Indian Journal of Medical Microbiology Medknow Publications on behalf of Indian Association of Medical Microbiology ISSN: 0255-0857 EISSN: 1998-3646 Vol. 23, Num. 2, 2005, pp. 95-101

Indian Journal of Medical Microbiology, Vol. 23, No. 2, April-June, 2005, pp. 95-101

Original Article

Epidemiological, Clinical and Prognostic Profile of Acute Bacterial Meningitis among Children in Alexandria, Egypt

Farag H.F.M., Abdel-Fattah M.M., Youssri A.M.

Medical Statistics and Clinical Epidemiology Department, Medical Research Institute, Alexandria University

Correspondence Address: Medical Statistics and Clinical Epidemiology Department, Medical Research Institute, Alexandria University, mezo106@hotmail.com

Code Number: mb05026

Abstract

Purpose: To address the epidemiological characteristics and clinical indices that may predict the prognostic profile of meningitis among children.
Methods: Children admitted to Alexandria fever hospital with clinical diagnosis of meningitis/meningoencephalitis during the period 2002-2003 were recruited for the study. They were subjected to clinical examination as well as CSF bacteriological and serological investigations
Results: Three hundred and ten patients (195 males and 115 females) were included. About 65.2% of them were infected with acute bacterial meningitis (ABM) and 34.8% were infected with aseptic meningitis. In this study, ABM was caused by Haemophilus influenzae (21%), Streptococcus pneumoniae (13.9%), Neisseria meningitidis (14.2%) and other undetermined bacteria (16.1%). ABM showed significant association with age group 1-9 years (66.3%), low socio-economic class (96%), working mother (83.2%), more than two smokers in the family (62.9%) and cold seasons(fall 35.1% and winter 48.5%). Aseptic meningitis showed significant association with age group 3-15 months (100%) and previous immunization(81.5%). The overall case fatality rate was 10.3%; 13.9% for ABM and 3.4% for aseptic meningitis. 7.1% of all survivors developed epileptic attacks. Predictors for death or epilepsy events were high WHO meningitis score (> 9), decreased CSF glucose level (<10 mg/dL), more smokers in the family, generalised seizures, infancy (<1year of age) and working mothers.
Conclusion: This study highlights the importance of several predictors of the outcome of meningitis in children. It is concluded that quick and simple scoring scales, such as the WHO scale, are not only applicable but valuable prognostic tools for meningitis in children.

Keywords: Meningitis, children, Egypt, risk factors, survival

Meningitis is a serious public health problem demanding early diagnosis, effective treatment, prevention and control. It is a major cause of morbidity and mortality among infants and children below the age of five years.[1]-[3] There are three main organisms that account for over 90% of the world′s cases of meningitis. These are N. meningitidis, S. pneumoniae and H. influenzae type b (Hib).[4],[5] In Africa, meningitis prevails in the semiarid sub-Saharan area between 10° and 12° North latitude from West to East Africa, dubbed the meningitis belt.[6]

Endemic meningitis among children takes the form of sporadic cases or small clusters with an endemicity rate of 1.5/100,000 and 20/100,000 population in the developed and developing countries, respectively.[7] At least 890,000 cases [500,000 in Africa; 210,000 in pacific countries; 100,000 in Europe and 80,000 in America] are estimated to occur annually. Of these cases, 160,000 and 135,000 of them are disabling and fatal, respectively.[7],[8] Meningitis is considered as an endemic disease in Egypt.[7] Apart from endemicity, a violent periodic epidemic occurs every 6-12 years in the African meningitis belt; usually at the end of the dry season, peaking in April and declining with the start of the rainy season in May or June.[8],[9] The largest epidemic occurred during the 1996 season with over 140,000 cases and 16,000 fatalities. It extended during the 1997 season afflicting over 60,000 cases of whom over 6000 died.[10] Before the era of antibiotics, meningitis had a case fatality rate of almost 100%.[11] Today, despite the availability of nontoxic and affordable antibiotics worldwide, the case fatality rate and neurological sequelae among survivors are still high, ranging from 15 to 70% and 10 to 35%, respectively; depending on the interaction between multiple factors related to the causative agent, host, and management.[12]-[14] In Egypt, case-fatality rate ranged from 8.5 to 55%.[15]

The purpose of this prospective research was to gain further understanding of the epidemiological characteristics and clinical indices that may predict the prognostic profile of meningitis among cases admitted at Alexandria fever hospital.

