literature, the document was drafted by a multidisciplinary panel with identified key

clinical questions related to diagnosis, treatment of the acute episode, management of complications and

prevention.

� 2010 Elsevier Ireland Ltd. All rights reserved.

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology

journa l homepage: www.e lsev ier .com/ locate / i jpor l

* Corresponding author. Tel.: +39 0255032690.

E-mail address: paola.marchisio@unimi.it (P. Marchisio).1 On behalf of the Italian AOM Guideline Multidisciplinary Working Group. See Appendix A.

0165-5876/$ – see front matter � 2010 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.ijporl.2010.08.016

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–12161210

3.4. Duration of antibiotic therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

3.5. Therapeutic failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

3.6. Long-term benefits of antibiotic treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

3.7. Usefulness of other treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

4. Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

5. Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214

5.1. Reducing risk factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214

5.2. Influenza vaccines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214

5.3. Pneumococcal vaccine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215

1. Introduction

Acute otitis media (AOM) is the most common bacterial diseaseoccurring in infants and children: almost all children experience atleast one episode, and about one-third experience two or moreepisodes in their first 3 years of life [1,2]. The European data, whichalso include a sample of Italian subjects, indicate that the annualincidence of AOM in children is 268 episodes per 1000 children [3].If we consider the Italian pediatric population and the incidencerates in different age ranges, it can be calculated that almost onemillion cases of AOM are diagnosed in Italy every year.

The disease has major medical, social and economic effects [4].It always requires considerable clinical and financial implicationsas it involves at least one examination by a doctor, and theprescription of antipyretics and, depending on the country,antibiotics (more than 80% of cases in Italy and about 40% inGermany) [5,6]. Furthermore, it can cause major complicationssuch as mastoiditis. Indirect costs are just as high and often higher,mainly because of the loss of working days by parents [7,8]. Finally,its acute symptoms and frequent recurrences mean that AOM has aconsiderable impact on the quality of life of children and theirfamilies.

The way in which AOM is approached has greatly changed overrecent years because of the availability of new instruments thatprovide a more precise diagnosis, studies that have better definedthe efficacy of antibiotic therapy and the selection of the patients tobe treated, and vaccines that have undeniable preventive efficacy.However, acceptance of these variations by healthcare workers hasbeen slower, and this means that there is a need to translate thenewly acquired knowledge into precise and scientifically correctrecommendations that combine the latest findings with whatshould be preserved from the past.

1.1. Reasons for an Italian guideline

Various attempts have been made nationally and internation-ally to clarify the most appropriate ways in which AOM should bemanaged (see Appendix B). Large differences exist with respect todiagnostic criteria and diagnostic methods used (e.g. pneumaticotoscopy, tympanometry and otoscopy). In most guidelinesantibiotic are indicated for children with AOM below the age of6 months. For older children (6–24 months) indications differbetween countries: some guidelines recommend antibiotics in allchildren with AOM, whereas others recommend to prescribeantibiotics according to the presence of effusion, fever, duration ofcomplaints and comorbidity. In most guidelines amoxicillin is thefirst choice [9,10]. No international consensus exist. In Italy, theproblem of the management of AOM has been faced at local orregional level with a large variety of recommendations, without anofficial endorsement by any Scientific society. There was on overallconsensus of pediatricians that the 2004 AAP and AAFP ClinicalPractice guideline [11] could be somewhat applicable for Italianpatients, but this attitude was not shared by otolaryngologists. In

addition, the 2004 AAP guideline has not been updated and, inmeanwhile, new data literature has become available. Finally,peculiarities of microbiological data exclude the immediatetransfer of other European guidelines to the Italian reality. Thegap was filled by the Italian Society of Pediatrics, driven by theItalian Preventive Pediatrics Society and the Italian Society ofPediatric Otolaryngology, which coordinated the preparation ofthe Italian National Multidisciplinary guideline. The participationof both pediatricians and otolaryngologists in the same projectwould not have been possible until a few years ago, but it wasagreed that this strategy was indispensable, as AOM is encounteredby both types of specialists, and their use of different diagnosticand therapeutic approaches may cause confusion and prevent theadoption of a common clinical language, thus hampering anoptimal multidisciplinary approach. The synthesis of the guidelineis here reported. It includes the most relevant recommendations(highlighted by quotation marks) for the management of AOM, andbrief comments concentrating on differences with internationalguidelines. The entire document, in Italian, and full references, isaccessible at the websites of the Italian Society of PediatricOtolaryngology (www.siop.it) and of the Italian Society ofPediatrics (www.sip.it).

1.2. Objective of the guideline

The Italian guideline examines only the management of AOM inchildren, whereas the management of recurrent acute otitis media,otitis media with effusion or chronic suppurative otitis media areintentionally excluded [12]. The guideline deals only with childrenaged 2 months to 12 years, excluding subjects with acquired orcongenital immunodepression, spontaneous chronic perforation ofthe eardrum or tube, an underlying chronic disease favouringnasopharyngeal colonisation by unusual pathogens (e.g. cysticfibrosis), or cranio-facial malformations. The neonatal period(extended to the first 2 months of life) is also excluded because thedisease is rare during this period, its etiology is somewhat differentfrom that commonly observed in children, and there are fewpublished studies.

