Epidemiology of bacterial meningitis among children in Brazil , 1997-1998 *

Correspondence to: Débora PL Weiss, MS, RPh Merck & Co., Inc. 10 Sentry Pkwy, Mail Stop, BL 1-7 Blue Bell, PA 19422, USA E-mail: debora_petersonleite@merck.com *Partially supported by Merck & Co., Inc. West Point, PA, USA. Submitted on 5/6/2000. Reviewed on 2/3/2001. Approved on 19/3/2001.


INTRODUCTION
Bacterial infections remain an important cause of morbidity and mortality in children despite the availability of highly effective antibiotics.In many countries where vaccination against Haemophilus influenzae type b (Hib) disease is not yet widely adopted, this organism is responsible for a substantial fraction of the cases of meningitis in children under five years of age. 4,16,20,23In the United States, before the era of Hib vaccines, more than 70% of bacterial meningitis among children less than five years old was attributable to Hib. 19 Published data from the U.S. show that the incidence of Hib meningitis among children under 5 years of age decreased from 25 cases per 10, 5 in 1984, to less than 2.5 cases per 10 5 in 1993 after the introduction of conjugate vaccines since 1988. 21 Brazil, several studies have been performed over the past ten years evaluating the etiology, treatment, or outcomes of bacterial meningitis in newborns and children. 2,5,7,10,14,18However, only limited information on the etiology, incidence and treatment outcomes of bacterial meningitis in geographically defined regions are available.The present study was conducted to provide an overview of the etiology, incidence rates, antimicrobial use, and outcome of reported bacterial meningitis cases among children in Campinas, SP, Brazil, during a two-year period immediately preceding the implementation of the Hib vaccination as a part of the national immunization program in July 1999.The information may serve as a baseline for further studies to evaluate the effectiveness of the Hib vaccination program.

Geography and study population
In 1998, Campinas, in southeast Brazil, had 930,659 residents of which 316,570 were under 20 years of age (50.8% male) and 77,150 were under 5 years of age (51.3% male).*

Epidemiological surveillance
The "Secretaria de Saúde do Município de Campinas" (Campinas Health Department) was in charge of the epidemiological surveillance.There was a mandatory requirement for all cases of meningitis to be reported to the Campinas Health Department.All reports of meningitis arising from hospital inpatients, emergency room patients, death certificates, autopsies, outpatients, and telephone calls were recorded in a standardized meningitis abstract form.Information collected included: demographics, date of onset, previous contact with similar cases, type of contact, clinical manifestations, *Personal communication: Seade (Sistema Estadual de Análises de Dados Estatísticos); http://www.seade.gov.brlaboratory data, final diagnosis, criteria used to define diagnosis, hospitalization, prior antimicrobial use, antimicrobial treatment, and outcome.The notifications were regionally computerized and submitted to the coordinating center (CoVISA -Coordenadoria de Vigilância e Saúde, Health and Surveillance Coordinating Center), where the information was centralized into the notification system database (SINAN -Sistema de Informações de Agravos de Notificação).

Data collection
The Campinas Health Department provided a copy of the SINAN database with data on meningitis for this study.Respecting standard ethical procedures, this study reviewed all reported cases of bacterial meningitis reported in Campinas among individuals under 20 years of age, notified from January 1, 1997 to December 31, 1998.All cases whose residence was Campinas, but who were eventually seen by a health professional outside the city were reported by this system, since the complete investigation process and control measures were responsibility of the Campinas Health Department.Cases whose residence was outside Campinas were excluded.Sex, age, date of onset, laboratory data, final diagnosis, criteria used to define diagnosis, hospitalization, prior antimicrobial use, antimicrobial treatment, and outcome were extracted, reviewed and analyzed.

Case definition and diagnostics
The diagnostic criteria for bacterial meningitis included at least one of the following:

Statistical analysis
Two years of collected data (1997-1998) were analyzed for this study.Overall incidence and agespecific incidence rates were calculated in cases per 10 5 pys (person-years) according to the Campinas population during the two-year period.Tests of statistical significance were performed using chi-square, Fisher's exact test, t-Test and ANOVA and 95% confidence intervals were computed when applicable.

Epidemiology and etiology
Over the two-year period 274 cases of bacterial meningitis were found.N. meningitidis was the most common cause of bacterial meningitis (19.3%), followed by Hib, (9.5%), Streptococcus pneumoniae (S. pneumoniae) (7.7%), and other etiologic agents (8.0%).No specific organisms were isolated in 152 cases (55.5%).Table 1 shows the diagnostic criteria were used to identify the different organisms.Most of the Hib cases were identified by antigen detection with latex agglutination in the CSF.

Sex and age distribution
One hundred and sixty cases were male (58.4%).

Other etiologic agents
Other reported etiological agents included: P. aeruginosa (4 cases, age 7 days-5 months), K. pneumoniae (4 cases, age 3-7 months), and 14 cases with a variety of organisms typical of skin contaminants.
From 1997 to 1998 the overall incidence of bacterial meningitis increased by 61%, from 33.2 cases/10 5 pys to 53.4 cases/10 5 pys (Table 3).For Hib, the incidence rates overall and for children <5 were stable.The incidence rate of Hib for children under age one decreased by 62%.The incidence rates for N. meningitidis, S. pneumoniae, other etiologies, and unspecified meningitis increased across nearly all age groups and overall, ranging from an increase of 32% for N. meningitidis to 118% for other etiologies.

Morbidity and mortality
There were 24 deaths among the population studied.The total case fatality rate (CFR) was 8.8%.Seventyfive percent of deaths occurred in children less than five years old and 45.8% occurred in children less than one year old.There were four deaths among the Hib cases (CFR=15.4%,100% <5 years old, 50% <1 Among the 252 survivors, 16 (5.8%)had sequelae: three among Hib cases (11.5%, 100% under five years old, 33.3% under one year old); five among N. meningitidis cases (9.4%, 100% under five years old, 40% under one year old); three among S. pneumoniae cases (14.3%, 66.7% under one year old); and five among unspecified cases (3.3%, 40% under one year old).Data on sequelae were missing for 84 cases (30.7%).The type of sequelae was not recorded.

Hospitalization
The sum of hospitalization patient-days for all cases was 3,374.The total average length of stay was 12.4 days (SD 9.0, range 1-64).Table 4 shows the average length of stay (ALOS) and ranges of stay for each of the types of meningitis.The difference between these type-specific average lengths of stay was not statistically significant.

Antimicrobial treatment
Antimicrobial treatment was recorded for 67.5% of the cases, including 88.5% of the Hib cases.Ampicillin plus chloramphenicol was the most widely used antimicrobial treatment.Cephalosporins were the second most widely used antimicrobial treatment for all agents, except S. pneumoniae, for which penicillin was the second most widely used drug (19.0%).Table 5 provides a more detailed listing by etiologic agent and age group of the various combinations of antibiotics used.

Table 1 -
Diagnostic criteria used to identify the different etiologic organisms.

Table 2 -
Age-specific incidence rates* by category of bacterial meningitis among children in Campinas, SP, Brazil, over a two-year period(1997-1998).

Table 3 -
Age-specific incidence rates* of bacterial meningitis among children in Campinas, SP, Brazil, by year.
**% ∆ refers to percent change in the total rates from 1997 to 1998.

Table 4 -
Hospitalization of cases of bacterial meningitis among children in Campinas, SP, Brazil, from 1997-1998 by etiologic agent.

Table 5 -
Most common received antimicrobial treatments for bacterial meningitis by etiologic agents and age groups among children in Campinas,SP, Brazil, over 1997-1998.