Epidemiology of acute otitis media in children of Latin America and the Caribbean: A systematic review and meta-analysis

International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl

Epidemiology of acute otitis media in children of Latin America and the Caribbean: A systematic review and meta-analysis Ariel Bardach a,*, Agustı´n Ciapponi a, Sebastian Garcia-Marti a, Demian Glujovsky a, Agustina Mazzoni a, Alicia Fayad a, Romulo E. Colindres b, Angela Gentile c a b c

Institute for Clinical Effectiveness and Health Policy, Buenos Aires, Argentina GlaxoSmithKline Biologicals, Rio de Janeiro, Brazil Hospital de Nin˜os ‘‘Dr. Ricardo Gutierrez’’, Buenos Aires, Argentina

A R T I C L E I N F O

A B S T R A C T

Article history: Received 10 December 2010 Received in revised form 12 May 2011 Accepted 14 May 2011 Available online 12 June 2011

Background: Acute otitis media (AOM) is one of the most common childhood diseases requiring antimicrobial prescription in pre-school children. This systematic review aimed to estimate the AOM incidence, bacterial etiology and use of resources in children aged <6 years in Latin America and the Caribbean (LA&C). Methods: A systematic search using keywords otitis or middle ear and inflammation was performed for articles published during 1988-2008 in MEDLINE, Cochrane Library, EMBASE, LILACS, generic and academic internet searches, Ministries of Health, PAHO, regional proceedings, reference lists and consulting experts. Pairs of reviewers independently selected articles and assessed their methodological quality with a checklist of essential items from the STROBE statement according to pre-specified criteria. Studies involving immune-competent children with AOM were considered. Arcsine transformations were used for proportion meta-analyses. Results: Annual AOM incidence in four studies in children aged <5 years ranged from 1,171–36,000 episodes/100,000 children. Meta-analysis on etiology and pneumococcal serotypes included 18 studies and 125, 519 children with AOM from six LA&C countries. Meta-analysis per serotype showed that Streptococcus pneumoniae (32.4%; 95%CI = 27.1-38.0%) and Haemophilus influenzae (26%; 95%CI = 19.533.1%), including non-typeable H. influenzae (18.3%; 95%CI = 9.5-33.1%) were the most prevalent. The most commonly observed pneumococcal serotype was 19F (24.0%; 95% CI 17.0-32.0%). Data on use of health resources were scarce. Conclusions: Streptococcus pneumoniae and H. influenzae were the most frequent AOM bacterial pathogens, consistent with the international literature from other regions. Future studies on AOM incidence and health resources usage will help better define the impact of this disease. ß 2011 Elsevier Ireland Ltd. All rights reserved.

Keywords: Acute otitis media Epidemiology Latin America Caribbean Meta-analysis

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and methods . . . . . . . . . . . . . . . . . . 2.1. Search strategy and eligibility criteria . Screening and data abstraction . . . . . . 2.2. Methodological quality assessment . . . 2.3. Statistical analysis. . . . . . . . . . . . . . . . . 2.4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Incidence . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Etiology . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Pneumococcal serotypes. . . . . . . . . . . . 3.3. Use of health resources . . . . . . . . . . . . 3.4.

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

* Corresponding author. Tel.: +54 11 4777 8767. E-mail address: [email protected] (A. Bardach). 0165-5876/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2011.05.014

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

1063 1063 1063 1063 1063 1063 1064 1066 1066 1066 1067

4.

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

1063

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1068 1069 1070

1. Introduction Acute otitis media (AOM) is one of the most common childhood infections and a frequent reason for pediatricians to prescribe antimicrobials [1]. It generally affects children <3 years of age and, if not properly treated, can cause significant hearing loss, which may potentially result in speech, language, and cognitive skills impairment [2]. The disease occurs in all ages, but it is more common in infancy. Fluid may persist for weeks or months after the onset of illness, despite appropriate antimicrobial treatment. In the United States (US) and Europe, AOM incidence has been estimated to range from 0.1 to 1.8 episodes/year per child <2 years of age, commonly with multiple consultations for an individual episode [3–6]. In the first two years of life, more than 25% of children in almost all settings suffer from AOM caused by Streptococcus pneumoniae [6–8], while close to 80% of children have at least one AOM episode by the time they are three years old in the US [6,8,9]. A significant increase in the incidence of AOM has been observed due to socialization of children at younger ages (by attending day care centers or kindergartens), which increases the risk of infection [10]. In addition, other important risk factors are age, time of year, socioeconomic factors and large numbers of members within a household [10–12]. In developing countries, suppurative infections, including mastoiditis and meningitis, are still major complications of AOM [13]. According to estimates from the World Health Organization, annually 50,000 children aged <5 years die due to AOM-related complications in developing countries [14,15]. This high mortality may be attributed to a late recognition of such complications (perforation and mastoiditis) or due to poor nutritional status [15]. In 2000, the annual estimates of healthcare expenditure to treat otitis media alone in the US were 5 billion dollars of which 40% were for children aged 1–3 years [16]. In addition, data from a US national survey in 1992 indicated that 30% of all antibiotic prescriptions were to treat otitis media in that age group [17]. In routine practice, obtaining etiological culture-confirmed diagnosis is uncommon, thus epidemiological information becomes crucial to determine empirical therapies based on evidence or for the implementation of vaccines that could prevent AOM. No comprehensive systematic reviews assessing the incidence, etiology and use of health resources for AOM in children <5 years of age in LA&C is available to date. Therefore, a detailed systematic review of the overall epidemiological data on AOM was performed to evaluate the burden of this disease. 2. Materials and methods 2.1. Search strategy and eligibility criteria We conducted a systematic search of articles published in the following electronic databases: CENTRAL, MEDLINE, EMBASE, and LILACS from January1988 to January 2008 (Supplementary data). The apparent stability of the bacterial etiology of AOM [18] over time (no secular trends) supported the utilization of an unrestrictive 20-year time window. We also performed a generic and academic Internet search and meta-search. An annotated search strategy for grey literature was included to retrieve information from relevant sources like LA&C countries’ Ministries of Health, PAHO, reports from hospitals, databases containing regional proceedings or congresses’ annals and doctoral theses, reference lists of included studies and consulting experts and associations

