Cryptococcal meningitis in Rio de Janeiro State , Brazil , 1994-2004

The objective of this article was to evaluate the epidemiology of cryptococcal meningitis in Rio de Janeiro State, Brazil, from 1994 to 2004. Six hundred and ninety-six cases of cryptococcal meningitis were reported, with a mean incidence of 0.45 per 100,000 inhabitants. Patients were predominantly male; mean age was 35.9 years; AIDS was practically the only underlying disease, reported in 61.2% of cases; case-fatality was 51.8%. No decline in incidence was observed during the study period. AIDS is the main predisposing condition for cryptococcal meningitis, and thus the profile of most patients mirrors that of HIV infection. Missing information prevented the evaluation of other underlying diseases. Cryptococcal Meningitis; Incidence Rate; Epidemiologic Surveillance Introduction The principal etiological agent of fungal meningitis in the world is an encapsulated yeast that presents two distinct species recognized in the Cryptococcus neoformans complex: C. neoformans and C. gattii 1,2. C. neoformans is distributed worldwide and predominantly affects immunocompromised individuals. C. gattii infection generally affects immunocompetent individuals and predominates in tropical and subtropical regions. The disease is endemic in Northeast Brazil, Paraguay, Mexico, Australia, several African countries, India, Southeast Asia, and California (USA) 3. Due to the important tropism of this fungus for the central nervous system, meningitis is the principal clinical manifestation in some 85% of cases of cryptococcal infection 4,5. Until the late 1970s, cryptococcal disease was described in patients with underlying diseases like lymphoma, sarcoidosis, liver disease, and kidney transplant patients 6, and was considered a rare condition. Beginning in the 1980s, with the emergence of HIV infection, cryptococcal disease acquired a relevant role as an opportunistic infection, with increasing morbidity and mortality rates 6,7. The use of prophylactic antifungals and the introduction of HAART (highly active antiretroviral therapy) in 1995-1996 modified the course of cryptococcal disease in HIV patients, with reductions in the incidence rates varying from 46% to 75% 7,8,9,10,11. Despite a decrease in the numCad. Saúde Pública, Rio de Janeiro, 24(11):2582-2592, nov, 2008 2582 ARTIGO ARTICLE


Introduction
The principal etiological agent of fungal meningitis in the world is an encapsulated yeast that presents two distinct species recognized in the Cryptococcus neoformans complex: C. neoformans and C. gattii 1,2 .C. neoformans is distributed worldwide and predominantly affects immunocompromised individuals.C. gattii infection generally affects immunocompetent individuals and predominates in tropical and subtropical regions.The disease is endemic in Northeast Brazil, Paraguay, Mexico, Australia, several African countries, India, Southeast Asia, and California (USA) 3 .Due to the important tropism of this fungus for the central nervous system, meningitis is the principal clinical manifestation in some 85% of cases of cryptococcal infection 4,5 .
Until the late 1970s, cryptococcal disease was described in patients with underlying diseases like lymphoma, sarcoidosis, liver disease, and kidney transplant patients 6 , and was considered a rare condition.Beginning in the 1980s, with the emergence of HIV infection, cryptococcal disease acquired a relevant role as an opportunistic infection, with increasing morbidity and mortality rates 6,7 .The use of prophylactic antifungals and the introduction of HAART (highly active antiretroviral therapy) in 1995-1996 modified the course of cryptococcal disease in HIV patients, with reductions in the incidence rates varying from 46% to 75% 7,8,9,10,11 .Despite a decrease in the num-ber of AIDS cases, HIV is still the principal risk factor for cryptococcosis.According to data from the Centers for Disease Control and Prevention (CDC), 85% of cryptococcosis cases in the United State occurred in HIV patients 12 , while in France, Chrétien et al. 13 report 80%.An estimated 5% to 10% of HIV patients have cryptococcosis as their AIDS-defining disease 14 .
In addition to HIV infection, the principal predisposing conditions for cryptococcal disease are those that also involve cellular immunodeficiency, like treatment for hematological cancer, the use of immune suppressors in transplant patients, and prolonged corticosteroid therapy 5,15 .Other conditions not related directly to cellular immunity, like diabetes, alcoholism, and sarcoidosis, are also defined as risk factors for cryptococcosis 16 .