Materials and Methods

In a prospective study, 310 patients (age ranged from 3 months to 15 years), admitted at Alexandria fever hospital (700-bed ministry of health hospital) during the seasons 2002-2003 with clinical diagnosis of meningitis/meningoencephalitis were recruited for the study. This hospital is the only one in Alexandria caring for patients with suspected meningitis i.e., with clinical symptoms and/or signs of meningitis.

Diagnostic inclusion criteria were clinical symptoms and signs of meningitis (e.g., fever, severe headache, severe irritability, photophobia, vomiting drowsiness, neck stiffness, bulging fontanel, Kernig and Brudziniski signs[16] plus either a positive culture from CSF or a negative culture with a positive CSF antigen study or Gram stain in conjunction with a CSF leukocyte concentration (WB) of > 10/mm³; blood culture positive with CSF WBC of > 100/mm³; or, in the absence of bacterial isolate, CSF WBC of > 4000/mm³.[17] Data collected from cases, after getting an informed consent, included demographic features, clinical presentations, medications taken, and duration of illness. The severity of meningitis in our patients was assessed according to the WHO meningitis prognostic scoring system using seven parmeters [Table - 1].[18]

CSF samples collected on admission were subjected to laboratory analysis including cell count, glucose level, Gram stain using direct microscopy, bacteriological culture and ELISA (Meningitis EIA test, Karo-Bio Diagnostics, Sweden).[19]

Cultures on blood chocolate agar and MacConkey agar were incubated at 37°C aerobically and read after 48 hours. Chocolate agar plates were incubated in candle jars. Isolates were identified following standard procedures. A case presenting with clinical meningitis (fever, headache, neck and back rigidity, and photophobia) whose CSF showed pleocytosis but was sterile was considered as aseptic meningitis.[20] On discharge and one month later, patients were reexamined clinically for assessing their prognosis.

Statistical analysis

Data were analyzed using SPSS, version 11 (Chicago, USA, 2004) and Epi-info, version 6 (WHO, Geneva, 2000) statistical software. Student′s t-test was utilized for comparison of the means of quantitative continuous variables of two groups. For comparison between categorical variables, chi-square test was used. A bad prognosis (death or epilepsy events) was treated as a dependent variable in logistic regression analysis. Age, sex, WHO meningitis score, CSF glucose level, CSF WBCs count, convulsions, history of familial smoking, and working mothers were treated as independent categorical variables. The cut-off level for WHO meningitis score, CSF glucose level, and CSF WBCs count was chosen based on the 90^th percentile among non-confirmed patients. Multiple associations were evaluated in multiple logistic regression model based on the backward stepwise selection. This procedure allowed the estimation of the strength of the association between each independent variable and the dependant variable while taking into account the potential confounding effects of the other independent variables. The covariates were removed from the model if the likelihood ratio statistic based on the maximum likelihood estimates had a probability >0.10. Each category of the predictor variables was contrasted with the initial category (reference category). The adjusted measure of association between risk factors and bad meningitis prognosis was expressed as the odds ratio (OR) with 95% confidence interval (95% CI). An adjusted OR with 95% CI that did not include 1.0 was considered significant.

Survival study

Six months survival study for all cases was done using Kaplan Meier estimates of survival curves. Cases were traced every month to confirm their vital status. The log rank test was used to compare survival rates of acute bacterial meningitis and aseptic meningitis.