related to the topic. Authors were contacted to obtain missing or extra information when needed. Studies involving immune competent patients <6 years of age with a diagnosis of AOM were considered. Three age range categories were selected: 0–6 months, 7–23 months and 24– 71 months. AOM was defined as acute inflammation (72 h) of the middle ear together with signs or symptoms of infectious disease and confirmed by standardized methods (clinical, otoscopic, pneumatoscopic, tympanocentesis). Labyrinthitis, chronic or recurrent otitis media, and serous otitis media were excluded from this review. Studies including children up to 12 years of age were included in this review only if they also contained information for children <5 years of age. Studies with a sample size below 20 patients were excluded. The following outcomes were assessed: incidence of AOM, distribution of pathogens and health resources used during episodes. 2.2. Screening and data abstraction Pairs of independent researchers screened titles and abstracts of all identified citations. They categorized the articles into one of the following categories: potentially eligible, related reviews, related references, and excluded. The full-text versions of all articles that were not ‘excluded’ were obtained. Two independent researchers assessed the full-texts of potentially eligible articles to evaluate whether they met the inclusion/exclusion criteria. Data was abstracted using previously piloted electronic charts. 2.3. Methodological quality assessment The risk of bias of observational studies was assessed by a checklist of essential items from the STROBE statement (strengthening the reporting of observational studies in epidemiology) [19], MOOSE (meta-analysis of observational studies in epidemiology) [20], randomized clinical trials (RCT) [21] and from Sanderson et al. [22] and Fowkes and Fulton [23]. Two reviewers independently assessed the risk bias of each included study. Discrepancies in quality evaluation were resolved by team consensus, and authors of potentially eligible articles were contacted to provide missing information if needed. We used an algorithm (Supplementary data), programmed in an Excel spreadsheet, to estimate a summary risk of bias considering three major criteria (methods for selecting study participants, methods for measuring exposure and outcome variables, and methods to control confounding) and two minor criteria, excluding confounding (design-specific sources of bias, and statistical methods). 2.4. Statistical analysis The incidence of AOM was expressed as AOM episodes per 100,000 children-years. To analyze etiology and pneumococcal serotype data, we performed meta-analysis of proportions per serotype. We applied an Arcsine transformation to stabilize the variance of proportions (Freeman–Tukey variant of the Arcsine square root of transformed proportions method), where y = arcsine[H(r/(n + 1))] + arcsine[H(r/(n + 1)/(n + 1)], with a variance of 1/ (n + 1), and where n is the population size [24]. The pooled proportion was calculated as the back-transformation of the weighted mean of the transformed proportions, using inverse

1064

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

Potentially relevant studies identified (n = 195) - International literature databases, such as Cochrane Library, MEDLINE, EMBASE and LILACS, without language restrictions. - Regional databases: generic and academic search, Internet meta-trackers, abstracts from congresses and scientific thesis. - Grey literature was identified in national ministries of health, PAHO, scientific societies and generic trackers Not included per title/summary (n = 140) Articles were excluded after reviewing the abstract (if available) or the title (if the abstract was unavailable) Full-text evaluated studies (n = 59)

Excluded studies (n = 41) - Criteria not met (n = 26) - Full text unavailable (n = 1) - Related references (n =2) - Related reviews (n = 11) - Data duplicates (n = 1) Included studies (n = 18) - Incidence information (n = 4) - Etiology information (n = 13) - SPN serotype information (n = 4) - Use of resources information (n = 4)

Fig. 1. Study Flowchart.

arcsine variance weights for the fixed and random effects model. We applied DerSimonian–Laird weights for the random effects model [25] where heterogeneity between studies was found. We calculated the I2 statistic as a measure of the proportion of the overall variation in the proportion that was attributable to between-study heterogeneity [26]. StatsDirect and Stata8.01 were used for all analyses. Data was insufficient to perform the planned subgroup analyses (age group, factors related to AOM development and etiology, country, geographic region and national gross income categories). However, a cumulative meta-analysis, with increasing age range, was a better approach to deal with this scarce data. We included different studies in every single meta-analysis for each pathogen and for each serotype in order to include all the available information. This explains why the cumulative point estimates do not reach exactly 100% in the tables shown. 3. Results Of the 195 published studies identified and reviewed, 18 studies met the inclusion criteria (Fig. 1). Four studies each (not necessarily the same studies) provided information on AOM

incidence, pneumococcal serotypes, and health resource usage; and13 studies provided information on etiological agents. The study designs were classified as case-series (8), cohort (6), cross sectional (2) and before–after (2). Studies were conducted between 1979 and 1999, in the following six countries in LA (Latin America): Mexico (5), Argentina (4), Chile (4), Costa Rica (3), Brazil (1) and Colombia (1). We did not find any study from the Caribbean region. The studies were conducted at referral centers (8), non-referral centers (5), in the overall population (1) and in the rural population (1). The setting of the study was not reported in three studies. Table 1 summarizes the main characteristics of the included studies. A total of 125,519 children with AOM were included. As per protocol, the lower limit of age was 0 months (median 1.5, mean 1.8  1.6) and the upper limit was 167 months (median 59.5, mean 73.8). The median of index patients per study was 111 (range 9–925; mean 193.9) excluding the study from Arevalo Silva et al. [18] which included 122,222 patients with AOM. The duration of each study was between 3 and 72 months (median 35.0, mean 37.2). In three studies reporting gender, the proportion of males with AOM was 51.1–53.7% [18,27,28]. Risk factors reported in three studies were the following: non-breastfed (5.9–23.6%) [27,28] passive smoking (45.9–70.3%) [27,28], had colds or upper airway infection (31.2%) [29], and day

Table 1 Characteristics of studies included in the analysis. Start-end

Duration (months)

Socio economic level

Risk factors

Parameter assessed

9 (1–24)

01/01/2001–30/12/2004

48

Not reported

–

Etiology

432

01/01/1990–30/12/1995

72

Not reported

–

Etiology

–

925

01/01/1994–30/12/1999

72

Not reported

–

Etiology

Before-After

–

41

–

Not reported

–

Etiology

Non-probabilistic

Cross sectional

–

Non-probabilistic

Case series

–

Non-reference center

Non-probabilistic

Cohorts

246

Etiology serotype Etiology serotype Incidence

Non-reference center Non-reference center

Non-probabilistic

Cohorts

Non-probabilistic

Country (city)

Field

Sample

Design

Children N (males)

Comisso et al. [34]

Argentina (Cordoba) Argentina

Reference center

Non-probabilistic

Case series

–

174

Reference center

Non-probabilistic

Case series

–

Reference center

Non-probabilistic

Case series

Reference center

Non-probabilistic

Non-reference center Reference center

Orlando et al. [38]

Orlando et al. [39] Desio et al. [36] Sih [40] Cofre et al. [35] Lopez et al. [27]

(Buenos Aires) Argentina (Buenos Aires) Argentina (Multicenter) Brazil (San Pablo) Chile (Santiago) Chile

Children with AOM

(3–120)

–

300

(2–60)