In Brazil, Almeida & Lacaz 7 reported the first case of cryptococcosis in 1941.Oliveira-Netto et al. 4 , in a retrospective study of 302 cases of cryptococcosis published in the Brazilian literature from 1941 to 1992, reported that 179 (59.3%) had occurred in the last 10 years.In 72% of the 153 patients with some diagnosed predisposing condition, the latter was AIDS.As elsewhere in the world, in Brazil the introduction of HAART in 1995-1996 and the use of prophylactic antifungals led to an important decrease in cryptococcosis incidence 7,9 .
There are few data on the epidemiology of cryptococcosis in the various States of Brazil.The published articles refer to single hospital-based studies that are complete 17,18 but do not reflect the profile of the disease in the general population.
In the State of Rio de Janeiro, Brazil, the available studies on cryptococcal meningitis were published in 1984 19 and 1994 20 , and there is still no information on incidence temporal trends in the State or repercussions from the introduction of HAART for AIDS patients or as a result of the increase in the number of transplants.
The objective of the current study was to determine the epidemiological profile of cryptococcal meningitis in the State of Rio de Janeiro from 1994 to 2004.

Material and methods
This was a descriptive exploratory study on the incidence of cryptococcal meningitis in the State of Rio de Janeiro from 1994 to 2004.Data on cryptococcal meningitis in the State are available in two databases: the Advisory Board on Meningitis, a division of the Epidemiological Surveillance Center of the Rio de Janeiro State Health Depart-ment, and in the Information System on Notifiable Diseases (SINAN).Both databases are fed by information from the "individual meningitis investigation form".This data collection instrument does not have a field with a specific code for cryptococcal meningitis, and the latter is recorded as open information in the field "meningitis of another etiology".Considering that the database of the Advisory Board on Meningitis covers a longer time period, we chose to use it as the data source.
The Advisory Board on Meningitis receives notifications of all meningitis cases from the Municipal Health Departments and directly from the Epidemiological Surveillance Unit of the São Sebastião State Institute of Infectology (IEISS), which is the referral center for meningitis diagnosis in the State.Cases reported by the IEISS include both patients treated at the Institute itself and those whose cerebrospinal fluid (CSF) samples are collected at other units and analyzed at the IEISS laboratory.
The criterion adopted in this study for definition of an incident case was the first notification with diagnostic confirmation of cryptococcal meningitis present in the database during the period.Subsequent diagnoses for the same individual during the study were considered as relapse of cryptococcal meningitis.
The study variables were: age, gender, municipality (county) of residence, residential zone (rural versus urban), preexisting diseases, type of diagnostic confirmation (direct India ink test, CSF culture of Cryptococcus sp. or latex agglutination, clinical, therapeutic challenge, autopsy, death certificate), and type of evolution (categorized as discharge, death, or death from another cause).In relation to age, individuals under 20 years were stratified in two strata: 0-12 and 13-19 years.The first age bracket included children up to 12 years of age, in accordance with the national 21,22 and international literature 23 .To complement the information on preexisting diseases, AIDS notification data were accessed in the SINAN STD/AIDS database.Due to the high missing information rate, the race/color variable was not analyzed.
Excel and Access (Microsoft Corp., USA) and SPSS (SPSS Inc., Chicago, USA) were used for the data analysis.The proportions were compared using the chi-square test, means using the Student's t test, and temporal trends using simple linear regression.Significance level was set at 5%.

Results
In the State of Rio de Janeiro, according to the cases reported to the Advisory Board on Meningitis from 1994 to 2004, 696 new cases of cryptococcal meningitis were recorded, representing a mean annual incidence rate of 0.45 cases per 100 thousand inhabitants, varying from 0.30 in 1994 to 0.58 per 100 thousand inhabitants in 2003 (Table 1).Gender distribution showed a predominance of males during the study period (male/female ratio: 2.6), with values ranging from 62.3% in 2000 (male/female ratio: 1.6) to 82.4% in 1995 (male/ female ratio: 4.0) (Table 1).Place of residence was urban in 98.9% of cases.
Incidence rates according to municipality of residence are shown in the map in Figure 1.Ninety-five percent of cases (656) were observed in the Metropolitan meso-region, with 92.9% of these in the Rio de Janeiro micro-region.Of the 13 municipalities with reported cases, from the total of 16 in the micro-region, the municipality or city of Rio de Janeiro contributed with more than half (59.9%) of the total of 696 cases in the State, followed by Duque de Caxias with 40 cases (7.1%), Niterói with 42 (6.1%),São Gonçalo with 33 (4.8%),São João de Meriti with 31 (4.5%), and Nova Iguaçu with 29 (4.2%).Each of the other municipalities had fewer than 2.5% of the cases.Cases were distributed homogeneously over the months of the year during the study period.
