Annals of Tropical Paediatrics International Child Health
ISSN: 0272-4936 (Print) 1465-3281 (Online) Journal homepage: http://www.tandfonline.com/loi/ypch19
Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria George O. Akpede, Roger M. Sykes & Philip O. Abiodun To cite this article: George O. Akpede, Roger M. Sykes & Philip O. Abiodun (1992) Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria, Annals of Tropical Paediatrics, 12:4, 385-389, DOI: 10.1080/02724936.1992.11747603 To link to this article: http://dx.doi.org/10.1080/02724936.1992.11747603
Published online: 13 Jul 2016.
Submit your article to this journal
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ypch19 Download by: [Cornell University Library]
Date: 29 June 2017, At: 01:59
Annals of Tropical Paediatrics (1992) 12, 385-389
Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria GEORGE O. AKPEDE, ROGER M. SYKES & PHILIP O. ABIODUN Department of Child Health, College of Medical Sciences, University of Benin, Nigeria (Received 25 September 1991)
Summary A total of 522 children, aged 1 month to 6 years, who presented with convulsions and fever of acute onset at the Children's Emergency Room of the University of Benin Teaching Hospital over a 1-year period, were prospectively evaluated. Bacterial meningitis was diagnosed in 22 (4.2%) on bacteriological and/or biochemical evidence. The causative organisms were cultured from the CSF in 13 (Neisseria meningitidis=7, Streptococcus pneumoniae=5 and Haemophi/us infiuenzae= 1) and identified by Gram stain only in three (Gram-positive diplococci = 2 and Gram-negative diplococci = 1). No organisms were identified in the CSF of six of the children with meningitis. The prevalence of meningitis declined sharply after 6 months of age. Six of the children with bacterial meningitis lacked classical meningeal signs but had other indications for lumbar puncture. The following were significantly associated with meningitis: age under 6 months; focal or multiple seizures; absence of a past or family history of seizures; unrousable coma; and an extracranial focus of infection. 1t is concluded that bacterial meningitis occurs in a good proportion of children, even beyond infancy, with convulsions associated with fever of acute onset, and that decision on the need for lumbar puncture should be guided by clinical features such as age and the presence of complex febrile seizures.
Introduction Fever in a child in the tropics is alarming to parents because it is often the herald of convulsions, other serious illness, or death. 1 Convulsions associated with fever are a cornmon emergency in the tropics 1-4 and, although commonly due to malaria, 2•3•5 can also be due to viral infections and bacterial meningitus. 4 •6 •7 Febrile convulsions occur predominantly in children under 3 years of agé•8 and are a common cause of morbidity in the tropics. 2' 6 Convulsions occur in Reprint requests to: Dr George O. Akpede, Dept. of Paediatrics, College of Medical Sciences, University of Maiduguri, PMB 1069, Maiduguri, Borna State, Nigeria.
25-30% of children with bacterial meningitis which in 30-40% of cases, especially those under 2 years of age, may not show meningeal signs.9- 11 There is the risk therefore of mistaking meningitis for febrile convulsions. Because of the serious implications of missing a diagnosis of meningitis, lumbar puncture (LP) has been advocated for ali children presenting with febrile convulsions,8 •12 •13 especially those under 2 years or presenting with their first seizure. 11 •14 The procedure of LP is not without risks such as the initiation of meningitis in a bacteraemic child 15•16 and the development of cerebellar coning. 17 Lumbar puncture might also provide false reassurances, with delay in diagnosing meningitis when the initial
386
G. 0. Akpede et al.
cerebrospinal fluid (CSF) is normal. 16•17 Sorne authors therefore question the wisdom of non-selective LPs in children with febrile convulsions, 17' 18 but this attitude bas been sharply criticised by others 19-21 and criteria have been formulated to aid in deciding on the need for LP in children with convulsions associated with fever . 11 ' 13 ' 14' 22 Cerebral and other forms of severe malaria are common in the tropics 5•23 and can influence decisions regarding the need for LPs. The high incidence of malaria and widespread shortage of manpower and facilities in developing countries might argue for a policy of selective LPs in children presenting with convulsions associated with fever. The present study was undertaken to ascertain the risk factors associated with a diagnosis of bacterial meningitis and to investigate the proportion of cases which would be missed if lumbar puncture were performed only when clinical signs are present.