01/01/1990–01/12/1995

60

Not reported

–

172

–

18

Not reported

–

82

31 (4–86) (0–59)

01/01/1991–01/01/1996

60

Low

–

154

(0–59)

01/05/1997–30/04/1999

24

Low

Not breastfed (24%), passive smoking (70%), day care center (100%) –

Cohorts

–

170

36 (3–60)

01/07/1998–30/06/1999

12

Low

(Santiago) Lopez Bravo et al. [44]

Chile (Santiago)

Rosenblut et al. [29]

Chile (Santiago)

Trujillo et al. [41]

Colombia (Medellı´n)

Not reported

Not reported

Cross sectional

–

111

(1–3)

25/08/1979–26/02/1985

58

Not reported

Arguedas, 2003 [30]

Costa Rica

Reference center

Non-probabilistic

Case series

–

102

36 (4–123)

01/01/1999–1/12/2001

35

Not reported

Arguedas et al. [33] Arguedas et al. [32]

Costa Rica Costa Rica (San Jose´) Mexico

Reference center Reference center

Non-probabilistic Non-probabilistic

Case series Case series

– –

69 398

– 01/03/1992–01/01/1997

24 58

Not reported Not reported

Rural population

Non-probabilistic

Cohortsb

1571

(3–49) 42 (4–144) (0–59)

Day care center (9%), bilateral otitis media (56%) – –

01/03/1982–28/02/1983

12

Low

–

Nandi-Lozano et al. [28]

Mexico (Mexico DF)

Non-reference center

Not reported

Cohortsb

85

(1–48)

01/04/1999–01/10/1999

7

Not reported

Villasenor-Sierra et al. [42]

Mexico

Not reported

Non-probabilistic

Case series

–

57

(0–167)

01/05/1991–31/01/1993

21

Not reported

Not breastfed (5.9%), passive smoking (45.9%) –

(Mexico DF) Mexico (Mexico DF)

Not reported

Non-probabilistic

Before–After

–

30

19 (3–71)

01/08/1999–31/10/1999

3

Not reported

–

Mexico (Nation)

Population-wide

Non-applicable

Cohortb

10,941, 493 (5,587, 001)

122,222

(0–48)

01/01/1995–31/12/1998

48

Not reported

–

Gutierrez Trujillo et al. [31]

Morayta Ramirez et al. [37] Arevalo Silva et al. [18]

a b

9

Etiology Health resource Etiology Serotype

Etiology Etiology Incidence Health resource Incidence

Etiology

Etiology Health resource Incidence

1065

Age ranges or inclusion age limits (in case age range is not reported). Surveillance interval: 7 days, all studies were conducted between1979 and 1999.

71

Recurrent AOM (14.7%), colds (31.2%) –

Health resource Etiology erotype

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

Mean age (rangea) (months)

Study

1066

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

Table 2 AOM annual episodes per 100,000 children by age group. AOM rate per 100,000 children-years

Study

Arevalo Silva et al. (Mexico) [18] Lopez et al. (Chile) [27] Gutierrez Trujillo et al. (Mexico) [31] Nandi-Lozano et al. (Mexico) [28]

0–1 year of age

1–4 years of age

0–4 years of age

1260 16,300 – –

1083 4030 – –

1171 2698 4519 36,000

Table 3 Meta-analysis of the etiological agents of AOM. Etiologic agent

S. pneumoniae Total H. influenzaeb - Non-typeable H. influenzae - H. influenzae type b S. pyogenes A S. epidermidis M. catarrhalis S. aureus P. aeruginosa S. a-hemolytic Others Bacterial Co-infection No growth

Overall prevalence % (95% CI)

32.40 26.03 18.26 3.75 5.98 3.66 3.47 2.77 1.48 1.20 4.57 3.46 30.71

(27.06–37.99) (19.52–33.12) (10.00–28.32) (1.27–7.48) (2.00–11.91) (2.31–5.32) (2.08–5.19) (1.87–3.83) (0.50–2.96) (0.41–2.40) (0.92–10.75) (2.34–4.80) (24.01–37.83)

I2%

88.3 93.6 93.3 88.9 84.1 0 77.5 46.5 62.5 0 97 0 93.1

Prevalence % (95% CI) 1–3 months

1–60 months

20.72 33.33 28.83 7.20 0.90 3.60 0.90 2.70 – – 9.01 2.70 26.13

28.79 24.64 17.99 2.66 5.84 4.15 2.87 1.90 2.00 1.76 11.09 3.04 28.58

(13.61–29.45) (24.67–42.91) (20.63–38.20) (3.16–13.71) (0.02–4.92) (0.99–8.97) (0.02–4.92) (0.56–7.70)

(4.41–15.94) (0.56–7.70) (18.25–35.32)

(16.07–3.50) (9.88–43.45) (5.39–35.83) (0.46–6.57) (0.01–22.65) (1.51–8.02)a (1.27–5.09) (0.10–2.99) (0.74–4.30) (0.37–5.07) (0.0–38.67) (1.58–4.96) (15.86–43.32)

I2 = inconsistency. a 72 months was established as upper limit [37] since no data were available for 60 months. b Total H. influenzae included 11 studies (only 3 could be analyzed within this group since typing was not available).

care attendance (9.0–100.0%) [27,30]; clinical variants or complications were bilateral otitis media (56.0%) [30] and recurrent AOM (14.7%) [29]. The summary risk of bias was generally high in 13 studies and average in five studies (see Supplementary data). 3.1. Incidence Incidence of AOM was available from four studies [18,27,28,31] (Table 2). Three of these studies were conducted in Mexico while one was from Chile. The Mexican AOM study by Arevalo Silva et al. [18] was a population-based study conducted from 1995 to 1998. The rate of reported cases per 100,000 persons-years, according to age groups, within all the Mexican states was as follows: <1 year of age (924–1688); 1–4 years of age (756–1384); 5–14 years of age (425– 829) and overall ages (366–745). Gutierrez Trujillo et al. [31] studied the antimicrobial prescription patterns in 8002 acute respiratory infection (ARI) episodes involving 1359 families living in 137 rural or semi-rural districts in Mexico. The surveillance was conducted during 1982–1983 through weekly home visits of 10 families randomly selected from each district. During the surveillance period, 133/1571 (8.5%) children <5 years of age presented AOM. In the last Mexican study included, Nandi-Lozano et al. [28] studied ARI in a cohort of 85 children <4 years of age attending the day care center of a Pediatric Hospital in Mexico City, from April to October 1999. There were 258 acute respiratory infection events: 246 cases of rhinopharyngitis (95.3%), nine of AOM (3.5%), three of bronchiolitis (1.2%). The annual mean incidence rate in Mexico was 36,000 per 100,000 children aged <4 years. In Lopez et al. [27] studied the frequency of AOM in 246 children <5 years of age at an outpatient center in Santiago de Chile from 1992 to 1997. The annual mean rate of AOM was 16,300 per 100,000 children in infants below 1 year of age and 13,488 per 100,000 children <4 years of age. 3.2. Etiology Thirteen studies [29,30,32–42] provided information on middle ear effusion collected and cultured to document the bacteria