Results

The study included 310 patients with confirmed meningitis. Aseptic meningitis represented approximately one third of cases (108=34.8%) and acute bacterial meningitis (ABM) represented the other two thirds (202=65.2%). Sixty five (21%) of the ABM cases were caused by H. influenzae , 43 (13.9%) by S. pneumoniae , 44 (14.2%) by N. meningitidis and 50 (16.1%) by other undetermined bacteria. Both aseptic meningitis and ABM prevailed among urbans (63% vs. 66.3%), males (57.4% vs. 65.8%), cases belonging to families with more than two smokers (71.3% vs. 62.9%), and cases having either blood group with no statistically significant difference in any of the categories (P >0.05). All aseptic meningitis cases occurred below the age of five (3-15 months) while 52.4% of ABM cases were above five years (P <0.0001). A significant association between aseptic meningitis and previous vaccination (81.5%, P <0.0001) and between ABM and each of low socio-economic class (96%, P <0.0001) and working mother (83.2%, P <0.0001) is shown in [Table - 2]. Aseptic meningitis showed no significant difference between expected and observed values for all season while ABM showed a significant difference (X² = 133.3, P <0.001) as shown in [Table - 2]. The presenting features of ABM included high fever (92.1%), vomiting (75.2%), and seizures (64.9%). Meningeal signs, cranial nerve palsies and coma were elicited in 23.6%, 16.8% and 11.9% of cases respectively. For those below one year old, irritability and refusal of feeds were encountered among 92.9% and 78.6% respectively. The mean levels of CSF WBCs, CSF protein and CSF sugar were 121±5/cu mm, 169±142 mg/dL and 19±3 mg/dL respectively as shown in [Table - 3]. There is a significant association between scoring of the severity of meningitis and the causative organism, where the number of cases with high scores ³9 is higher with S. pneumoniae (34.9%), N. meningitidis (40.0%) and other bacteria (38.0%) as compared to other types (H. influenzae and aseptic meningitis) (P <0.05) [Table - 4]. In [Table - 5], CSF WBCs count did not vary significantly by scores of meningitis (P >0.05), while CSF glucose level dropped with high score (greater than or equal to 9) in all ABM categories (P <0.001). Also, a significant negative correlation between CSF glucose level and WHO scoring of meningitis is clearly obvious (R= - 0.4113, P <0.01). The cases of ABM and aseptic meningitis that recovered were 78.2% and 90.8%, respectively. Death was absolutely exclusive to a score of > 9 in both types; the case fatality rate for ABM was 13.9% and for aseptic meningitis was 3.4%. As shown in [Table - 6], 22 (7.1%) cases experienced recurrent epileptic fits after 1 month. Of them, 18 cases had a score of ≥9 (X² = 144.2, P <0.001). Kaplan Meier curves revealed that children diagnosed with aseptic meningitis had better survival rates than those with ABM [Figure - 1]. The 6-months survival rate was 95.4% among those with aseptic meningitis as compared to 86.1% with ABM cases. This difference was statistically significant (log rank test=6.29, P =0.012). In [Table - 7] the logistic regression analysis revealed that the risk of bad prognosis (death and epilepsy) encountered among our cases increased significantly with meningitis score ³9 (OR=22.7; 95% CI 18.3-69.2), CSF glucose level < 10 mg/dL (OR=13.3; 95% CI 9.01-51.1), occurrence of generalized convulsions (OR=8.05; 95% CI 7.11-36.6), infancy (≤1 year of age) (OR=7.7; 95% CI 8.81-20.03), presence of more than one family smoker (OR=3.01; 95% CI 1.07-6.9) and working mother (OR=2.9; 95% CI 1.17-26.6). Gender and CSF WBCs level were not significantly related to prognosis of meningitis.

Discussion

Acute bacterial meningitis appeared in the current study as an important type of meningitis, it accounted for approximately two thirds of cases. Similar finding has been reported by others.[21] As worldwide,[1]-[5],[15],[22] in our study, the main three causative bacteria were H. influenzae, S. pneumoniae and N. meningitidis . Nonetheless a considerable percentage was attributed to other undetermined bacteria which may be Klebsiella pneumoniae , E. coli and enterobacteria and others as shown by Gebremaiiam in Addis Ababa.[23] However, in India, a low incidence of infection with N. meningitidis has been reported.[24]-[25] Unlike meningococcal infection in other areas,[6],[14] ABM in our cases was found to peak in winter and fall. This may augment and explain the significant connection of ABM to increased number of smokers, in families having high crowding index and belonging to lower socioeconomic status, with working mothers (sellers, housemaids and cleaners) and dwelling urban slums. These circumstances increase the likelihood of carriage and transmission of the causative pathogens among studied cases. While blood group and previous vaccination had no influence on the occurrence of ABM, gender and age identified double predilection of studied ABM to male (2:1 male to female ratio) in the age group 1-7 years. Of note is that our study clarified that ABM was not the lone player in the playground of infantile and childhood meningitis. Aseptic meningitis caused by different viral infection was another strong contender accounting for one third of all cases. Unlike ABM, all aseptic meningitis cases occurred in a younger age group (3-15 months old). But similar to ABM, most vulnerable of aseptic meningitis were males (1.4:1 male to female ratio), those found in urban areas, those having more than two family members who smoke, and those belonging to a lower socio-economic class. However such similar demographic features between ABM and aseptic meningitis can be ascribed to conduction of the study in a public hospital. The strong association between aseptic meningitis and history of previous vaccination (81.5%) among our cases provides an additional support to the findings of Khalfan et al in Bahrain.[20] They reported occurrence of two epidemics of aseptic meningitis in 1995 (n=286) and 1999 (n=169) due to echoviruses 30 and 4 (with male to female ratios of 1.8:1 and 2.1:1 respectively) following national immunization days (NIDs) for eradication of poliomyelitis. Similar 1.7:1 and 2.1:1 male to female ratios were reported by Make et al[26] and Hsu et al[27] in two outbreaks which occurred in Kaohsiung (1988) and in Taipei (1993) in Taiwan, respectively. Their explanation to such epidemics was that feeding trivalent oral polio vaccine (OPV) to children within a short period predisposes them to dramatic intestinal ecologic changes in the form of suppression of the naturally occurring echoviruses with dominance and excretion of the Sabine polio virus in the stool for around two months. This is followed by colonization and invasiveness of the potentially epidemic strains. Conduction of four rounds of NIDs in Egypt all over the year may elucidate lack of seasonal affinity of aseptic meningitis among our cases.