Patients and methods Between 24 October 1988 and 23 October 1989, 653 children under 16 years of age presented with convulsions associated with fever at the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria. Of these, 522 were recruited for the study based on the criteria of age 1 month-6 years, rectal temperature of at least 38°C and fever of not more than 7 days duration. Ali the children were evaluated by a detailed history and physical examination. Unrousable coma was defined as non-localizing or absent motor response to noxious stimuli. 23 Ali the children bad an LP done irrespective of the presence or absence of features of meningeal irritation. CSF was analysed for glucose and protein and examined for total and differentiai white blood cell counts and Gram stain appearance of any organisms. Samples for culture were collected into sterile botties and innoculated onto blood, chocolate and MacConkey agar plates and incubated at 37°C for 48 h under
both aerobic and anaerobie conditions. Isolates were identified by standard techniques. The diagnosis of meningitis was based on the presence of CSF pleocytosis ( > 5 WBC/mm 3). The diagnosis of bacterial meningitis was based on the presence of a bacterial pathogen identified by Gram stain and/or culture of the CSF. 9 A presumed diagnosis of bacterial meningitis was made in children with no bacterial pathogen identified in the CSF but with pleocytosis (mainly polymorphs) and typical biochemical changes in the CSF. The latter included CSF protein ?: 80 mg/dl and glucose ~ 40~ 0 of the simultaneous blood glucose. 24 The risk factors were assessed using the x2 test. The relative risk of a child developing meningitis if the risk factor is present was calculated together with 95% confidence limits.
Results Twenty-two of the 522 children studied (4.2%) bad bacterial meningitis. Bacteria were cultured from the CSF in 13 (Neisseria meningitidis 7, Streptococcus pneumonia 5 and Haemophilus influenzae 1) and identified by Gram stain only in three (Gram-positive diplococci 2 and Gram-negative diplococci 1). No bacteria were identified either by Gram stain or culture of CSF in six children. The clinical features of the children with bacterial meningitis are shown in Table 1. The prevalence of meningitis declined sharply after the age of 6 months. The difference in the age distributions of the children with meningitis and those with CAF only is highly significant (p ~ 0.0001). Seven (31.8%) of the children with bacterial meningitis bad associated extra-cranial infections, mostly localized to the respiratory tract. Classical clinical signs of meningitis (neck stiffness, Kernig's sign, bulging fontanelle)9 were definitely present in 13, equivocal in three and absent in six of the children with meningitis. Three of the children with no clinical signs of meningitis were aged over 2 years.
Lumbar puncture in convulsions withfever
387
TABLE 1. Clinical features of children presenting with convulsions and fever (CAF) who bad bacterial meningitis Age (yrs) No.withCAF No. (%) with meningitis No. with meningitis with: Family history of convulsion Past history of convulsion Multiple seizures Focal seizures Extracranial focus of infection Convulsion > 15 min duration Unrousable coma Signs of meningeal irritation
1-6months 23 8 (34.8) 0 0 5 4 3 0 7
Ali the factors, with the exception of the duration of the seizure, were significant, with relative risk greater than one (Table Il). The relative risk for a diagnosis of meningitis was highest for those aged under 6 months, intermediate for those with a past or family history of convulsion, unrousable coma and focal seizures, and lowest for those with multiple seizures and an extracranial focus of infection, in the approximate order of magnitude of 5.2:1.8-2.8:1. Six out of22 (27.2%) cases would have been missed if lumbar puncture had been performed only when there were clinical signs of meningitis. The numbers were too small to investigate these further.
Discussion The 4.2% prevalence of meningitis in children with convulsions associated with fever (CAF) found in this study is comparable with the prevalence (4%) reported in a similar population of children in U ganda. 4 Classical signs ofmeningitis were absent in six (27%) of the children with meningitis in the present study, three of whom were over 2 years of age, and ali of them had other indications for LP such as focal seizures 13 and/or unrousable coma. 23 In contrast with other reports/• 10 signs of meningitis were elicited in ali but one child under 6 months of age.