responsible for AOM used in the meta-analysis on AOM etiology and four of these studies [29,33,35,40] provided information on specific AOM pneumococcal serotypes. Characteristics of these studies are presented in Table 1 and details of each study in Supplementary data. Cumulative meta-analysis per age range is reported to provide the most precise estimate of etiologic prevalence according to age. They show the three most relevant age ranges: 0–1 years (1– 3 months), 1–4 years, and overall 0–4 years, inclusive of children above 1 month of age or without age data. An etiology summary table in decreasing order of frequency is shown (Table 3). The meta-analysis per pathogen using the random-effect model demonstrated that the main causative agents for purulent AOM in children >3 months of age were, S. pneumoniae (32.4%; 95% CI: 27.1–38.0) [29,30,32,34–41], H. influenzae (26.0%; 95% CI: 19.5– 33.1) [29,30,32,34–42], non-typeable H. influenzae (18.3%; 95% CI: 10.0–28.3) [29,30,32,40–42], Streptococcus pyogenes A (6.0%; 95% CI: 2.0–11.9) [29,30,35,41], Staphylococcus epidermidis (3.7%; 95% CI: 2.3–5.3) [32,37,41,43], Moraxella catarrhalis (3.5%; 95% CI: 2.1– 5.2) [29,30,32,34–36,38–41], and to a lesser extent, S. aureus (2.77%; 95% CI: 1.9–3.8) [29,32,34,36–41,43], Pseudomonas aeruginosa (1.5%; 95% CI: 0.5–3.0) [32,36,38,40] and S. alpha haemolytic (1.2%; 95% CI: 0.4–2.4) [29,32] (Fig. 2). In a study conducted in Argentina by Commisso et al. [34], H. influenzae (57.7%) was the most common causative organism in children <6 months of age and S. pneumoniae (32.3%) in children >6 months of age. No bacterial growth was very common (30.7%). Overall, co-infection of several bacteria was rare, but observed in 2.7% of children aged 1– 3 months. 3.3. Pneumococcal serotypes A random effects model was used in the meta-analysis of proportions per serotype presented in decreasing order of frequency in Table 4. Altogether, 14 pneumococcal serotypes were detected among the four studies that reported this outcome. The most common serotypes were, 19F (24%) [29,33,35,40], 6B

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

A. S. pneumoniae

B. Total H. influenzae

1.

Trujillo 1989

0.21 (0.14, 0.29)

2.

Comisso 2006

0.44 (0.36, 0.51)

3

Sih 2001

0.16 (0.12, 0.21)

4. Rosenblut 2001

0.37 (0.30, 0.45)

1.

Trujillo 1989

2.

4.

0.36 (0.27, 0.46)

Comisso 2006

0.40 (0.32, 0.47)

3.

Sih 2001

0.07 (0.04, 0.11)

Rosenblut 2001

0.24 (0.18, 0.31)

Morayta 2000

0.33 (0.17, 0.53)

5.

5. Morayta 2000

0.37 (0.20, 0.56)

6. Cofre 2004

0.42 (0.35, 0.50)

6.

Cofre 2004

0.28 (0.21, 0.35)

7. Desio 2000

0.49 (0.33, 0.65)

7.

Desio 2000

0.29 (0.16, 0.46)

Arguedas 2003

0.14 (0.08, 0.22)

8. Arguedas 2003

0.42 (0.32, 0.52)

8.

9. Arguedas 1998

0.30 (0.25. 0.34)

9.

Arguedas 1998

0.14 (0.11, 0.18)

10. Orlando 1996

0.27 (0.23, 0.31)

10. Morayta 2001

0.21 (0.11. 0.34)

11. Orlando 1996

0.24 (0.20. 0.28)

11. Orlando 2000

0.32 (0.29, 0.35)

12. Orlando 2000

0.28 (0.25. 0.31)

12. Villaseñor 1999

0.42 (0.28, 0.58)

0.32 (0.26. 0.37)

combined

0.26 (0.20, 0.33)

combined 0.0

0.2

0.4

0.6

0.0

0.8

C. H. influenzae type B

5.

Sih 2001

0.06 (0.04, 0.09)

Rosenblut 2001

0.23 (0.17, 0.30)

4.

Arguedas 2003

0.13 (0.07, 0.21)

5.

Arguedas 1998

0.09 (0.06, 0.12)

0.07 (0.03, 0.14)

Sih 2001

0.01 (0.01, 0.03)

Rosenblut 2001

0.01 (0.01, 0.04)

4.

Desio 2000

0.20 (0.09, 0.34)

Arguedas 1998

0.05 (0.03, 0.07)

6. Orlando 1996

6.VillaseñorSierra1999

0.01 (0.00, 0.02)

combined 0.0

0.8

0.29 (0.21, 0.38)

3. 3.

0.6

Trujillo 1989

2. 2.

0.4

D. Non typeable H. influenzae 1.

Trujillo 1989

0.2

Proportion (95% confidence interval)

Proportion (95% confidence interval)

1.

1067

0.04 (0.01, 0.07) 0.1

0.2

0.3

0.42 (0.28, 0.58)

combined

0.4

0.18 (0.10, 0.28) 0.0

Proportion (95% confidence interval)

0.2

0.4

0.6

0.8

Proportion (95% confidence interval)

Fig. 2. Proportion meta-analyses of etiologic agents among tested AOM samples.