The most frequent diagnostic confirmation for the study period as a whole was culture (42.8%), followed by direct India ink test (26.3%) and latex agglutination (23.7%).Death certificate was the criterion in 6.6% of cases, autopsy in 0.4%, and therapeutic challenge in 0.1%.Beginning in 2001, there was an upward trend in the proportion of diagnoses by India ink test and by latex agglutination, and beginning in 2003, latex agglutination became the principal confirmatory criterion.
The entire sample of cases showed a mean age of 35.9 years and median of 35 years.There was no significant difference in mean age between genders (males: 36.06;females: 35.48; p = 0.58).Age ranged from 4 months to 92 years.Age was unknown in four cases.Except for one 89-year-old patient and two 92-year-olds, the maximum age for both genders was 70.Among individuals under 12 years, three were less than a year old (four, six, and eight months), while the others varied from 2 to 11 years, with a mean and median of 6 years.Age distribution remained homogeneous throughout the period.
Analysis of case distribution by age strata according to gender showed that for both males and females, the largest number of cases was in the 30-39-year bracket, with 41% of the male cases and 28.4% of the females.However, when considering the incidence rate per 100 thousand inhabitants, in males the highest rate (1.74) was in this same age bracket (30-39 years), while among females the highest rate (0.45) was in the 40-49year bracket.In the age brackets up to 19 years, the incidence rates were similar (0.06 in females and 0.07 in males).In females, the incidence rate in the over-60 bracket was as low as in the brackets up to 19 years (0.07) (Table 2).The high proportion of missing data jeopardized the analysis of preexisting diseases.Among the cases reported to the Advisory Board on Meningitis with a diagnosis of cryptococcal meningitis from 1994 to 2004, AIDS was the most frequent preexisting disease, present in 178 patients (25.6% overall and 98.3% for cases with information on preexisting disease).When the cryptococcal meningitis database was overlapped with the AIDS reporting data from the SINAN STD/ AIDS database, information on the variable AIDS as preexisting disease reached 426 cases, representing 99.3% of the cases with information on preexisting disease (61.2% of total patients).The percentage of missing information on preexisting disease remained high (35.8% in males and 45.3% in females).Except for one case of leukemia in the 30-39-year bracket and two cases of diabetes in the over-60 bracket, AIDS was the only preexisting disease recorded.The proportion of patients diagnosed with cryptococcal meningitis with AIDS as preexisting disease varied from 47.5% in 1994 to 72.7% in 1998.
Figure 2 shows the AIDS incidence rates (per 100 thousand inhabitants), according to Boletim Epidemiológico AIDST 24 , and the cryptococcal meningitis incidence rates from 1994 to 2004.
The incidence adjustment model for the period identified an upward trend in AIDS (p = 0.03) and stability in meningitis incidence (p = 0.98).
The temporal relationship between meningitis diagnosis and AIDS notification showed concordance with the AIDS notification year for the two events in 59.7% of cases, while in 16.7% of the cases meningitis occurred in years prior to AIDS notification, varying from 1 to 4 years prior, while in 23.6% of cases meningitis occurred after AIDS notification, varying from 1 to 11 years later.In 12.5% of meningitis cases with AIDS (53 patients), information on the year of AIDS notification was missing.
Of the total of 666 patients with cryptococcal meningitis (95.7%) in which the outcome was known, 345 (51.8%) evolved to death from meningitis and 8 (1.2%) to death from another cause.Temporal trend analysis for cryptococcal meningitis case-fatality showed a downward trend during the period, without statistical significance (p = 0.42) (Figure 3).
Case-fatality was high in all age brackets, showing a progressive increase with age: 25% under 12 years, approximately 47% in the 13-29-year bracket, 54% in the 30-59-year bracket, and 68% in the over-60 group.Except for individuals up to 19 years and over 60, in which the small number of cases precludes evaluation, case-fatality was stable over the period.
Time elapsed between hospitalization and discharge or death varied from 1 to 229 days.Analysis of time of evolution showed that 13% of the deaths occurred on the first day of hospitalization and 58.7% during the first 14 days.
