Original Article
Risk Factors for Acute Symptomatic Seizure in Bacterial Meningitis in Children
Journal of Child Neurology 1-4 ª The Author(s) 2014 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0883073814555907 jcn.sagepub.com
Ana Rosa Melo Correˆa-Lima, MD, MSc1, Demo´crito de Barros Miranda-Filho, MD, PhD2, Marcelo Moraes Valenc¸a, MD, PhD3, and Luciana Andrade-Valenc¸a, MD, PhD1,3
Abstract The aim of this study was to investigate the association between clinical and laboratory parameters at admission and the occurrence of in-hospital symptomatic seizures in children with acute bacterial meningitis in Brazil. A retrospective casecontrol study in 270 children with confirmed bacterial meningitis, aged from 1 month to 14 years, was conducted between January 2004 and December 2008. Associations with the occurrence of in-hospital epileptic seizures were adjusted using stepwise multiple logistic regression analysis. Sixty-seven children suffered at least one in-hospital epileptic seizure. After multivariate analysis, the independent predictors considered for in-hospital epileptic seizures were as follows: age less than 2 years (odds ratio ¼ 0.97; 95% confidence interval 0.97-0.98), pneumococcal etiology (odds ratio ¼ 4.55; 95% confidence interval 1.88-11.0); altered mental status (odds ratio ¼ 3.47; 95% confidence interval 1.66-7.26) and cerebrospinal fluid leukocyte count below 1000 cells (odds ratio ¼ 2.14; 95% confidence interval 0.99-4.60). Mortality was higher in patients with intrahospital epileptic seizures compared to those without (25/67 [37.3%] vs 9/203 [4.43%], P < .001]. Keywords epileptic seizures, bacterial meningitis, risk factors, childhood, case-control Received June 11, 2014. Received revised August 08, 2014. Accepted for publication September 11, 2014.
Acute symptomatic seizures, a delay in the diagnosis, severity of the disease at the moment of its recognition, and the type of pathogen can all have an unfavorable impact on the outcomes of children with acute bacterial meningitis.1-5 Seizures can be linked to eventual death, either through the increased risk of bronchial aspiration, or by progressing to intractable status epilepticus.3-5 In addition, there is an increased risk of developing late symptomatic epilepsy in those patients who have presented seizures during the acute infection.5,6 The aim of this study is to investigate the association between clinical and laboratory parameters at admission and the occurrence of in-hospital acute symptomatic seizures in a reference hospital specialized in the treatment of acute bacterial meningitis in northeast Brazil.
The Correia Picanc¸o Hospital is a referral pediatric institution for the treatment of community-acquired meningitis in the state of Pernambuco, Brazil, attends approximately 90% of all cases registered within the community, and routinely administers 4-day dexamethasone combined with an antibiotic therapy. Confirmed acute bacterial meningitis was defined by obtaining a positive cerebrospinal fluid bacterial culture or cerebrospinal fluid cell count greater than 10 leukocytes/mm3, plus at least 1 of the following parameters: a positive blood culture, bacterial antigen detection tests of cerebrospinal fluid, or a direct Gram stain in children with a clinical syndrome consistent with bacterial meningitis. The case group were those who presented at least 1 in-hospital acute symptomatic seizure (after initiating treatment for meningitis),
1
Methods Identification of Cases and Control A retrospective case-control study in children with acute bacterial meningitis, aged from 1 month to 14 years, was conducted at Correia Picanc¸o Hospital between January 2004 and December 2008. In Brazil, Haemophilus influenzae type b (Hib) vaccine and pneumococcal conjugated vaccine were introduced in 1999 and 2010, respectively.
Department of Neurology, University of Pernambuco, Recife, Brazil Department of Internal Medicine, Division of Infectious Diseases, University of Pernambuco, Recife, Brazil 3 Department of Neurosurgery and Neurology, Federal University of Pernambuco, Brazil 2
Corresponding Author: Ana Rosa Melo Correˆa-Lima, MD, MSc, Rua Arno´bio Marques 310, Santo Amaro, CEP 50100130, Recife-PE, Brazil. Email: [email protected]
2 whereas the control group consisted of patients with meningitis who did not present epileptic seizure. The clinical characterization of epileptic seizure was based on an interpretation of the descriptions found in the medical records.