(19%) [33,40], 19A (17%) [35,40], 5 (15%) [29], 9V (11%) [33,40], 14 (9%) [29,33,35], and 23F (8%) [29,33,40] accounting for 60–70% of the strains isolated from AOM in the age group 3–86 months. Other reported serotypes were, 3 (8%) [29,33,35], 18C (7%) [33,40], 1 (5%) [29,33], others (49%) [33,35,40], along with 6A [33], 4 [33] and 15C [33] (the most common forest plot in Fig. 3). Serotypes 12F, 7F, 33F and NT (non-typeable) were not reported in the included studies. 3.4. Use of health resources Information on the use of health resources was provided by four studies [31,37,41,44]. Lopez Bravo et al. [44] assessed the magnitude and characteristics of the morbidity cared for at

primary outpatient centers in 2000. A systematic random sample was collected monthly, representative for age and gender, until completing 468 children <6 years of age (53% were males). Each child was followed-up over one year. Of the 3122 primary consultations, AOM was the third most common reason for pediatric consultation (i.e. 29.1% of consultations in children <1 year of age and 8.2% of consultations in children <5 years of age). The most frequently diagnosed infections were acute bronchitis (17.3%), obstructive bronchitis (16.9%) and AOM (6.4%) [44]. Gutierrez Trujillo et al. [31] studied the antimicrobial prescription pattern in ARI presented by 1359 families in rural area of Mexico over a period of one year. The prescribed treatment was examined in 8002 ARI episodes observed. The annual rate of the

Table 4 Meta-analysis of pneumococcal serotype-specific prevalence for AOM. Serotype Prevalence (%) 95% CI

19F 24 17–32

6B 19 9–31

19A 17 4–35

5 15 7–24

9V 11 5–17

14 9 5–13

23F 8 5–13

3 8 5–12

18C 7 0.6–11

6A 6 2–13

1 5 0.3–10

4 2 0–7

15C 2 0–7

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

1068

A. 19F

B. 6B

Arguedas 2005 Costa Rica 3-49 m

0.26 (0.16, 0.38) Arguedas 2005Costa Rica 3-49 m

Cofre 2004 Chile 4-86 m

0.21 (0.12, 0.32)

Rosenblut 2001 Chile 3-60 m

0.16 (0.08, 0.27)

Sih 2001 Brazil 2-60 m

0.39 (0.20, 0.61)

0.14 (0.07, 0.25)

Sih 2001 Brazil 2-60 m

0.26 (0.10, 0.48)

combined combined

0.19 (0.09, 0.31)

0.24 (0.17, 0.32)

0.0

0.2

0.4

0.6

0.0

0.8

0.2

C. 9V

0.6

D. 1

Arguedas 2005 Costa Rica 3-49 m

Sih 2001 Brazil 2-60 m

combined

0.0

0.4

Proportion (95% confidence interval)

Proportion (95% confidence interval)

0.1

0.2

0.3

0.09 (0.03, 0.18)

Cofre 2004 Chile 4-86 m

0.10 (0.04, 0.19)

0.13 (0.03, 0.34)

Sih 2001 Brazil 2-60 m

0.26 (0.10, 0.48)

0.11 (0.05, 0.17)

combined

0.17 (0.04, 0.35)

0.4

Proportion (95% confidence interval)

0.0

0.2

0.4

0.6

Proportion (95% confidence interval)

Fig. 3. Proportion meta-analyses of frequent pneumococcal serotypes of AOM tested samples.

evaluated episodes was approximately 3.5 per person in children <2 years and 1.9 in children <5 years of age. AOM was the cause of 1.7% of the overall consultations in children <5 years of age, occupying fifth place as a cause of consultations. Specifically for AOM, 85.8% were treated with antibiotics, 29.3% with antihistamines and 61.7% with anti-pyretics. Penicillin (54.7%) was the most frequently used antibiotic in children >5 years of age and ampicillin (25%) in children <5 years of age. In the study conducted by Trujillo et al. in the late 1980s [41], 111 Colombian children diagnosed with AOM underwent tympanocentesis for middle ear culture. Although 18 children had received some antibiotic prior to culture, all isolates presented were penicillin sensitive and none of them were beta-lactamase producers. In another study in Mexico [37], 30 children were evaluated over a 3-month follow-up period. The procedures performed during the follow-up included otoscopy, culture of ear secretion and antibiotic therapy with amoxicillin/sulbactame 50 mg/kg during ten days. Clinical cure was obtained in 98.1% of the cases. 4. Discussion Reliable epidemiological data on etiology and burden of AOM are important to make well-informed health policy decisions. This systematic review and meta-analysis provide data and identify

information gaps about etiology, incidence and health economic impact of AOM in LA&C. This systematic review in LA&C countries showed that the reported annual incidence of AOM was considerably lower than previously reported from developed countries [8,45,46]. In a study conducted in Boston, children had a mean of 1.2 and 1.1 AOM episodes during the first and second years of age, respectively [8]. At least one episode of AOM was observed in 60–80% of children during the first year of life and in 80–90% of children between 2 and 3 years of age [8,45]. The highest incidence of AOM has been reported in the US, in children aged 6–24 months. Based on the evidence provided by studies in Finland and Boston, the probable AOM incidence up to 2 years of age was 0.93 (0.90–0.96) [46] to 1.28 episodes. This in turn shows that the probable AOM cumulative incidence up to 5 years of age will be at least 90%. The projected annual burden of disease using cumulative 5-year incidence in LA&C showed 10.5 million AOM cases for the region for a birth cohort of 11.7 million, resulting in a similar cumulative incidence percentage (89%), with 12% estimated to be caused by pneumococcus [47]. Total events of pneumococcal otitis media have been projected to be 108 events per 1000 children in Brazil, Chile and Uruguay (projected for the 2005 birth cohort during the first five years of life) [48]. Compared to this conservative assumption for this region, the reported incidence was 80-fold