Only three percent of patients presented relapse (n = 21).Time elapsed between hospital discharge and readmission due to relapse varied from 2 to 45 months, and 90% of cases with relapse had AIDS..Incidence rates (per 100,000 inhabitants) for cryptococcal meningitis and AIDS.Annual incidence rates in the State were consistent with CDC data that indicates that the annual cryptococcosis incidence rate varies from 0.4 to 1.3 cases per 100 thousand inhabitants.The municipalities with the most cases of cryptococcal meningitis recorded by the Advisory Board on Meningitis coincide with the ones that according to Ministry of Health data 26 showed the most reported AIDS cases during the study period.Considering that HIV infection is still the principal risk factor for cryptococcal disease, this concordance was expected.
Direct CSF microscopy with India ink is the most practical and economical diagnostic method for cryptococcal meningitis, with a sensitivity of 80% in patients with AIDS and 50% in HIV-negative patients 27 .Culture is the gold standard 7 , but the results are not immediate as are India ink or latex agglutination, since the colonies can only be observed after 48 to 72 hours.If the India ink test is negative and latex agglutination is not performed, this delay until identification of the etiological agent can lead to the meningitis case being classified as "unspecified etiology", in the absence of active surveillance.Sensitivity and specificity in the detection of capsular polysaccharide antigen by latex agglutination depend generally on the kit and vary from 93% to 100% and 93% to 98%, respectively 27 .The increase in the use of latex agglutination as the confirmatory criterion appears to indicate a distortion in the completion of notification forms.Considering the high case-fatality, one can conclude that patients are already quite compromised by the time they reach the health services, and it is unlikely that they would not present a positive culture for Cryptococcus.What may be happening, we believe, is that since the result of latex agglutination is released sooner, it is recorded first on the notification form, without subsequently verifying the result of the culture, which is the gold standard and as such should always be prioritized.Chayakulkeeree & Perfect 27 use the term "isolated cryptococcal polysaccharidemia" to describe a condition observed in high-risk patients, i.e., extremely immunodepressed, that present a positive cryptococcal antigen titer without positive culture or important symptoms.They conclude that these patients would probably benefit from antifungal therapy to prevent or delay the development of cryptococcosis.However, this does not appear to be the profile of the patients in the Rio de Janeiro State health services, who generally seek (or rather, manage to obtain) treatment when they already display important symptoms.
Our study did not show a decrease in the absolute number of cryptococcal meningitis cases or in the incidence per 100 thousand inhabit-ants over the period, as reported in the literature 8,28,29 .The authors report that this reduction is due to a decline in incidence among HIV-positive patients, since the incidence among HIV-negative patients did not change during the period 8,27,28 .Studies have shown that prophylactic antifungals and especially the advent of HAART have led to a 55% reduction in absolute cryptococcosis cases in the Netherlands 30 , a 70% decrease in incidence in the general population in Atlanta 28 , a 63% reduction in incidence in HIV-positive patients in San Francisco 10 , and a 46% drop in HIV-positive individuals in France 8 .Authors of the French study highlight that the decrease occurred mainly in the part of the population with regular access to the health system (Europeans), as opposed to a significant increase among patients with little access to the system (patients of African origin).They conclude that in the post-HAART era, lack of diagnosis and/or treatment of HIV infection is associated with increased risk of cryptococcosis.A study by Saraceni et al. 31 on AIDS mortality in Rio de Janeiro from 1995 to 2003 revealed deficient access to HIV diagnosis, impeding the early introduction of HAART, before the development of immune deficiency, as a basis for the occurrence of opportunistic infections.They highlight that patients frequently come to emergency services without knowledge of their HIV infection and already presenting opportunistic infections.In Brazil, data from the Emílio Ribas Institute in São Paulo showed a reduction in cryptococcal meningitis in AIDS patients, from 7.7% in 1995 to 3.1% in 2002 7 .At the Evandro Chagas Clinical Research Institute, Oswaldo Cruz Foundation, in Rio de Janeiro, Pinto Jr. 32 observed a reduction (from 7.6% in 1987-1994 to 1.8% in 1999-2002) in the incidence of cryptococcal meningitis among AIDS patients.These data appear to indicate that in patients monitored by well-structured services, allowing early access to HAART, there can be an important reduction in cryptococcosis, at rates similar to those reported in the international literature.Our study showed 61.2% of cryptococcosis patients with HIV infection.However, despite the harmonization with the SINAN STD/AIDS database, some patients may not have been identified due to limitations in the data cross-analysis and the existing underreporting in the STD/AIDS database itself 33 .Thus, it was not possible to conduct a segmentation for AIDS, comparing it with non-AIDS, which would have allowed us to make inferences for each of these segments, including for the role of HAART.