> 6 months-2 yrs 191
>2-6yrs 308
Total 522
5 (2.6)
9 (2.9)
22 (4.2)
1 4 2 1 2 2 3
4 3 5 3 3 3 3 6
5 4 14 9 7 6 5 16
No child with uncomplicated features of febrile convulsions 13 had meningitis in this study. Jaffe et al. in Israel also noted that bacterial meningitis was rare in children presenting with simple febrile convulsions. 13 Age under 6 months, multiple seizures, focal seizures, unrousable coma, presence of an extracranial focus of infection and an absence of a past or family history of convulsion showed associations with bacterial meningitis. The presence of one or more of these features in a child with CAF, even in the absence of classical signs of meningitis, would thus be good indicators of the need for lumbar puncture. Findings in this study support previous recommendations 17' 18 that careful clinical appraisal serves to discriminate between the presence or absence of meningitis in children with CAF. Thus, notwithstanding the disastrous consequences of missing a diagnosis of meningitis, we cannot support those authors who advocate an obligatory diagnostic LP in ali children presenting with CAF under the age of 2 years. 8 •11- 14' 19-21 One third of the children over 2 years of age with meningitis in the present study lacked the classical signs ofmeningeal irritation. We do not therefore advise reliance only on clinical signs of meningitis as the indicator of the need for LP in this age group, but would stress the presence
No Yes
No Yes
Yes No
Yes No
Yes No
Yes No
Yes No
Family history of convulsion
Past history of convulsion
Multiple seizures
Focal seizures
Extracranial focus of infection
Seizure > 15 min
Unrousable coma
5 (14.7) 17 (3.5)
34 (6.5) 488
7 (8.6) 15 (3.4)
81 (15.5) 441 6 (3.3) 16 (4.7)
9 (18.4) 13 (2. 7)
49 (9.4) 473
180 (34.5) 342
14 (6.3) 8 (2.7)
18 (8.6) 4 (1.3)
18 (6.9) 4 (1.5)
221 (42.3) 301
208 (39.9) 314
259 (49.6) 236
8 (34.8) 14 (2.8)
23 (4.4) 499
5 (2.3) 1 (0.3) 3 (6.1) 3 (0.6) 3 (3.7) 3 (0.7)
9 (4.1) 7 (2.3) 6 (12.2) 10 (2.1) 4 (4.9) 12 (2.7)
3 (8.8) 3 (0.6)
4(1.9) 2 (0.6)
14 (6.7) 4 (1.3)
2 (5.9) 14 (2.9)
6 (2.3) 0 (0.0)
12 (4.6) 4 (1.5)
3 (1.7) 3 (0.9)
1 (4.3) 5 (1.0)
7 (30.4) 9 (1.8)
3 (1.7) 13 (3.8)
Meningitis with no meningeal signs (n=6)
Meningitis with meningeal signs (n = 16)
9.9 (
ISSN: 0272-4936 (Print) 1465-3281 (Online) Journal homepage: http://www.tandfonline.com/loi/ypch19
Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria George O. Akpede, Roger M. Sykes & Philip O. Abiodun To cite this article: George O. Akpede, Roger M. Sykes & Philip O. Abiodun (1992) Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria, Annals of Tropical Paediatrics, 12:4, 385-389, DOI: 10.1080/02724936.1992.11747603 To link to this article: http://dx.doi.org/10.1080/02724936.1992.11747603
Published online: 13 Jul 2016.
Submit your article to this journal
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ypch19 Download by: [Cornell University Library]
Date: 29 June 2017, At: 01:59
Annals of Tropical Paediatrics (1992) 12, 385-389
Indications for lumbar puncture in children presenting with convulsions and fever of acute onset: experience in the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria GEORGE O. AKPEDE, ROGER M. SYKES & PHILIP O. ABIODUN Department of Child Health, College of Medical Sciences, University of Benin, Nigeria (Received 25 September 1991)
Summary A total of 522 children, aged 1 month to 6 years, who presented with convulsions and fever of acute onset at the Children's Emergency Room of the University of Benin Teaching Hospital over a 1-year period, were prospectively evaluated. Bacterial meningitis was diagnosed in 22 (4.2%) on bacteriological and/or biochemical evidence. The causative organisms were cultured from the CSF in 13 (Neisseria meningitidis=7, Streptococcus pneumoniae=5 and Haemophi/us infiuenzae= 1) and identified by Gram stain only in three (Gram-positive diplococci = 2 and Gram-negative diplococci = 1). No organisms were identified in the CSF of six of the children with meningitis. The prevalence of meningitis declined sharply after 6 months of age. Six of the children with bacterial meningitis lacked classical meningeal signs but had other indications for lumbar puncture. The following were significantly associated with meningitis: age under 6 months; focal or multiple seizures; absence of a past or family history of seizures; unrousable coma; and an extracranial focus of infection. 1t is concluded that bacterial meningitis occurs in a good proportion of children, even beyond infancy, with convulsions associated with fever of acute onset, and that decision on the need for lumbar puncture should be guided by clinical features such as age and the presence of complex febrile seizures.