1.3. Development and implementation of the guideline

This document was created in accordance with the recom-mendations made in the Manual for Writing Clinical PracticeGuidelines of the Programma Nazionale Linee Guida (PNLG: theItalian National Guideline Programme) [13]. The above mentionedScientific Societies convened a multidisciplinary group of expertsfrom all pertinent areas (including representatives of laypeopleand consumers), which identified key clinical questions related to(a) diagnosis, (b) treatment, (c) complications and (d) prevention ofAOM, concentrating on those about which there was the greatestuncertainty. The literature search included systematic reviews andprotocols, either developed by the Cochrane collaboration or not,existing guidelines and other documents providing evidences of

Table 1Levels of evidence.

Level of evidence Criteria

I Evidence from more than one randomised controlled clinical trial and/or from systematic reviews of randomised trials

II Evidence from a single randomised, controlled and soundly designed clinical trial

III Evidence from cohort studies with concurrent or historical control groups

IV Evidence from retrospective studies, such as case-control studies or their meta-analysis

V Evidence from case series with no control group

VI Evidence based on the opinions of renowned experts or expert committees, as indicated in guidelines or by

consensus conferences

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–1216 1211

the effectiveness of treatment, randomised controlled trials andother types of primary studies on clinical issues not covered in thesystematic reviews. The working group decided to consider the2004 Guideline of the American Academy of Pediatrics (AAP) anoptimal model and starting point. The search covered the periodJanuary 2003 to October 2009, with no language restrictions. Of theapproximately 700 references initially identified and shared in aprotected web site, data were extracted and summarised for 285studies. The published evidence was then reviewed and assessedusing the grading system adopted by the PNLG (see Tables 1 and 2).In case of conflict between the assessment of the two subspecia-lists (pediatricians versus otolaryngologists), it was decided toemphasize the differences, and make them live together instead ofhaving one position dominating the other one. The draft documentwas discussed with external reviewers, including representativesof professional association and then finalised. The document wasdeveloped from June 2009 to February 2010. Disseminationimplied multiple techniques, including mailing, publications,symposia, educational courses, internet and opinion leadersinvolvement. The impact on practice will be evaluated at countrylevel assessing the trend in the incidence of AOM, in variations ofantibiotic use and of inappropriate hospital admissions.

2. Diagnosis

2.1. Certain diagnosis

The panel stated that a correct diagnosis of AOM is essential inorder to avoid useless, unjustified, costly and potentially harmfultherapeutic procedures. Experts agreed that all the efforts of thepractitioners, either pediatricians or otolaryngologists, should bedevoted to a certain diagnosis of AOM. The recommendation was:‘‘The diagnosis of AOM is certain only when the following can be

simultaneously demonstrated: (1) acute, recent onset of symptoms;

(2) signs of inflammation in the tympanic membrane; (3) the

presence of middle ear effusion [I/A]’’. In addition, the guidelinereminded that the demonstration of tympanic membraneinflammation and middle ear effusion should be based on: (a)otoscopic findings of marked erythema of the tympanic mem-brane, with bulging and absence mobility due to the presence ofmiddle ear effusion or (b) otoscopic finding of a yellowishmembrane by observing in transparency the presence of purulentmaterial in the middle ear or (c) the presence of spontaneous

Table 2Strength of the recommendations.

Level of strength Criteria

A The procedure or diagnostic test is strongly reco

evidence is not necessarily I or II

B There are some doubts as to whether a particula

always be carefully examined

C There is substantial uncertainty in favour of or a

D The procedure is not recommended

E There is strong advice against the procedure

perforation with otorrhea. Bulging was the clinical sign consid-ered optimal for detecting middle ear effusion as, alone, has theclosest correlation with bacterial AOM confirmed by tympano-centesis [11,14]. Redness alone of the tympanic membrane wasconsidered insufficient for diagnosis.

2.2. Diagnostic instruments

The panel decided that ‘‘the use of a pneumatic otoscope was the

most simple and efficient means of supporting a diagnosis of AOM [II/

B]’’. It also made clear that ‘‘the otoscopic examination should lead to

the identification and description of all of the features of the tympanic

membrane on both sides [II/B]’’ and thus the pediatrician shouldalways detail otoscopic findings in child’s records [15]. In uncertaincases, and in the absence of a pneumatic otoscope, pediatriciansare allowed to use a static otoscope in combination with atympanometer or reflectometer. In alternative, the pediatricianshould refer the patient to an ENT who can use otomicroscopy and/or otoendoscopy in addition to the previously mentionedinstruments [II/A].

Pneumatic otoscope is indicated as the optimal instrument byseveral guidelines (Canada, England, Finland, France, Germany,Israel, Scotland, Sweden, USA) but so far it has been not beenlargely used in Italy [16]. However the multidisciplinary panel,having in mind the certain diagnosis, thought that the instrument,beyond the pneumatic phase, was the best aid in diagnosing AOM.The guideline highlighted the importance of often underestimatedcharacteristics of an optimal otoscope such as an appropriate lightsource (with batteries that should be periodically replaced) andequipment with non-coloured speculum of different sizes in orderto avoid light dispersion and allow the speculum to be selected onthe basis of the size of the auditory canal [17].