Journal of Child Neurology Table 1. Clinical and Laboratory Features in Children With and Without in-Hospital Seizures Among 270 Cases of Acute Bacterial Meningitis.a Characteristic
Clinical and Laboratorial Parameters The following clinical variables were studied: (1) impaired mental status at the time of admission; (2) skin rash; (3) prior use of antibiotics (at least 48 hours before admission); (4) signs of shock represented by systolic blood pressure below 70 mmHg, cold extremities, weak pulse, or reference to the use of vasoactive drugs in the emergency room; and (5) duration of symptoms (interval in days of initial symptoms of meningitis and admission to hospital). The cerebrospinal fluid parameters analyzed on admission were white cell count and glucose and protein concentrations. Hemoglobin concentration and platelet and total white cell counts were measured in the blood. In this analysis, the continuous variables were dichotomized using cutoff limits employed elsewhere.2
Statistical Analyses All analyses were conducted using Epi-Info, version 6.0. In order to calculate the sample size, 2 variables were considered, based on previous studies: impaired mental status and pneumococcal etiology.6,7 Considered a proportion of 1 case to 2 controls, a ¼ 5% and b ¼ 20%; the minimal sample estimated was 63 cases and 126 control patients. Because this was an exploratory study to identify other possible factors associated with the occurrence of epileptic seizure in pediatric patients, a sample of 289 individuals was taken. The Kolmogorov-Smirnov test was applied to determine the type of variable distributions to be studied. When the variables presented a normal distribution, the parametric Student t test was applied. In the case of variables that did not present normal distribution, the nonparametric Mann-Whitney test was applied. To compare categorical variables, the chi-square test with Yate correction or Fisher exact test, was applied. Data were presented as mean + standard deviation. A significant difference of P < .05 was considered. Associations with the occurrence of in-hospital epileptic seizures were adjusted using stepwise multiple logistic regression analysis. The variables with P < .20 in the univariate analysis were included in the multivariate model. Variables with P < .10 in the multivariate analysis were considered independent risk factors for the occurrence of epileptic seizures. The results were expressed in odds ratio with a confidence interval of 95%.
Results Over a period of 5 years, 289 children met the inclusion criteria. Of these, 19 patients were excluded because of a reported history of exclusive prehospital seizures. The case group was made up of 67 patients and the control group 203. Table 1 illustrates the results of the univariate analysis of the clinical and laboratory parameters observed in patients with acute in-hospital symptomatic seizures compared to those without seizures. The mean age was lower in the group with acute seizure (1.5+2.3 vs 5.8+4.4, P < .001) and 86.6% of the children who suffered seizures were under 2 years of age. Of the 79 children who presented skin rash on admission, only 5
Age, yc Age
Risk Factors for Acute Symptomatic Seizure in Bacterial Meningitis in Children
Journal of Child Neurology 1-4 ª The Author(s) 2014 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0883073814555907 jcn.sagepub.com
Ana Rosa Melo Correˆa-Lima, MD, MSc1, Demo´crito de Barros Miranda-Filho, MD, PhD2, Marcelo Moraes Valenc¸a, MD, PhD3, and Luciana Andrade-Valenc¸a, MD, PhD1,3
Abstract The aim of this study was to investigate the association between clinical and laboratory parameters at admission and the occurrence of in-hospital symptomatic seizures in children with acute bacterial meningitis in Brazil. A retrospective casecontrol study in 270 children with confirmed bacterial meningitis, aged from 1 month to 14 years, was conducted between January 2004 and December 2008. Associations with the occurrence of in-hospital epileptic seizures were adjusted using stepwise multiple logistic regression analysis. Sixty-seven children suffered at least one in-hospital epileptic seizure. After multivariate analysis, the independent predictors considered for in-hospital epileptic seizures were as follows: age less than 2 years (odds ratio ¼ 0.97; 95% confidence interval 0.97-0.98), pneumococcal etiology (odds ratio ¼ 4.55; 95% confidence interval 1.88-11.0); altered mental status (odds ratio ¼ 3.47; 95% confidence interval 1.66-7.26) and cerebrospinal fluid leukocyte count below 1000 cells (odds ratio ¼ 2.14; 95% confidence interval 0.99-4.60). Mortality was higher in patients with intrahospital epileptic seizures compared to those without (25/67 [37.3%] vs 9/203 [4.43%], P < .001]. Keywords epileptic seizures, bacterial meningitis, risk factors, childhood, case-control Received June 11, 2014. Received revised August 08, 2014. Accepted for publication September 11, 2014.
Acute symptomatic seizures, a delay in the diagnosis, severity of the disease at the moment of its recognition, and the type of pathogen can all have an unfavorable impact on the outcomes of children with acute bacterial meningitis.1-5 Seizures can be linked to eventual death, either through the increased risk of bronchial aspiration, or by progressing to intractable status epilepticus.3-5 In addition, there is an increased risk of developing late symptomatic epilepsy in those patients who have presented seizures during the acute infection.5,6 The aim of this study is to investigate the association between clinical and laboratory parameters at admission and the occurrence of in-hospital acute symptomatic seizures in a reference hospital specialized in the treatment of acute bacterial meningitis in northeast Brazil.