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

lower in Arevalo Silva et al. [18] (incidence 1.1%), 2.5-fold lower in Nandi-Lozano et al. [28] (incidence 36.0%), while intermediate values were observed in Gutierrez Trujillo et al. [31] (incidence 4.5%) and Lopez 1998 (incidence 2.7%) [27]. In the current systematic review, the highest incidence of otitis was found in Nandi-Lozano et al. [28] study (0.36 episodes per child-year), probably because it followed children attending day care center [28]. Nevertheless, this figure is lower than could be expected possibly due to under-reporting. Furthermore, a decreasing trend in the incidence in higher-age groups was apparent, which is consistent with the literature. It was observed that the incidence of AOM in children aged 1–4 years compared to those <1 year of age was 14% lower in Arevalo Silva et al. [18] study and 26% lower in Nandi-Lozano et al. [28] study. It is important to note that the clinical diagnosis of acute otitis media and the specific criteria used to make a diagnosis vary widely. In the absence of uniform criteria, drawing conclusions about the relative incidence of disease becomes very difficult. S. pneumoniae, H. influenzae, M. catarrhalis and S. pyogenes have been the most common bacterial causes of AOM for the past half century [49]. International literature reports showed that S. pneumoniae and H. influenzae are by far the most common pathogens isolated from middle ear fluid cultures obtained by tympanocentesis at any age [41]. The majority of H. influenzae strains currently causing AOM are non-typeable H. influenzae. In the current review, we found that S. pneumoniae (32.4%) and H. influenzae (26.0%) were consistently the two most commonly identified pathogens [34]. The prevalence of non-typeable H. influenzae found in our review was 18.3% (CI 95% 10.0–28.3), and even higher in the first three months of life (28.8%). This organism has been related to recurrent AOM and treatment failures, and continues to be a major therapeutic challenge in the treatment of AOM disease [49]. Generally for LA&C there is paucity of studies that bring information about the etiological contribution of each type of Haemophilus, mainly due to limitations in the capacity of countries’ laboratories to type samples, which has improved only in the last decade. Importantly, all studies included were performed before 1999, prior to full implementation of the Hib vaccine in the region. Some studies in other regions [49,50] suggest the etiological role for non-typeable H. influenzae to be higher than what is reported in our current review. Some studies were not included due to methodological reasons. Commisso et al. published in 2000 a cross-sectional study, but etiology percentages refer to middle ear samples and not patients with AOM [51], and thus selection bias could not be discarded. Rosenblut et al. [52] published an extension of a 2001 paper [29] but did not include raw serotype data. Historically, relatively few S. pneumoniae serotypes have been responsible for the majority of AOMs. In a multinational study of pneumococcal serotypes causing AOM in children based on nine large databases of middle ear cultures in different countries in Europe, United States and South Africa, Hausdorff et al. [53] found that the seven most prevalent serotypes found were: 19F (16.1%), 23F (14.4%), 14 (13.1%), 6B (10.1%), 6A (7.3%), 19A (6.6%) and 9 V (4.6%), representing 60–70% of AOM isolates in children 6– 59 months of age. In our review, the first seven most prevalent serotypes were: 19F (24%), 6B (19%), 19A (17%), 5 (15%), 9 V (11%), 14 (9%) and 23F (8%). Of these, serotype 5 which is uncommon in AOM was only documented in a Chilean study. Data on pneumococcal invasive disease from Latin America and the Caribbean showed that the distribution of the first seven S. pneumoniae serotypes observed were: 14 (36.2%), 1 (11.3%), 6B (7.2%), 19A (3.8%), 9 V (3.7%), 19F (3.6%) and 23F (3.3%) [54]. Thus, with the exception of serotype 1, the other serotypes observed were similar to that of the non-invasive cases of AOM (SIREVA 2000–2005) [54]. A review by Valenzuela et al. has shown that by

1069

vaccinating children with pneumococcal vaccine, approximately 660,000–1,100,000 pneumococcal cases and 6300–14,000 deaths can be avoided in a birth cohort of 10.8 million assuming the vaccination coverage to be 92% (same as DTP3 vaccination coverage in 2005) [47]. Information on health resource usage for AOM was very limited. Estimates in the US indicate that 90% of children below 3 years of age will undergo at least one episode of AOM over that period of their life, representing 33% of consultations in infants <1 year of age [16]. The frequency of use of health resources in Brazil, Chile, and Uruguay was recently assessed. In these three countries 100% of the patients required antibiotics, and 60–95% concomitant medication. About 7–8% needed laboratory tests and 28–87% underwent procedures like myringotomy or tympanoplasty. The main costs associated with AOM came from outpatient visits and medication. Costs for an episode of AOM ranged from $20 (Brazil) to $217 (Chile) 2004 [55,56]. In a study by Sinha et al. in Latin America, base case estimate for costs of AOM has been deemed to be $101 (78–234) according to physician and parent surveys [57]. Otologic illnesses are among the most frequent reasons for consultations during the first five years of life. The number of visits for otitis media in US increased from 9.5 to 24.5 million between 1975 and 1990 [58]. The study of Lopez Bravo et al. [44] found that AOM was the fourth reason for primary care visits in Chile. Given that this is one of the most common childhood infectious diseases and a major cause of antibiotic use, more studies should be done to better determine the economic and health resource impact. Considering our highly sensitive systematic search, including grey literature, it is possible but unlikely that we have omitted relevant studies. The main limitations of this systematic review were the risk of bias of incidence studies included, mainly due to under-reporting; the lack of suitable population-based welldesigned studies; and heterogeneity of data. Even after considering these limitations, etiological data were more reliable, although the quality of the microbiological procedures used, such as sample collection and transportation and bacterial isolation in some studies could explain some discrepancies in isolation rates. This is the first systematic review strictly designed to determine incidence, etiology, and use of health resources for AOM in children in LA&C. High quality data on AOM incidence and use of health resources in children is scarce for LA&C. AOM etiology reported from LA&C demonstrates trends similar to internationally published data, with S. pneumoniae and H. influenzae as the first and the second most common pathogens. Newly licensed vaccines against pneumococcal and H. influenzae disease could have a large impact on reducing the burden of AOM disease. Acknowledgements The authors would like to thank Farah Mecci, Eiman Jahangir, and Geetha Subramanyam for their writing assistance and Tatiana M Lanzieri for her editorial review. They also thank Patricia Aruj and Juan Calcagno (researchers) for helping with the screening and data abstraction, Daniel Comande´ (librarian) for bibliographic support, Luz Gibbons (biostatistician) for helping with the statistical analyses, Eduardo Ortega-Barria for critical reviewing of the manuscript, Fla´via R. M. Lamara˜o and Jessica Mattos (GSK Biologicals employer/contractor, respectively) to help in manuscript coordination and editorial assistance. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ijporl.2011.05.014.