In relation to other preexisting diseases, only three cases referred to other predisposing conditions for cryptococcal disease: leukemia Cad.Saúde Pública, Rio de Janeiro, 24 (11):2582-2592, nov, 2008 (one) and diabetes (two).We believe that lack of investigation of preexisting diseases, as we verified in relation to AIDS, is responsible for this low number of patients with other predisposing conditions.The importance of other conditions has increased in the last 10 to 15 years 5 , especially in developed countries where there has been extensive development in the field of transplants and immune-manipulating therapies 27 .The international literature reports percentages as high as 74.6% 8 or 82% 28 for the presence of underlying conditions like diabetes or cancer in patients without HIV infection.In Brazil, different authors highlight the relevance of predisposing conditions for the development of cryptococcal meningitis.Livramento et al. 34 , for the period from 1984 to 1992, observed some predisposing condition in 60% of cryptococcal meningitis patients without HIV infection, with renal transplant as the most frequent (73.3%), and Carvalho 17 , for the period from 1980 to 2000, reported 78.3% of cases having predisposing conditions, mainly diabetes and cancer.The series by Rozenbaum & Gonçalves 20 in the State of Rio de Janeiro (1960 to 1989) showed 29.6% of cases with another predisposing condition among patients without AIDS, half of whom in use of immune suppressing drugs.
Predominance of male gender has been reported consistently in cryptococcal disease 35 .In the early AIDS epidemic, the number of male cases of cryptococcosis reached eight times that of females, accompanying the HIV infection profile.In recent years, with the increase in AIDS among women, the distribution between the sexes is changing 5 .Our study showed a male/female ratio of 2.6:1, close to the 2:1 ratio in the State of do Rio de Janeiro during the same period for AIDS, the principal risk factor for cryptococcosis.The fact that we did not observe a difference in the mean age between men and women, unlike reports from the literature, in which female cases are significantly younger 5,8 , mirroring the age profile for HIV infection, could be explained by the persistent increase observed since 1998 in AIDS incidence among Brazilian women over 24 years of age 36 .
In our study, 2.3% of patients were in the under-12-year bracket and 25% had AIDS as a known preexisting disease.In the series by Rozembaum & Gonçalves 20 from 1960 and 1989, 1.8% of the patients (3/171) were under 12 years of age, one of whom had AIDS.In pediatric patients with AIDS, Py et al. 22 in Rio de Janeiro observed 8.3% of cases of cryptococcal meningitis.In Brazil, cryptococcal meningitis in children has been well studied in the Northeast Region, where C. gattii is endemic.Thus, Martins 18 in 53 cases under 14 years of age, Corrêa et al. 21in 19 cases under 13 years, and Cavalcanti 37 in 17 cases under 10 years and 27 from 11 to 20 years of age found no cases of AIDS and observed a significant predominance of isolation of C. gattii.According to the international literature, cryptococcosis is rare in childhood 1 , and the proportion of cryptococcal disease among children with HIV is much lower than in adults 12 .Abadi et al. 23 observed 1.4% prevalence of cryptococcal infection in American children with AIDS from 1985 to 1996.During this same period, data from studies in Europe and the United States reported 2.9% and 13.3% prevalence, respectively, in adults 11 .Goldman et al. 38 , with the objective of determining whether the low prevalence of cryptococcal disease in childhood is due to low exposure to Cryptococcus, analyzed serum samples from 185 children in the Bronx, New York City, and observed that for the majority, age at seroconversion was under ten years.Based on their results, they concluded that the low incidence of cryptococcosis in children with AIDS is not due to lack of exposure to the agent.Dromer et al. 8 reported 0.9% of patients younger than 15 years among 2,125 cases of cryptococcal infection from 1985 to 2001, 31% of whom were HIV-positive and 31% immunosuppressed.Chen et al. 1 , from 1994 to 1998, observed 0.6% of children among 350 patients with cryptococcal infection.The high percentage of children with cryptococcosis in our study and also reported by Py et al. 22 as compared to the international literature could result from greater environmental pressure for exposure to both species of Cryptococcus, both in patients with HIV infection and in immunocompetent patients.