2.3. Ear wax removal

In contrast to the majority of guidelines, which neglect theimportance of earwax, the panel decided to emphasize that ‘‘the

external canal has to be completely free and the tympanic membrane

fully visible’’. In order to overcome possible conflicts betweenspecialists and, at the same time, to value the different capabilities,the experts recommended that ‘‘Ear wax should be removed by the

pediatrician or an otolaryngologist depending on the methodological

difficulties and peculiarities of local settings; any other obstructions of

mmended and supported by good quality scientific trials, although the level of

r procedure or intervention should always be recommended, but its use should

gainst the procedure or intervention

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–12161212

the external auditory canal should only be removed by an

otolaryngologist [I/A]’’.However, no particular method was suggested, as no system of

ear eax removal (irrigation, eardrops or manual removal) has yetbeen shown to be superior to others [18,19].

3. Treatment

3.1. Earache

The panel accorded that the first aim of treatment is to limitsymptoms, of which pain is the most clinically relevant [11]. Therecommendations were: ‘‘The main treatment of earache is the

systemic administration of appropriate doses of analgesics (paraceta-

mol or ibuprofen) [I/A]; the use of topical analgesic formulations as the

only means of treating pain is not advised. Local anesthetics should be

systematically used in association with systemic analgesic therapy

only in children aged more than three years of age and only in the

documented absence of perforation [II/A]’’. It was suggested that ifthe earache is associated with fever, oral paracetamol or ibuprofenis sufficient to relieve pain at doses that control the fever.Considered the limited evidence of efficacy, a statement was addedagainst the use of natural extracts.

3.2. Selection of patients to be immediately treated with antibiotics

The panel agreed that the fact that a very high number of casesof AOM resolve spontaneously advises against the universalantibiotic treatment of all children affected by AOM. Moreover,the panel has considered that several studies have demonstratedthat watchful waiting has been shown to be able to reduce the needfor antibiotic prescription, thus reducing medical and treatmentcosts, limiting side effects secondary to treatment, and reducingselective pressure on the saprophytic flora, thus lowering the riskof the emergence of new drug-resistant strains [20–22]. On theother hand, contrasting data indicate that immediate treatmentcan have a significant beneficial effect in terms of pain durationand intensity, analgesic doses and medical visits, even taking intoaccount the increased risk of side effects related to antibioticadministration. Based on systematic reviews [23,24] which agreethat immediate antibiotic treatment is advantageous in the shortterm in patients with certain AOM aged less than 2 years, in thosewith a bilateral episode, and in those with spontaneous otorrhea,the recommendation was ‘‘Antibiotic treatment should be prescribed

immediately for severe cases of AOM, in children aged less than two

years with bilateral AOM, and in the case of spontaneous perforation.

In all other cases, and in agreement with the parents, it is possible to

wait watchfully and prescribe antibiotic treatment only if the episode

worsens or does not improve within 48–72 hours [I/A]’’ (Table 3). Thestatement differs from the 2004 AAP and AAFP clinical practiceguideline as it includes and emphasized the assessment oflaterality of the episode, as children with bilateral AOM are moreprone to a negative outcome [25], and excludes definitively all the

Table 3Treatment strategies for uncomplicated AOMa.

Diagnosis Certain

Bilateralb

Severec Mildc

Age <6 months Immediate antibiotics Immediate anti

Age 6–24 months Immediate antibiotics Immediate anti

Age >24 months Immediate antibiotics Watchful waitin

a Absence of otorrhea, intracranial complications or a history of recurrences.b Laterality.c Severity.

episodes of AOM diagnosed as uncertain, in order to avoidantibiotic misuse.

A limitation of our guideline is that there was no consensus onthe type of score to be used to grade the severity of the episode. Thescores used vary from one study to another [26–28] and most ofthe data validating the watchful waiting approach have come fromchildren suffering mild episodes. However, as it is known that theepisodes associated with a high fever and substantial earache mostfrequently have a negative outcome, the panel thought to besufficient the presence of these symptoms to consider severe acertain AOMt.

A special attention was given to the fact that the successful andriskless application of watchful waiting requires the activeparticipation of parents/caregivers, who need to be adequatelyinformed concerning the potential risks and benefits of treatmentand the management of disease outcome. It was emphasized thatparents/caregivers have to be made to feel part of the decision-making process and patient control during the disease, be able tocommunicate promptly with a physician and, if necessary, haveaccess to a clinical examination.

3.3. Choice of antimicrobial drug

The panel stated that the choice of the optimal drug for thetreatment of AOM should be primarily based on microbiologicalconsiderations: i.e. an analysis of the most frequent causativebacterial pathogens and their susceptibility to the most widelyused drugs. The ideal situation, the one in which it is possible toidentify the otopathogens responsible for the infection and theantibiotic treatment is based on susceptibility results, rarelyhappens in clinical practice: when the eardrum is ruptured and it ispossible to collect middle ear effusion within 12–24 h, or whentympanocentesis is possible. The panel recognized that theantibiotic treatment of AOM is mainly empirical, and based on aknowledge of the most common bacterial pathogens in a givengeographical area and their antibiotic susceptibility. Streptococcus

pneumoniae (S. p.), nontypable Haemophilus influenzae (NTHi),Moraxella catarrhalis (M. c.) and Streptococcus pyogenes (S. pys.) arethe most common causes of AOM in almost any geographical areaand age range [29]. Italy has specific microbiologic characteristicswhich influence the choice of the antibiotic: (a) resistance to S. p.penicillin is still limited to 20%, with only half of the resistantstrains having MIC values of more than 4 mcg/mL, whereas there is30% to 50% resistance to macrolides and azalides, (b) NTHi ischaracterised by 20–40% aminopenicillin resistance due to theproduction of b-lactamases, (c) M. c. is quite rare in Italian childrenbut penicillin and aminopenicillin resistance reaches 80%, (d) S.

pys. is responsible for a quite relevant percentage of AOM, is stillhighly susceptible to beta-lactam drugs, but has developedresistance to macrolides and azalides in about 40% of cases [30,31].