The Correia Picanc¸o Hospital is a referral pediatric institution for the treatment of community-acquired meningitis in the state of Pernambuco, Brazil, attends approximately 90% of all cases registered within the community, and routinely administers 4-day dexamethasone combined with an antibiotic therapy. Confirmed acute bacterial meningitis was defined by obtaining a positive cerebrospinal fluid bacterial culture or cerebrospinal fluid cell count greater than 10 leukocytes/mm3, plus at least 1 of the following parameters: a positive blood culture, bacterial antigen detection tests of cerebrospinal fluid, or a direct Gram stain in children with a clinical syndrome consistent with bacterial meningitis. The case group were those who presented at least 1 in-hospital acute symptomatic seizure (after initiating treatment for meningitis),
1
Methods Identification of Cases and Control A retrospective case-control study in children with acute bacterial meningitis, aged from 1 month to 14 years, was conducted at Correia Picanc¸o Hospital between January 2004 and December 2008. In Brazil, Haemophilus influenzae type b (Hib) vaccine and pneumococcal conjugated vaccine were introduced in 1999 and 2010, respectively.
Department of Neurology, University of Pernambuco, Recife, Brazil Department of Internal Medicine, Division of Infectious Diseases, University of Pernambuco, Recife, Brazil 3 Department of Neurosurgery and Neurology, Federal University of Pernambuco, Brazil 2
Corresponding Author: Ana Rosa Melo Correˆa-Lima, MD, MSc, Rua Arno´bio Marques 310, Santo Amaro, CEP 50100130, Recife-PE, Brazil. Email: [email protected]
2 whereas the control group consisted of patients with meningitis who did not present epileptic seizure. The clinical characterization of epileptic seizure was based on an interpretation of the descriptions found in the medical records.
Journal of Child Neurology Table 1. Clinical and Laboratory Features in Children With and Without in-Hospital Seizures Among 270 Cases of Acute Bacterial Meningitis.a Characteristic
Clinical and Laboratorial Parameters The following clinical variables were studied: (1) impaired mental status at the time of admission; (2) skin rash; (3) prior use of antibiotics (at least 48 hours before admission); (4) signs of shock represented by systolic blood pressure below 70 mmHg, cold extremities, weak pulse, or reference to the use of vasoactive drugs in the emergency room; and (5) duration of symptoms (interval in days of initial symptoms of meningitis and admission to hospital). The cerebrospinal fluid parameters analyzed on admission were white cell count and glucose and protein concentrations. Hemoglobin concentration and platelet and total white cell counts were measured in the blood. In this analysis, the continuous variables were dichotomized using cutoff limits employed elsewhere.2
Statistical Analyses All analyses were conducted using Epi-Info, version 6.0. In order to calculate the sample size, 2 variables were considered, based on previous studies: impaired mental status and pneumococcal etiology.6,7 Considered a proportion of 1 case to 2 controls, a ¼ 5% and b ¼ 20%; the minimal sample estimated was 63 cases and 126 control patients. Because this was an exploratory study to identify other possible factors associated with the occurrence of epileptic seizure in pediatric patients, a sample of 289 individuals was taken. The Kolmogorov-Smirnov test was applied to determine the type of variable distributions to be studied. When the variables presented a normal distribution, the parametric Student t test was applied. In the case of variables that did not present normal distribution, the nonparametric Mann-Whitney test was applied. To compare categorical variables, the chi-square test with Yate correction or Fisher exact test, was applied. Data were presented as mean + standard deviation. A significant difference of P < .05 was considered. Associations with the occurrence of in-hospital epileptic seizures were adjusted using stepwise multiple logistic regression analysis. The variables with P < .20 in the univariate analysis were included in the multivariate model. Variables with P < .10 in the multivariate analysis were considered independent risk factors for the occurrence of epileptic seizures. The results were expressed in odds ratio with a confidence interval of 95%.
Results Over a period of 5 years, 289 children met the inclusion criteria. Of these, 19 patients were excluded because of a reported history of exclusive prehospital seizures. The case group was made up of 67 patients and the control group 203. Table 1 illustrates the results of the univariate analysis of the clinical and laboratory parameters observed in patients with acute in-hospital symptomatic seizures compared to those without seizures. The mean age was lower in the group with acute seizure (1.5+2.3 vs 5.8+4.4, P < .001) and 86.6% of the children who suffered seizures were under 2 years of age. Of the 79 children who presented skin rash on admission, only 5
Age, yc Age