We did not observe a change in case-fatality from cryptococcal meningitis during the period studied.In Brazil, cryptococcosis case-fatality with or without underlying disease ranges from 45% to 65% 7,17,39 .In the series by Rozembaum & Gonçalves 20 in Rio de Janeiro from 1960 to 1989, overall case-fatality in the sample was 52.3% (65.7% in AIDS patients, 38.6% in non-immune suppressed patients, and 36.8% in patients with other predisposing conditions).Case-fatality greater than 40% is close to that observed in African countries 11 .For AIDS, the principal risk factor for cryptococcosis, a reduction in case-fatality was observed in the State of Rio de Janeiro beginning in 1997, shortly after the introduction of HAART, with subsequent stabilization beginning in 1999 36 .Evidence suggests that combination antiretroviral treatment can also alter the evolution of AIDS-associated conditions, since this treatment partially recovers the patient's immune defenses 9 .
Cad. Saúde Pública, Rio de Janeiro, 24(11):2582-2592, nov, 2008 We observed a low relapse rate, which may suggest the efficacy of adequate maintenance therapy, equaling international standards 40,41 .However, considering the precarious state of health services, we cannot rule out the possibility that the surveillance system is failing to detect cases of relapse.According to Mamidi et al. 42 , even after effective initial treatment, relapse occurs in 25% to 60% of HIV-infected patients with cryptococcal meningitis if a maintenance regimen is not implemented.This regimen, proposed by the Infectious Diseases Society of America, consists of the administration of an antifungal agent for one year in patients without HIV infection, and in those with AIDS, administration of an antifungal for at least two years and suspension if there is sustained immune recovery with HAART 27,43 .With the introduction of maintenance therapy, the risk of relapse dropped to values ranging from 2% to 4% 39,40 .
Two key aspects should be considered when analyzing the results of this study: underreporting of cases in the epidemiological surveillance system and the high number of patients with unknown preexisting disease.The underreporting of various diseases is well known in Brazil 33,36 and should always be taken into consideration.As for cryptococcal meningitis, an active search showed 24.4% of cases not recorded by the Ad-visory Board on Meningitis from 2000 to 2004 25 .As for preexisting diseases, in conditions like cryptococcosis, in which the underlying diseases that affect cellular immunity play a central role, missing information hinders the identification of patients at risk and negatively affects risk factor assessment over time.The proportion we managed to detect of cryptococcal meningitis cases with HIV (61.2%) is close to the 63% observed by Livramento et al. 34 in São Paulo from 1984 to 1992, but is lower than reported by Carvalho 17 in Ribeirão Preto from 1980 to 2000, where 73.1% of patients with cryptococcosis were HIV-positive.Missing information also prevents the evaluation of cryptococcosis incidence in immunocompetent patients.
The exploratory analysis of the database from the Advisory Board on Meningitis at the Rio de Janeiro State Health Department, although subject to the limitations discussed above, allowed an evaluation of the trends in cryptococcal meningitis in the State.The incidence adjustment model for the period identified stability in meningitis incidence.AIDS is still the greatest risk factor, and thus the profile for the majority of patients accompanies that of HIV-infected individuals.Meanwhile, missing information jeopardizes the evaluation of other preexisting diseases and cryptococcosis in immunocompetent individuals.
Meningite Criptocócica; Taxa de Incidência; Vigilância Epidemiológica Contributors B. C. Q. Leimann was responsible for the literature review, data analysis, and organization and writing of the article.R. J. Koifman contributed to the data analysis and organization and writing of the article.

Table 1
Percentage distribution of cryptococcal meningitis cases by gender, and incidence rates (per 100,000 inhabitants) from 1994 to 2004.Cryptococcal incidence rates (per 100,000 inhabitants) by meso-region, micro-region, and municipality in the State of Rio de Janeiro, Brazil, from 1994 to 2004.

Table 2
Distribution of cryptococcal meningitis cases and incidence rates (per 100,000 inhabitants) by gender for seven age strata from 1994 to 2004.
DiscussionIn this study, which aimed to trace the profile of patients with cryptococcal meningitis in the State of Rio de Janeiro, we opted to use the database of the Rio de Janeiro State Advisory Board on Meningitis, since it was more comprehensive than the SINAN database.SINAN only had data on cryptococcal meningitis starting in the year 2000, and from 2000 to 2004, it captured 82% of the cases reported to the Advisory Board on Meningitis