Introduction Fever in a child in the tropics is alarming to parents because it is often the herald of convulsions, other serious illness, or death. 1 Convulsions associated with fever are a cornmon emergency in the tropics 1-4 and, although commonly due to malaria, 2•3•5 can also be due to viral infections and bacterial meningitus. 4 •6 •7 Febrile convulsions occur predominantly in children under 3 years of agé•8 and are a common cause of morbidity in the tropics. 2' 6 Convulsions occur in Reprint requests to: Dr George O. Akpede, Dept. of Paediatrics, College of Medical Sciences, University of Maiduguri, PMB 1069, Maiduguri, Borna State, Nigeria.
25-30% of children with bacterial meningitis which in 30-40% of cases, especially those under 2 years of age, may not show meningeal signs.9- 11 There is the risk therefore of mistaking meningitis for febrile convulsions. Because of the serious implications of missing a diagnosis of meningitis, lumbar puncture (LP) has been advocated for ali children presenting with febrile convulsions,8 •12 •13 especially those under 2 years or presenting with their first seizure. 11 •14 The procedure of LP is not without risks such as the initiation of meningitis in a bacteraemic child 15•16 and the development of cerebellar coning. 17 Lumbar puncture might also provide false reassurances, with delay in diagnosing meningitis when the initial
386
G. 0. Akpede et al.
cerebrospinal fluid (CSF) is normal. 16•17 Sorne authors therefore question the wisdom of non-selective LPs in children with febrile convulsions, 17' 18 but this attitude bas been sharply criticised by others 19-21 and criteria have been formulated to aid in deciding on the need for LP in children with convulsions associated with fever . 11 ' 13 ' 14' 22 Cerebral and other forms of severe malaria are common in the tropics 5•23 and can influence decisions regarding the need for LPs. The high incidence of malaria and widespread shortage of manpower and facilities in developing countries might argue for a policy of selective LPs in children presenting with convulsions associated with fever. The present study was undertaken to ascertain the risk factors associated with a diagnosis of bacterial meningitis and to investigate the proportion of cases which would be missed if lumbar puncture were performed only when clinical signs are present.
Patients and methods Between 24 October 1988 and 23 October 1989, 653 children under 16 years of age presented with convulsions associated with fever at the Children's Emergency Room of the University of Benin Teaching Hospital, Nigeria. Of these, 522 were recruited for the study based on the criteria of age 1 month-6 years, rectal temperature of at least 38°C and fever of not more than 7 days duration. Ali the children were evaluated by a detailed history and physical examination. Unrousable coma was defined as non-localizing or absent motor response to noxious stimuli. 23 Ali the children bad an LP done irrespective of the presence or absence of features of meningeal irritation. CSF was analysed for glucose and protein and examined for total and differentiai white blood cell counts and Gram stain appearance of any organisms. Samples for culture were collected into sterile botties and innoculated onto blood, chocolate and MacConkey agar plates and incubated at 37°C for 48 h under
both aerobic and anaerobie conditions. Isolates were identified by standard techniques. The diagnosis of meningitis was based on the presence of CSF pleocytosis ( > 5 WBC/mm 3). The diagnosis of bacterial meningitis was based on the presence of a bacterial pathogen identified by Gram stain and/or culture of the CSF. 9 A presumed diagnosis of bacterial meningitis was made in children with no bacterial pathogen identified in the CSF but with pleocytosis (mainly polymorphs) and typical biochemical changes in the CSF. The latter included CSF protein ?: 80 mg/dl and glucose ~ 40~ 0 of the simultaneous blood glucose. 24 The risk factors were assessed using the x2 test. The relative risk of a child developing meningitis if the risk factor is present was calculated together with 95% confidence limits.
Results Twenty-two of the 522 children studied (4.2%) bad bacterial meningitis. Bacteria were cultured from the CSF in 13 (Neisseria meningitidis 7, Streptococcus pneumonia 5 and Haemophilus influenzae 1) and identified by Gram stain only in three (Gram-positive diplococci 2 and Gram-negative diplococci 1). No bacteria were identified either by Gram stain or culture of CSF in six children. The clinical features of the children with bacterial meningitis are shown in Table 1. The prevalence of meningitis declined sharply after the age of 6 months. The difference in the age distributions of the children with meningitis and those with CAF only is highly significant (p ~ 0.0001). Seven (31.8%) of the children with bacterial meningitis bad associated extra-cranial infections, mostly localized to the respiratory tract. Classical clinical signs of meningitis (neck stiffness, Kernig's sign, bulging fontanelle)9 were definitely present in 13, equivocal in three and absent in six of the children with meningitis. Three of the children with no clinical signs of meningitis were aged over 2 years.