1070

A. Bardach et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 1062–1070

References [1] M. Uhari, K. Mantysaari, M. Niemela, A meta-analytic review of the risk factors for acute otitis media, Clin. Infect. Dis. 22 (6) (1996) 1079–1083. [2] Rubinstein A and Col (Ed.), Medicina Familiary Pra´ctica Ambulatoria, Segunda Edicio´n, Editorial Me´dica Panamericana, Buenos Aires, 2006. [3] B. Fireman, S.B. Black, H.R. Shinefield, J. Lee, E. Lewis, P. Ray, Impact of the pneumococcal conjugate vaccine on otitis media, Pediatr. Infect. Dis. J. 22 (2003) 10–16. [4] R. Prymula, P. Peeter, V. Chrobok, P. Kriz, E. Novakova, E. Kaliskova, Pneumococcal capsular polysaccharides conjugated to protein D for prevention of acute otitis media caused by both Streptococcus pneumoniae and non-typable Haemophilus influenzae: a randomised double-blind efficacy study, Lancet 367 (2006) 740–748. [5] J. Eskola, T. Kilpi, A. Palmu, J. Jokinen, J. Haapakoski, E. Herva, et al., Efficacy of a pneumococcal conjugate vaccine against acute otitis media, N. Engl. J. Med. 344 (2001) 403–409. [6] D.W. Teele, J.O. Klein, B. Rosner, Epidemiology of otitis media during the first seven years of life in children in greater Boston: a prospective, cohort study, J. Infect. Dis. 160 (1989) 83–94. [7] P.M. Riquelme, V.P. Rincon, R.L. Garcia, C.J. Tusset, T.R. Bravo, P.C. Fernandez, et al., Acute otitis media in a pediatric primary care unit, An. Pediatr. (Barc.) 61 (2004) 408–412. [8] M.J. Caceres Udina, J.A. Alvarez Martinez, J. Argente del Castillo, M.A. Chumilla Valderas, E. Ferna´ndez A´lvarez, A. Garrido Romera, et al., Incidence, air pollution and risk factors of acute otitis media in the first year of life: a prospective study, An. Pediatr. (Barc.) 60 (2004) 133–138. [9] A.W. Cripps, D.C. Otczyk, J.M. Kyd, Bacterial otitis media: a vaccine preventable disease? Vaccine 23 (2005) 2304–2310. [10] M.M. Rovers, G.A. Zielhuis, K. Ingels, G.J. van der Wilt, Day-care and otitis media in young children: a critical overview, Eur. J. Pediatr. 158 (1999) 1–6. [11] R.E. Cervio, H.E. Maccarone, Otitis media aguda: algunos aspectos de su diagnostico y tratamiento, Arch. Argent. Pediatr. 105 (2007) 43–49. [12] E.R. Wald, B. Dashefsky, C. Byers, N. Guerra, F. Taylor, Frequency and severity of infections in day care, J. Pediatr. 112 (1988) 540–546. [13] B.C. Decker, in: C.D. Bluestone, J.O. Klein (Eds.), Otitis Media in Infants and Children, 4th ed., Hamilton, Ontario, 2007. [14] World Health Organization, World Development Report 1993: Investing in Health, Oxford University Press, Oxford, 1993, pp. 215–222. [15] S. Berman, Otitis media in developing countries, Pediatrics 96 (1995) 126–131. [16] J. Bondy, S. Berman, J. Glazner, D. Lezotte, Direct expenditures related to otitis media diagnoses: extrapolations from a pediatric medicaid cohort, Pediatrics 105 (2000) E72. [17] A.C. Nyquist, R. Gonzales, J.F. Steiner, M.A. Sande, Antibiotic prescribing for children with colds, upper respiratory tract infections, and bronchitis, JAMA 279 (1998) 875–877. [18] C.A. Arevalo Silva, A. Villasenor Sierra, P. Kuri Morales, R. Tapia Conyer, J.I. Santos Preciado, Otitis Media Aguda en Me´xico: casos reportados durante el periodo de 1995 a 1998, Gac. Med. Mex. 135 (1999) 541–543. [19] E. Elm, D.G. Altman, M. Egger, S.J. Pocock, P.C. Gotzsche, J.P. Vandenbroucke, The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies, Lancet 370 (2007) 1453–1457. [20] D.F. Stroup, J.A. Berlin, S.C. Morton, I. Olkin, G.D. Williamson, D. Rennie, Metaanalysis of Observational Studies in Epidemiology: a proposal for reporting. Metaanalysis of Observational Studies in Epidemiology (MOOSE) group, JAMA 283 (15) (2000) 2008–2012. [21] J.P.T. Higgins, S. Green (Eds.), Cochrane Handbook for Systematic Reviews of Interventions 4.2.5, The Cochrane Library, Issue 3, John Wiley & Sons, Ltd, Chichester, UK, 2005. [22] S. Sanderson, I.D. Tatt, J.P. Higgins, Tools for assessing quality and susceptibility to bias in observational studies in epidemiology: a systematic review and annotated bibliography, Int. J. Epidemiol. 36 (2007) 666–676. [23] F. Fowkes, P. Fulton, Critical appraisal of published research: introductory guidelines, Br. Med. J. 302 (1991) 1136–1140. [24] M.F. Freeman, J.W. Tukey, Transformations related to the angular and the square root, Ann. Math. Stat. 21 (1950) 607–611. [25] R. DerSimonian, N. Laird, Meta-analysis in clinical trials, Control. Clin. Trials 7 (1986) 177–188. [26] J.P.T. Higgins, S.G. Thompson, J.J. Deeks, D.G. Altman, Measuring inconsistency in meta-analyses, Br. Med. J. 327 (2003) 557–560. [27] B.I. Lopez, B.H. Sepulveda, I.I. Valdes, Frecuencia de otitis media aguda en menores de 5 an˜os bajo control, Pediatrı´a (Santiago de Chile) 41 (1998) 21–25. [28] E. Nandi-Lozano, L.E. Espinosa, L. Vinas-Flores, C. Avila-Figueroa, Incidence of acute respiratory infections in a cohort of infants and children attending a daycare center in Mexico City, Salud Publica de Mexico 44 (2002) 201–206 (in Spanish). [29] A. Rosenblut, M.E. Santolaya, P. Gonzalez, V. Corbalan, L.F. Avendano, M.A. Martinez, et al., Bacterial and viral etiology of acute otitis media in Chilean children, Pediatr. Infect. Dis. J. 20 (2001) 501–507. [30] A. Arguedas, R. Dagan, C. Soley, C. Loaiza, K. Knudsen, N. Porat, et al., Microbiology of otitis media in Costa Rican children, 1999 through 2001, Pediatr. Infect. Dis. J. 22 (2003) 1063–1068. [31] G. Gutierrez Trujillo, Md.C. Martinez Gonzalez, H. Guiscafre Gallardo, G. Gomez, A. Peniche, O. Munoz, Patrones de prescripcio´n de antimicrobianos en infecciones respiratorias agudas: incidencia en la poblacio´n rural, Rev. Fac. Med. Univ. Nac. Colomb. 32 (1989) 134–136.