On the basis of the pharmacokinetic characteristics of thevarious antibiotics and their MICs for the potential otopathogens,the panel decided that the usual 40–50 mg/kg/day dosage of

Monolateralb

Severec Mildc

biotics Immediate antibiotics Immediate antibiotics

biotics Immediate antibiotics Watchful waiting

g Watchful waiting Watchful waiting

Table 4Antibiotic choices for AOM.

Episode characteristics Recommended Alternative

Mild symptoms

No otorrhea

No recent recurrences

No risk factorsa

Amoxicillin (50 mg/kg/day, 2–3 doses) Cefaclor (40–50 mg/kg/day, 2 doses)

Severe symptoms

Otorrhea

Recent recurrences

Risk factorsa

Amoxicillin + clavulanic acid

(80–90b mg/kg/day, 2–3 doses)

Cefpodoxime proxetil (8 mg/kg/day, 2 doses),

cefuroxime axetil (30 mg/kg/day, 2 doses)

a Risk factors for bacterial resistance: age <3 years, day-care attendance, older siblings, recent antibiotic therapy (<1 month), no PCV-7.b Amoxicillin referred dosage.

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–1216 1213

amoxicillin maintain useful concentrations for long enough toeradicate the pathogen in the case of infections due to penicillin-susceptible S. p. or S. p. with intermediate susceptibility. Thusamoxicillin was indicated as the first choice drug in the case ofAOM in the absence of complications in children at low risk ofresistant pathogens (those aged more than 3 years who are notattending a day-care centres, have not received antibiotics in theprevious 30 days, do not have older siblings attending day-care,and have not been vaccinated with PCV-7 in the first year of life)and without a recent history of recurrent AOM. As alternative toamoxicillin, and in contrast to the 2004 AAP guidelines, the panedindicated cefaclor, which is characterised by good palatability andbioavailability [32]. In all other cases, amoxicillin + clavulanic acidwere recommended as the first choice drug, and, as alternatives,cefpodoxime proxetil and cefuroxime axetil were suggested (Table4). Macrolides and azalides were considered unsuitable choices.

3.4. Duration of antibiotic therapy

AOM treatment is traditionally recommended for 10 days,although some national guidelines (Finland, Germany, Sweden,Scotland, Australia) recommend short 5-day regimens and others(The Netherlands or South Africa) recommend 7 days. In othercountries (Canada, France, USA and Luxembourg), the durationmay be differentiated on the basis of patient age: 10 days forchildren aged less than 2 (or 5) years, and 5 days for children agedmore than 2 (or 5) years.

The panel consented that among the oral beta-lactams,controlled randomised clinical studies demonstrating the possi-bility of reducing AOM therapy to 5 days are available only foramoxicillin + clavulanic acid, cefpodoxime proxetil, cefuroximeaxetil and cefaclor. Moreover, as it has been shown that an age ofless than 2 years, the presence of otorrhea due to spontaneousperforation of the tympanic membrane, and a history of recurrentAOM episodes increase the risk of therapeutic failure whentreatment is shortened to 5 days [33], the final recommendationwas ‘‘Antibiotic therapy should be administered for 10 days. This can

be reduced to five days in children aged more than two years without

risking a negative outcome [I/B]’’.

3.5. Therapeutic failure

While considering that the clinical course of antibiotic-treated AOM is usually characterised by an improvement ofsymptoms within 48–72 h, with a progressive reduction in feverand pain, it was stated that ‘‘After 72 hours of adequate antibiotic,

the children with AOM episodes who do not show any improvement

or who worsen should be considered therapeutic failures. If they

were treated with amoxicillin or cefaclor, they should receive

amoxicillin plus clavulanic acid or cefpodoxime proxetil or

cefuroxime axetil. If they were being treated with a broad-spectrum

antibiotic, they should be prescribed intramuscular or intravenous

ceftriaxone [II/B]. The use of quinolones should be avoided in

treatment failures [IV/E]’’.

3.6. Long-term benefits of antibiotic treatment

The panel stated that immediate antibiotic therapy does notprevent the development of otitis media with effusion, and doesnot reduce the persistence of middle ear effusion [34]. Similarly, itdoes not seem to reduce the risk of recurrent episodes (althoughthe data are less robust). As regards complications, there was anagreement on the fact that even if the incidence of complicationsand complication-related mortality were both significantly re-duced after the introduction of antibiotic treatment, the correla-tion between the incidence of acute mastoiditis and antibioticprescription remains debatable because, in some countries (suchas Australia and the UK), a significant reduction in antibioticprescriptions was not associated with any increase in thefrequency of complications and thus ‘‘the universal use of antibiotics

in every case of AOM cannot be considered a valid means of reducing

the risk of mastoiditis [IV/D]’’ [35,36].