Lumbar puncture in convulsions withfever
387
TABLE 1. Clinical features of children presenting with convulsions and fever (CAF) who bad bacterial meningitis Age (yrs) No.withCAF No. (%) with meningitis No. with meningitis with: Family history of convulsion Past history of convulsion Multiple seizures Focal seizures Extracranial focus of infection Convulsion > 15 min duration Unrousable coma Signs of meningeal irritation
1-6months 23 8 (34.8) 0 0 5 4 3 0 7
Ali the factors, with the exception of the duration of the seizure, were significant, with relative risk greater than one (Table Il). The relative risk for a diagnosis of meningitis was highest for those aged under 6 months, intermediate for those with a past or family history of convulsion, unrousable coma and focal seizures, and lowest for those with multiple seizures and an extracranial focus of infection, in the approximate order of magnitude of 5.2:1.8-2.8:1. Six out of22 (27.2%) cases would have been missed if lumbar puncture had been performed only when there were clinical signs of meningitis. The numbers were too small to investigate these further.
Discussion The 4.2% prevalence of meningitis in children with convulsions associated with fever (CAF) found in this study is comparable with the prevalence (4%) reported in a similar population of children in U ganda. 4 Classical signs ofmeningitis were absent in six (27%) of the children with meningitis in the present study, three of whom were over 2 years of age, and ali of them had other indications for LP such as focal seizures 13 and/or unrousable coma. 23 In contrast with other reports/• 10 signs of meningitis were elicited in ali but one child under 6 months of age.
> 6 months-2 yrs 191
>2-6yrs 308
Total 522
5 (2.6)
9 (2.9)
22 (4.2)
1 4 2 1 2 2 3
4 3 5 3 3 3 3 6
5 4 14 9 7 6 5 16
No child with uncomplicated features of febrile convulsions 13 had meningitis in this study. Jaffe et al. in Israel also noted that bacterial meningitis was rare in children presenting with simple febrile convulsions. 13 Age under 6 months, multiple seizures, focal seizures, unrousable coma, presence of an extracranial focus of infection and an absence of a past or family history of convulsion showed associations with bacterial meningitis. The presence of one or more of these features in a child with CAF, even in the absence of classical signs of meningitis, would thus be good indicators of the need for lumbar puncture. Findings in this study support previous recommendations 17' 18 that careful clinical appraisal serves to discriminate between the presence or absence of meningitis in children with CAF. Thus, notwithstanding the disastrous consequences of missing a diagnosis of meningitis, we cannot support those authors who advocate an obligatory diagnostic LP in ali children presenting with CAF under the age of 2 years. 8 •11- 14' 19-21 One third of the children over 2 years of age with meningitis in the present study lacked the classical signs ofmeningeal irritation. We do not therefore advise reliance only on clinical signs of meningitis as the indicator of the need for LP in this age group, but would stress the presence
No Yes
No Yes
Yes No
Yes No
Yes No
Yes No
Yes No
Family history of convulsion
Past history of convulsion
Multiple seizures
Focal seizures
Extracranial focus of infection
Seizure > 15 min
Unrousable coma
5 (14.7) 17 (3.5)
34 (6.5) 488
7 (8.6) 15 (3.4)
81 (15.5) 441 6 (3.3) 16 (4.7)
9 (18.4) 13 (2. 7)
49 (9.4) 473
180 (34.5) 342
14 (6.3) 8 (2.7)
18 (8.6) 4 (1.3)
18 (6.9) 4 (1.5)
221 (42.3) 301
208 (39.9) 314
259 (49.6) 236
8 (34.8) 14 (2.8)
23 (4.4) 499
5 (2.3) 1 (0.3) 3 (6.1) 3 (0.6) 3 (3.7) 3 (0.7)
9 (4.1) 7 (2.3) 6 (12.2) 10 (2.1) 4 (4.9) 12 (2.7)
3 (8.8) 3 (0.6)
4(1.9) 2 (0.6)
14 (6.7) 4 (1.3)
2 (5.9) 14 (2.9)
6 (2.3) 0 (0.0)
12 (4.6) 4 (1.5)
3 (1.7) 3 (0.9)
1 (4.3) 5 (1.0)
7 (30.4) 9 (1.8)
3 (1.7) 13 (3.8)
Meningitis with no meningeal signs (n=6)
Meningitis with meningeal signs (n = 16)
9.9 (