[32] A. Arguedas, C. Loaiza, A. Perez, F. Vargas, M. Herrera, G. Rodriguez, et al., Microbiology of acute otitis media in Costa Rican children, Pediatr. Infect. Dis. J. 17 (1998) 680–689. [33] A. Arguedas, R. Dagan, S. Guevara, S. Guevara, E. Brilla, Middle ear fluid Streptococcus pneumoniae serotype distribution in Costa Rican children with otitis media, Pediatr. Infect. Dis. J. 24 (2005) 631–634. [34] R. Commisso, R. Romero Dı´az, F. Romero Moroni, F. Romero Orellano, D. Centeno, J. Escalera (h), et al., Bacteriologı´a y sensibilidad antibio´tica en otitis media aguda, Arch. Argent. Pediatr. 104 (2006) 333–338. [35] G.J. Cofre, R.A. Rosenblut, M.C. Borel, A.P. Gonza´lez, M.A. Va´squez, A.M.T. Siri, et al., Eficacia clı´nica de la asociacio´n amoxicilina/sulbactam (4:1) en otitis media aguda en nin˜os, Rev. Chil. Infectol. 21 (2004) 186–198. [36] E. Desio, S. Lopez Mina, O. Meder, R. Pa´ez, A. Salinas, V. Zanzi, Actividad terapeutica de la asociacio´n amoxicilina-sulbactam administrada cada 12 horas en ninos con otitis media aguda: estudio multicentrico, Rev. Enferm. Infecc. Ped. 14 (2000) 16–22. [37] A. Morayta Ramirez, E. Gonzalez Arenas, C. Gomez Altamirano, J. Gutie´rrez Mun˜oz, A. Ramos Garcı´a, Eficacia y seguridad de amoxicilina/sulbactam en nin˜os con otitis media aguda, Rev. Enferm. Infecc. Ped. 14 (2000) 50–54. [38] N. Orlando, L. Torrado, M. Turco, Prevalence of bacterias isolated from middle ear fluid in Argentina, in: First World Congress of Pediatric Infectious Diseases, Fifteenth InterAmerican Congress of Pediatric Infectious Diseases Acapulco, Mexico, 1996 [Abstract E186]. [39] N. Orlando, M. Vazquez, A. Merediz, L. Torra- D.O., A. Procopio, E.L. Lo´pez, Prevalence and antibiotic susceptibility of organisms isolated from acute otitis media in Argentinean children, in: 9th International Congress on Infectious Diseases, Bue- nos Aires, Arg., (2000), pp. 10–13, Abstract # 76.020. [40] T.M. Sih, Acute otitis media in Brazilian children: analysis of microbiology and antimicrobial susceptibility, Ann. Otol. Rhinol. Laryngol. 110 (7 (Pt 1)) (2001) 662–666. [41] H. Trujillo, R. Callejas, G.I. Mejia, L. Castrillon, Bacteriology of middle ear fluid specimens obtained by tympanocentesis from 111 Colombian children with acute otitis media, Pediatr. Infect. Dis. J. 8 (1989) 361–363. [42] A. Villasenor-Sierra, J.I. Santos, Outer membrane protein profiles of paired nasopharyngeal and middle ear isolates of nontypeable Haemophilus influenzae from Mexican children with acute otitis media, Clin. Infect. Dis. 28 (1999) 267– 273. [43] R.A. Morayta, S.N. Gonzalez, H.L. Alvarez, A.C. Go´mez, Estudio multice´ntrico de la eficacia y seguridad de amoxicilina/sulbactam en nin˜os con otitis media aguda, Rev. Enferm. Infecc. Ped. 14 (2001) 127–135. [44] I. Lopez Bravo, B.H. Sepulveda, C.F. Guerra, S.R. Nazar, P. Valle, Perfil de morbilidad anual de menores de 6 an˜os consultantes, Rev. Chil. Pediatr. 71 (2000) 321–327. [45] J.L. Paradise, H.E. Rockette, D.K. Colborn, B.S. Bernard, C.G. Smith, M. Kurs-Lasky, Otitis media in 2253 Pittsburgh-area infants: prevalence and risk factors during the first two years of life, Pediatrics 99 (1997) 318–333. [46] O.P. Alho, M. Koivu, M. Sorri, P. Rantakallio, P. Rantakallio, The occurrence of acute otitis media in infants. A life-table analysis, Int. J. Pediatr. Otorhinolaryngol. 21 (1991) 7–14. [47] M.T. Valenzuela, R. O’Loughlin, F. De La Hoz, E. Gomez, D. Constenla, A. Sinha, The burden of pneumococcal disease among Latin American and Caribbean children: review of the evidence, Rev. Panam. Salud Publica 25 (2009) 270–279. [48] D.O. Constenla, Economic impact of pneumococcal conjugate vaccination in Brazil, Chile, and Uruguay, Rev. Panam. Salud Publica 24 (2008) 101–112. [49] E. Leibovitz, M.R. Jacobs, R. Dagan, Haemophilus influenzae: a significant pathogen in acute otitis media, Pediatr. Infect. Dis. J. 23 (2004) 1142–1155. [50] C. Marrs, G.P. Krasan, K.W. McCrea, D.L. Clemans, J.R. Gilsdorf, Haemophilus influenzae – human specific bacteria, Front. Biosci. 6 (2001) e41–e60 (September 1). [51] R. Commisso, F. Romero-Orellano, P.B. Montanaro, F. Romero-Moroni, R. RomeroDiaz, Acute otitis media: bacteriology and bacterial resistance in 205 pediatric patients, Int. J. Pediatr. Otorhinolaryngol. 30 (56) (2000) 23–31. [52] A. Rosenblut, M.E. Santolaya, P. Gonzalez, C. Borel, J. Cofre, Penicillin resistance is not extrapolable to amoxicillin resistance in Streptococcus pneumoniae isolated from middle ear fluid in children with acute otitis media, Ann. Otol. Rhinol. Laryngol. 115 (2006) 186–190. [53] W.P. Hausdorff, G. Yothers, R. Dagan, T. Kilpi, S.I. Pelton, R. Cohen, et al., Multinational study of pneumococcal serotypes causing acute otitis media in children, Pediatr. Infect. Dis. J. 21 (2002) 1008–1016. [54] Informe Regional de SIREVA II: datos por paı´s y por grupos de edad sobre las caracterı´sticas de los aislamientos de Streptococcus pneumoniae, Haemophilus influenzae y Neisseria meningitidis, en procesos invasores, 2000–2005, Publicaciones Electronicas 2006–7: OPS/OMS, 2006. [55] D. Constenla, Evaluating the costs of pneumococcal disease in selected Latin American countries, Rev. Panam. Salud Publica 22 (2007) 268–278. [56] D. Constenla, E. Gomez, F. de la Hoz, Rosalyn O’Loughlin, Anushua Sinha, Juan E. Valencia, et al., The burden of pneumococcal disease and the cost effectiveness of a pneumococcal vaccine in Latin America and the Caribbean: a review of the evidence and a preliminary economic analysis, Albert B. Sabin Vaccine Institute, Washington, DC, 2007. [57] A. Sinha, D. Constenla, J.E. Valencia, R. O’Loughlin, E. Gomez, F. de la Hoz, et al., Cost-effectiveness of pneumococcal conjugate vaccination in Latin America and the Caribbean: a regional analysis, Rev. Panam. Salud Publica 24 (2008) 304– 313. [58] S.M. Schappert, Office visits for otitis media: United States, 1975–90, Adv. Data 214 (1992) 1–19.