3.7. Usefulness of other treatments

As regards possible adjunctive treatments the panel agreed that‘‘The use of other treatments (excluding analgesics) in association with

antibiotics is not recommended [VI/D]; the use of systemic or topical

decongestants should be avoided [I/D] and the administration of

steroids or antihistamines is not recommended [II/D]’’. This statementcontrasts with other guidelines, which usually do not judge therole of other treatments. In addition, the group highlighted theimportance of removal of nasal secretions by means of nasalwashings as a useful complementary treatment [VI/B]. Theusefulness of nasal washing is still limited by the absolute paucityof scientific evidence, particularly in children and relating tomiddle ear diseases. However, given the close inter-relationshipbetween the nose, nasopharynx and middle ear, it was hypothe-sised that the removal of nasal secretions by means of salineirrigation may contribute to improving Eustachian tube functionand clearing middle ear effusion, as has already been demonstratedin the case of rhinosinusitis and rhinitis.

Lastly, the panel decided to discourage treatments belonging tocomplementary and alternative medicine [VI/D].

4. Complications

A special section of the Italian guidelines was devoted tocomplications of AOM, which are neglected by most of the otherguidelines. As regards acute mastoiditis, the recommendation wasto base the primary diagnosis on clinical examination and toemphasize the need for close monitoring of the patients. Therecommendation was ‘‘The diagnosis of acute mastoiditis is based on

clinical criteria. CT scan of the mastoid area is helpful in evaluating the

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–12161214

extension of the process and the presence of complications, and should

be systematically performed when acute mastoiditis with periosteitis

or intracranial complications are suspected [IV/A]. Medical treatment

is indicated in all non-complicated cases, whereas medical/surgical

treatment is required in the presence of mastoid empyema and/or

intracranial complications [IV/A]’’.

5. Prevention

In this multidisciplinary group the role of prevention of AOM, interms of avoiding the first episode in otherwise healthy children,was recognized as one of the primary goals of pediatric care. Thus aspecial section was devoted to this issue, in contrast with 2004 AAPand other guidelines which give a limited or no space toprevention. As AOM is favoured by a wide range of predisposingfactors, and usually follows a viral infection of the upperrespiratory tract, attempts to prevent it mainly rely on reducingrisk factors, viral respiratory infections, and nasopharyngealbacterial colonisation.

5.1. Reducing risk factors

Reviewing the literature, it was clear that (a) home-care canavoid one in five AOM episodes in the general pediatric population,and two in five in children with middle ear diseases, (b) carefulhandwashing and the use of alcoholic solutions in pre-schoolersreduces AOM episodes by 27%, (c) exclusive breastfeeding for atleast 3 months reduces the incidence of AOM by 13%, and 6months’ breastfeeding increases this to 50%, thus providingprotective cover for the entire first year of life and (d) a 30%increased risk of AOM has been found in children who use a pacifiercontinuously, and a 29% decrease when its use is limited to thetime immediately before falls asleep [37,38]. Starting from thesepremises, the panel recommended that ‘‘infants are breastfed for at

least three months [V/B], attend day-care centres at which accurate

hygiene measures are practised [III/B], reduce the use of pacifier to a

minimum [III/A], and are not exposed to passive smoking [VI/B]’’.

5.2. Influenza vaccines

The multidisciplinary group agreed that influenza vaccine reducesthe incidence of AOM when the virus is circulating among healthychildren. However, the results vary depending on vaccine type andthe child’s age: a vaccine containing attenuated live viruses leadsto better results (with a reduction of up to 90% in children aged lessthan 18 months) than an inactivated influenza vaccine (a reductionof up to 50%), especially in children aged less than 18 months [39].The experts thought that if the only goal of influenza vaccinationwere AOM prevention, vaccine administration would only bejustified in children with a high frequency of recurrent episodes.However, as influenza prevention has the more important aim ofpreventing a disease whose complications are often more severethan AOM, doubts about the limited efficacy of influenzavaccination in preventing a first episode of AOM in young childrenare largely outweighed by its other medical, social and economicadvantages. Therefore, the recommendation was ‘‘Influenza vaccine

has to be encouraged because, beyond other more important

advantages, it can be useful in preventing a first AOM episode [I/A]’’.

5.3. Pneumococcal vaccine

The three pneumococcal vaccines that are known to beimmunogenic are different in terms of the carrier proteins usedto conjugate the capsular polysaccharides and/or the number ofserotypes included in the formulation (7, 10 and 13). There are

robust data concerning the efficacy of PCV-7 and the impact of itsuniversal use on the incidence of ambulatory AOM visits andantibiotic prescriptions [40]. When administered within the firstyear of live, PCV-7 prevents 6–7% of all AOM episodes, more than30% of pneumococcal episodes, and more than 50% of the episodescaused by the serotypes it contains. The impact of PCV-7 on theprevalence of AOM in the pediatric population has been studied inthe USA, where the vaccine has been universally administeredsince 2000: the group acknowledged that comparisons of the dataobtained before and after its introduction show that the significantreduction in the number of AOM-related visits (average 20%, range4–43%) and antibiotic prescriptions is greater than its efficacy rates[6]. It was also noted that there is still a lack of data concerning theeffect of PCV-13 on AOM, but the efficacy data so far produced for a10-valent vaccine (coming from studies of its unmarketedexperimental 11-component precursor vaccine) indicate a 33.6%reduction in all AOM episodes, a 57.6% reduction in the episodescaused by pneumococcal serotypes, and a 35.3% reduction in thosedue to nontypable H. influenzae [41]. The final recommendationwas ‘‘Pneumococcal conjugate vaccine can significantly reduce the

incidence and burden of AOM [I/A]’’.