From clinical point of view, ABM appeared more prone to an aggressive course and bad outcome. It had a case fatality rate of 13.9% close to that (13%) reported by Shembesh et al ,[28] yet much lower than that detected (24%) by Youssef et al .[15] In agreement with another study,[15] the case fatality was highest with unknown bacteria, modest with N. meningitidis and lowest with S. pneumoniae and H. influenzae . This necessitates the importance of determination of such unknown organisms to avoid empirical management. Most of the deaths in our study occurred within the first few days of hospitalization reflecting the critical condition of the patients at admission. In developed countries, inspite of availability of all facilities, the case fatality rate of meningitis in early childhood approaches 10%.[21]

Recurrent epileptic fits were the only complication encountered in 7.1% of participants. This figure is much lower than that documented by Gebremariam[23] who found that 21% of survivors suffered hydrocephalus, spastic paresis and seizures.

It is concluded that this study highlights the importance of some predictors of the outcome of meningitis in children. Also it is concluded that quick and simple scoring scales, as WHO scale, are not only applicable but valuable prognostic tools for meningitis in children.

Acknowledgement

The authors would sincerely like to thank Mr. Harold Trupos and Mr. Richards Timothy for revising of the English text.

References

1.	Sarair HA, Aman H, All-Madan M, Al-Khater M. Bacterial meningitis in Saudi children. Indian J Pediatric 1992;59 :719-21.  Back to cited text no. 1
2.	Liu CC, Chen JS, Lin Chattel, et al . Bacterial meningitis in infants and children in southern Taiwan. J Formosan Med Assoc 1993;92 :884-8.  Back to cited text no. 2
3.	Commey JO, Rodrigues OP, Akita FA, Newman M. Bacterial meningitis in children in Southern Ghana. East Afr Med J 1994;71 :113-7.  Back to cited text no. 3  [PUBMED]
4.	Dagan R, Isaachoson M, Lang R, et al . Epidemiology of pediatrics meningitis caused by Haemophilus influenzae type b, Streptococcus pneumoniae and Neisseria meningitides in Israel. J Infect Dis 1994;169 :912-6.  Back to cited text no. 4
5.	Nabi G, Khan TA, Raj AB, et al . Bacterial meningitis in children. Saudi Med J 1992;13 :348-51.  Back to cited text no. 5
6.	Muhe L, Tilahun M, Lulseged S, et al . Clinical signs and etiologic agents of pneumoni clinical outcome predictors. Ethiop Med J 1994;32 :206.  Back to cited text no. 6
7.	Anonymous. Control of epidemic meningococcal diseases. WHO practical guidelines. World Health Organization and Foundation Marcel Merieux, Edition Fondation Marcel Merieux, Lyon 1995.  Back to cited text no. 7
8.	Tikhomirov E, Santa MM, Eastern K. Meningococcal diseases: Public health burden and control. Wld hlth Statist Quart 1997;50 :170-7.  Back to cited text no. 8
9.	Salih MA, Sid AH, Kamar ZA, et al . Features of large epidemic of group A meningococcal meningitis in Khartoum, Sudan in 1988. J Infect Dis 1990;22 :161-70.  Back to cited text no. 9
10.	Greenwood BM, Blakebrough IS, Bradley Ak, et al . Meningococcal diseases and season in sub- Sahran Africa. Lancet 1984;1 :1393-42.  Back to cited text no. 10
11.	Pecoul B, Varaine F, Kieta M, et al . Long acting chloramphinicol versus intravenous ampicilline for treatment of bacterial meningitis. Lancet 1991;338 :862-6.  Back to cited text no. 11