Appendix A

Italian AOM Guideline Multidisciplinary Working Group: A.

Affinita (Movimento Italiano Genitori, MOIGE); L. Bellussi* (Societa

Italiana di Otorinolaringoiatria Pediatrica, SIOP); G. Conforti (Feder-

azione Italiana Medici Pediatri, FIMP and Societa Italiana di Cure

Primarie Pediatriche, SICuPP); D. Cuda (Societa Italiana di Otorino-

laringoiatria, SIO); E. Cunsolo (SIOP); G. de Vincentiis (SIO); MC. Diana

(Societa Italiana di Medicina Emergenza e Urgenza Pediatrica,

SIMEUP); G. di Mauro (SIPPS, FIMP); P. Di Pietro (SIP; SIMEUP); M.

Doria (FIMP, Associazione Culturale Pediatri, ACP); E. Dusi (pediatri-

cian, responsible for data retrieval); G. Felisati (SIO); F. Festini

(Nursing Society); E. Genovese (SIOP); R. Longhi (SIP); N. Mansi

(SIOP); P. Marchisio* (SIP, Italian Society of Pediatric Infectivology,

SITIP); L. Mariniello (FIMP); G. Mele (FIMP); M. Miraglia del Giudice

(Italian Society of Pediatric Respiratory Diseases, SIMRI and Italian

Society of Pediatric Allergology and Immunology, SIAIP); G. Muttillo

(IPASVI); C. Navone (SIP); G. Nicoletti (Italian Society of Microbiology,

SIM); A. Novelli (Italian Society of Chemotherapy, SIC); F. Paravati

(FIMP); M. Piemonte (SIO); L. Pignataro (SIO); P. Pisani (SIO); N.

Principi* (SITIP); D. Radzik (FIMP); S. Renna (SIP); F. Scaglione (SIC); A.

Speciale (SIM); G. Vitali-Rosati (FIMP, SICuPP). *, coordinators

Appendix B. Guidelines on acute otitis media

� Australia. Morris P, Ballinger D. Leach a et al. Recommendationsfor clinical care guidelines on the management of otitis media inaboriginal and Torres Strait Islander populations. Canberra:Office for aboriginal and Torres Strait Islander Health. Common-wealth Department of Health and Aged Care. ACT 2001� Australia. Antibiotic expert Group. Therapeutic guidelines.

Antibiotic. Version 13. Melbourne: Therapeutic Guidelineslimited. 2006� Belgium. Chevalier P, Van Lierde S, Janssens-de Varebeke S.

Acute middenoorontsteking. Belgian Antibiotic Policy Coordina-tion Committee 2001. (accessibile: http://www.health.fgov.be/antibiotics)� Canada (Alberta). Forgie S, Zhanel G, Robinson J. Management of

acute otitis media. Paediatr Child Health 2009; 14 (7): 457–460.available from: http://www.topalbertadoctors.org

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–1216 1215

� Canada (BC). Acute otitis media guideline (revised 2004).Available from: http://www.healthservices.gov.bc.ca/msp/protoguides� Denmark. Blegvad S, Ehlers P, Jakobsen M, et al. Øvre

luftvejsinfektion. State of art. Arhus County 2005. (availablefrom: http://www.sundhed.dk/praksis)� Finland. Puhakka H, Hagman E, Heikkinen T, et al. Akillisen

valikorvatulehduksen hoitosuositus. Duodecin. 1999; 115:2151–2157. from: http://www.kaypahoito.fi. [English summary:Mikstra. Acute otitis media. 2000. (available from: http://www.thl.fi)].� France. Antibiotherapie par voie generale en pratique courante:

Otite Moyenne Aigue. Agence Francaise de Securite Sanitaire desProduits de Sante. Recommendations francaises. 2001. Accessi-bile da: http://www.afssaps.sante.fr. [English summary: AgenceFrancaise de Securite Sanitaire des Produits de Sante. Systemicantibiotic treatment in upper and lower respiratory tractinfections: official French guidelines. Clin Microb Infect 2003;9: 1162–1178].� Germany. Koneczny N, Schmidt-Troschke S, Berger T, et al. Akute

Otitis media (AOM) bei Kindern (ICD H66.0): Eine evidenzba-sierte Leitlinie. Klinisch Padiatrie 2004; 216: 215–224� Institute for Clinical System Improvements. Diagnosis and

Treatment of otitis media in children. January 2008. (accessibile:http://www.icsi.org)� Netherlands. Damoiseaux RAMJ, Van Balen FAM, Leenheer WAM,

Kolnaar BGM. Huisarts Wet NHG-Standaard Otitis media acuta(Tweede herziening). Huisarts Wet 2006; 49(12): 615–621.� Norway. National Board of Health. Antibiotics in general practice:

Acute otitis media. 2000. (available from: http://www.helsetilsynet.no/webpubl/antibiotika)� New Zealand. New Zealand Guidelines Group. Acute Otitis Media.

1998. (available from: http://www.nzgg.org.nz)� Scotland. SIGN. Diagnosis and management of childhood otitis

media in primary care. A national clinical guideline 2003.(available from: http://www.sign.ac.uk)� Spain. Llor C, Gonzalez I, Boada A, Luque A. Terapeutica de les

infeccions de les vies aeries altes. Societat Catalana de Medicinade Famılia. Recomanacions sobre l’us d’antimicrobians enl’atencio primaria. 5th ed. Barcelona: Dassoy, 2005. pp.17–34� South Africa. Brink AJ, Cotton MF, Feldman C, et al. Working Group

of the Infectious Diseases Society of Southern Africa. Guidelinefor the management of upper respiratory tract infections. SouthAfr Med J 2004; 94: 475–482.� Sweden. Swedish Medical Research Council. Treatment for acute

inflammation of the middle ear. Consensus statement 2000.(available from: http://en.strama.se/dyn/84.html)� United Kingdom. NICE. National Institute for Health and Clinical

Excellence. Respiratory tract infections – antibiotic prescribing.Prescribing of Antibiotics for Self-limiting Respiratory TractInfections in Adults and Children in Primary Care. NICE Clinicalguideline No. 69. 2008. London, UK: National Institute for Healthand Clinical Excellence. (available from: http://www.nice.org.uk)

References

[1] M.M. Rovers, A.G.M. Schilder, G.A. Zielhius, R.M. Rosenfeld, Otitis media, Lancet363 (2004) 465–473.

[2] S.I. Pelton, Otitis media: re-evaluation of diagnosis and treatment in the era ofantimicrobial resistance, pneumococcal conjugate vaccine, and evolving morbid-ity, Pediatr. Clin. North Am. 52 (2005) 711–728.

[3] J. Liese, A. Carmona, L. Cantarutti, S.A. Silfverdal, A. Fuat, J. Vollmar, et al., Incidenceof acute otitis media in young children seen in European medical practices, in: 6thWorld Congress of the World Society for Pediatric Infectious Diseases, BuenosAires, Argentina, 18–22 November, 2009.

[4] M.M. Rovers, The burden of otitis media, Vaccine 26 (Suppl. 7) (2008) G2–G4.[5] P.S. Mattila, Antibiotics in childhood acute otitis media, Lancet 368 (2006) 1397–

1398.

[6] C.G. Grijalva, J.P. Nuorti, M.R. Griffin, Antibiotic prescription rates for acuterespiratory tract infections in US ambulatory settings, JAMA 302 (2009) 758–766.

[7] A.M. Capra, T.A. Lieu, S.B. Black, H.R. Shinefield, K.E. Martin, J.O. Klein, Costs ofotitis media in a managed care situation, Pediatr. Infect. Dis. J. 19 (2000) 354–355.

[8] D. Greenberg, N. Bilenko, Z. Liss, The burden of acute otitis media on the patientand the family, Eur. J. Pediatr. 162 (2003) 576–581.

[9] A. Schilder, W. Lok, M. Rovers, International perspectives on management of acuteotitis media: a qualitative review, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 29–36.

[10] A. Schilder, Comparison of acute otitis media guidelines across the word, MasterClass Brussels 2008, Presented at ESPO 2008, Budapest (personal communica-tion).

[11] American Academy of Pediatrics and American Academy of Family Physicians,Diagnosis and management of acute otitis media, Pediatrics 113 (2004) 1451–1465.

[12] Manuale Metodologico, Come produrre, diffondere e aggiornare raccomandazioniper la pratica clinica. Istituto superiore di sanita. Agenzia per i servizi sanitariregionali. Programma nazionale per le linee guida, Zadig, Milano, 2002, availableat http://www.snlg-iss.it.

[13] C.D. Bluestone, Definitions, terminology and classification, in: C.D. Bluestone, K. Jo(Eds.), Otitis Media in Infants and Children, BC Decker Inc., 2007, pp. 1–19.

[14] R. Rothman, T. Owens, D.L. Simel, Does this child have acute otitis media? JAMA290 (2003) 1633–1640.

[15] P.H. Kaleida, The COMPLETES exam for otitis, Contemp. Pediatr. 4 (1997) 93–101.[16] P. Marchisio, E. Mira, C. Klersy, F. Pagella, S. Esposito, S. Bianchini, et al., Medical

education and attitudes about the acute otitis media guidelines: an Italian surveyof paediatricians and otolaryngologists, Pediatr. Infect. Dis. J. 28 (2009) 1–4.

[17] S.L. Block, Acute otitis media: bunnies, disposables, and bacterial original sin!Pediatrics 111 (2003) 217–218.

[18] N. Ahman, I. Wacogne, Does the clinical finding of ear wax exclude the finding ofotitis media? Arch. Dis. Child. 94 (2009) 912–913.

[19] P.S. Roland, T.L. Smith, S.R. Schwartz, R.M. Rosenfeld, B. Ballachanda, J.M. Earll,et al., Clinical practice guideline: cerumen impaction, Otolaryngol. Head NeckSurg. 139 (3 (Suppl. 2)) (2008) S1–S21.

[20] F. Marchetti, L. Ronfani, S. Conti Nibali, G. Tamburlini, for the Italian Study Groupon Acute Otitis Media, Delayed prescription may reduce the use of antibiotics foracute otitis media, Arch. Pediatr. Adolesc. Med. 159 (2005) 679–684.

[21] D.P. McCormick, T. Chonmaitree, C. Pittman, K. Saeed, N.R. Friedman, T. Uchida,et al., Nonsevere acute otitis media: a clinical trial comparing outcomes ofwatchful waiting versus immediate antibiotic treatment, Pediatrics 115 (2005)1455–1465.

[22] D.M. Spiro, K.Y. Tay, D.H. Arnold, J.D. Dziura, M.D. Baker, E.D. Shapiro, Wait-and-see prescription for the treatment of acute otitis media: a randomized controlledtrial, JAMA 296 (September 10) (2006) 1235–1241.