12.	Carrol KJ, Carrol C. A prospective investigation for the long -term auditory- neurological sequelae associated with bacterial meningitis: A study from Vanuatu. J Top Med Hyg 1994;97 ;145-50.   Back to cited text no. 12
13.	Ciana G, Parmar N, Antonio C, et al . Effectiveness of adjunctive treatment with steroids in reducing short term mortality in a high risk population of children with bacterial meningitis. J Trop Med Hyg 1995;41 :164-8.  Back to cited text no. 13
14.	Moore PS. Meningococcal meningitis in sub-Sahran Africa: A model for the epidemic process. Clin infect diseases 1992;11 :515-25.  Back to cited text no. 14
15.	Youssef FG, El-Sakka H, Azab A, Eloun S, Chapman GD, et al . Etiology, antimicrobial susceptibility profiles, and mortality associated with bacterial meningitis among children in Egypt. Ann Epidemiol 2004;14 :44-8.  Back to cited text no. 15
16.	Wolf SM. Laboratory evaluation of the child with a febrile convulsion. Pediatrics 1978;62 :1074-6.  Back to cited text no. 16  [PUBMED]
17.	Berkley JA, Mwangi I, Ngetsd CJ, et al . Diagnosis of acute bacterial meningitis at a distant hospital in sub-saharan Africa. Lancet 2001;357 :1753-7.   Back to cited text no. 17
18.	Ajayi-Obe EK, Lodi E, Alkali AS, et al . Prognostic scores for use in African meningococcal epidemics. WHO Bulletin OMS 1998;76 :149-52.  Back to cited text no. 18
19.	Salih MA, Ahmad AA, Ahmad HS, Olcen P. An ELISA assay for the rapid diagnosis of acute bacterial meningitis. Ann Trop Paediatr 1995;15 :273-8.  Back to cited text no. 19
20.	Khalfan S, Ayhmard M, Lina B, et al . Epidemics of aseptic meningitis due to enteroviruses following national immunization days in Bahrain. Ann Trop Paediatr 1998;18 :101-9.  Back to cited text no. 20
21.	Quagliarello V, Scheld WM. Bacterial meningitis: Pathogenesis, pathophysiology and progress. N Engl J Med 1992;327 :864-72.  Back to cited text no. 21  [PUBMED]
22.	Goetghebeur T, Eoin WT, Wermenbol V, et al . Outcome of meningitis caused by streptococcus pneumonia and Haemophilus influenzae type b in children in the Gambia. Trop Med Int Hlth 2000;5 :207-14.  Back to cited text no. 22
23.	Gebremariam A. Neonatal meningitis in Addis Ababa: A 10-year reviews. Ann Trop Paediatr 1998;18 :279-83.  Back to cited text no. 23  [PUBMED]
24.	Chinchankar N, Mane M, Bhave S, Bapat S, Bavdekar A, Pandit A, et al . Diagnosis and outcome of acute bacterial meningitis in early childhood. Indian Pediatr 2002;39 :914-21.  Back to cited text no. 24
25.	Grimwood K, Anderson P, Anderson V, Tan L, Nolan T. Twelve years outcomes following bacterial meningitis: Further evidence for persisting effects. Arch Dis Child 2000;83 :111-6.  Back to cited text no. 25  [PUBMED]  [FULLTEXT]
26.	Mak SC, Jeng JE, Jongy J, et al . Clinical observation and biological study of aseptic meningitis in the kohsiung area. J Formosan Med Assoc 1990;89 ;868-72.  Back to cited text no. 26
27.	Hsu CM, Chen JM, Hunang LM, et al . Outbreak of aseptic meningitis in Taipei in spring 1993. J Formosan Med Assoc 1995;94 :14-8.  Back to cited text no. 27
28.	Shembesh NM, EL Bargathy SM, Rao BN, Kashbur IM. A Prospective study of bacterial meningitis in children from North - Eastern Libya. Ann Trop Pediatr 1998;18 :203-7.  Back to cited text no. 28  [PUBMED]

Copyright 2005 - Indian Journal of Medical Microbiology

The following images related to this document are available:

Photo images