[23] M.M. Rovers, P. Glasziou, C.L. Appelman, P. Burke, D.P. McCormick, R.A. Damoi-seaux, et al., Antibiotics for acute otitis media: a meta-analysis with individualpatient data, Lancet 368 (2006) 1429–1435.

[24] S. Sanders, P.P. Glasziou, C. DelMar, M. Rovers, Antibiotics for acute otitis media inchildren, Cochrane Database Syst. Rev. (1) (2004), doi:10.1002/14651858.CD000219.pub2 (Art. No.: CD000219).

[25] D.P. McCormick, S.M. Chandler, T. Chonmaitree, Laterality of acute otitis media:different clinical and microbiologic characteristics, Pediatr. Infect. Dis. J. 26 (2007)583–588.

[26] N. Le Saux, I. Gaboury, M. Baird, T.P. Klassen, J. MacCormick, C. Blanchard, et al., Arandomized, double-blind, placebo-controlled noninferiority trail of amoxicillinfor clinically diagnose acute otitis media in children 6 months to 5 years of age,CMAJ 172 (2005) 335–341.

[27] N.R. Friedman, D.P. McCormick, C. Pittman, T. Chonmaitree, D.C. Teichgraeber, T.Uchida, et al., Development of a practical tool for assessing the severity of acuteotitis media, Pediatr. Infect. Dis. J. 25 (2006) 101–107.

[28] N. Shaikh, A. Hoberman, J.L. Paradise, H.E. Rockette, M. Kurs-Lasky, D.K. Colborn,et al., Responsiveness and construct validity of a symptom scale for acute otitismedia, Pediatr. Infect. Dis. J. 28 (2009) 9–12.

[29] A. Vergison, Microbiology of otitis media: a moving target, Vaccine 26 (Suppl. 7)(2008) G5–G10.

[30] P. Marchisio, L. Claut, S. Gironi, L. Lambertini, S. Tosi, E. Ghisalberti, et al., Role ofgroup A streptococcus in acute otitis media, in: Lim, Bluestone, Casselbrant (Eds.),Recent Advances in Otitis Media, BC Decker Inc. Publisher, 2005, pp. 258–260.

[31] S. Stefani, M.L. Mezzatesta, G. Fadda, R. Mattina, G. Palu, F. Rossano, et al.,Antibacterial activity of cefditoren against major community-acquired respira-tory pathogens recently isolated in Italy, J. Chemother. 20 (5) (2008) 561–569.

[32] A. Novelli, S. Fallani, M.I. Cassetta, S. Conti, Pharmacokinetics and pharmacody-namics of oral cephalosporins as critical factors in choice of antibiotics, Int. J.Antimicrob. Agents 16 (4) (2000) 501–505.

[33] P. Ovetchkine, R. Cohen, Shortened course of antibacterial therapy for acute otitismedia, Pediatr. Drugs 5 (2003) 133–140.

[34] L. Koopman, A.W. Hoes, P.P. Glasziou, C.L. Appelman, P. Burke, D.P. McCormick,et al., Antibiotic therapy to prevent the development of asymptomatic middle eareffusion in children with acute otitis media: a meta-analysis of individual patientdata, Arch. Otolaryngol. Head Neck Surg. 134 (2008) 128–132.

[35] D. Ho, B.W. Rotenberg, R.G. Berkowitz, The relationship between acute mastoid-itis and antibiotic use for acute otitis media in children, Arch. Otolaryngol. HeadNeck Surg. 134 (2008) 45–48.

[36] P.L. Thompson, R.E. Gilbert, P.F. Long, S. Saxena, M. Sharland, I.C. Wong, Effect ofantibiotics for otitis media on mastoiditis in children: a retrospective cohort study

P. Marchisio et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 1209–12161216

using the United Kingdom general practice research database, Pediatrics 123(2009) 424–430.

[37] J.F. Lubianca Neto, L. Hemb, D.B. Silva, Systematic literature review of modifiablerisk factors for recurrent acute otitis media in childhood, J. Pediatr. (Rio J) 82(2006) 87–96.

[38] M.M. Rovers, M.E. Numans, E. Langenbach, D.E. Grobbee, T.J. Verheij, A.G. Schilder,Is pacifier use a risk factor for acute otitis media? A dynamic cohort study, Fam.Pract. 25 (2008) 233–236.

[39] L. Manzoli, F. Schioppa, A. Boccia, P. Villari, The efficacy of influenza vaccine forhealthy children: a meta-analysis evaluating potential sources of variation in

efficacy estimates including study quality, Pediatr. Infect. Dis. J. 26 (2) (2007) 97–106.

[40] A.G. Jansen, E. Hak, R.H. Veenhoven, R.A. Damoiseaux, A.G. Schilder, E.A. Sanders,Pneumococcal conjugate vaccines for preventing otitis media, Cochrane DatabaseSyst. Rev. 15 (April (2)) (2009) (CD001480).

[41] R. Prymula, P. Peeters, V. Chrobok, P. Kriz, E. Novakova, E. Kaliskova, et al.,Pneumococcal capsular polysaccharides conjugated to protein D for preventionof acute otitis media caused by both Streptococcus pneumoniae and non-typableHaemophilus influenzae: a randomised double-blind efficacy study, Lancet 367(2006) 740–748.