ORIGINAL CONTRIBUTION
meningitis, aseptic; meningitis, bacterial
Clinical Predictors of Bacterial Versus Aseptic Meningitis in Childhood From the Department of Pediatrics, Medical College of Wisconsin, Children's Hospital of Wisconsin, Milwaukee. Received for publication May 13, 1991. Revisions received November 12, 1991, and January 2, 1992. Acceptedfor
publication February 3, 1992.
Christine Walsh-Kelly, MD David B Nelson, MD, MSc Douglas S Smith, MD, MSc Joseph D Losek, MD, FACEP Marlene Melzer-Lange, MD Halim M Hennes, MD Peter W Glaeser, MD
Study objective: Toassess the reliability of meningeal signs and other physical findings in predicting bacterial and aseptic meningitis at various ages. Design: Children requiring lumbar puncture were evaluated prospectively for meningeal signs and other physical parameters before lumbar puncture. Setting: Emergency department of Children's Hespital of Wisconsin. Participants: One hundred seventy-two children, aged 1 week to 17 years, with meningitis (53 bacterial and 119 aseptic). Measurements and main results: Nuchal rigidity was present in 27% of infants aged 0 to 6 months with bacterial meningitis versus 95% of patients 19 months or older (P= .0001). Three percent of infants 0 to 6 months old with aseptic meningitis had nuchal rigidity versus 79% of patients 19 months or older (P= .0005). Seventy-two percent of infants 12 months of age or younger with bacterial meningitis had at least one positive meningeal sign versus 17% of infants with aseptic meningitis (P= .0001). Eighty-five percent of children older than 12 months with meningitis had at least one positive meningeal sign, 93% with bacterial meningitis, and 82% with aseptic meningitis. Conclusion: Despite a lack of meningeal signs, a high index of suspicion for meningitis is essential when evaluating the febrile infant 12 months of age or younger. [Walsh-Kelly C, Nelson DB, Smith DS, Losek JD, Melzer-Lange M, Hennes HM, Glaeser PW: Clinical predictors of bacterial versus aseptic meningitis in childhood. Ann EmergMed August 1992;21:910-914.]
24/910
ANNALS OF EMERGENCY MEDICINE 2 1 : 8
AUGUST1992
MENINGITIS Walsh-KeUy et al
I
INTRODUCTION Bacterial meningitis is a serious pediatric infection requiring expedient diagnosis and treatment to minimize morbidity and mortality. Opinions differ as to the age at which specific locahzing signs and symptoms of meningitis are commonly manifested. Lorber and Sunderland I stated that very few children more than 6 months old have bacterial meningitis without abnormal meningeal signs or evidence of serious illness. j However, Rutter and Smales 2 suggested that children less than 18 months old may not manifest meningeal signs or bulging fontanel in the presence of meningitis. Results of various studies conducted to determine the frequency of positive meningeal signs in childhood meningitis are difficult to i n t e r p r e t because of age differences in the population, indications for l u m b a r puncture, and etiology of meningitis.l-s This study was u n d e r t a k e n to determine the age at which meningeal signs are likely to be positive in both bacterial and aseptic meningitis. In addition, other variables such as altered mental status, toxic a p p e a r a n c e , a n d presence of shock were studied to determine their use in establishing the diagnosis of bacterial and aseptic meningitis~
MATERIALS AND METHODS From August 1985 through F e b r u a r y 1988, clinical data were recorded prospectively for all children undergoing lumbar puncture after examination by one of six pediatric emergency attending physicians in the emergency d e p a r t m e n t at Children's Hospital of Wisconsin, Milwaukee. The child's degree of illness was classified as well, mildly ill, toxic, and moribund. Mildly ill children were defined as having stable vital signs, decreased activity, or increased irritability but were responsive and consolable. Toxic children were defined as being lethargic, inconsolable, and uninterested in their environment and showing significant alterations in respiratory or heart rates or decreased p e r i p h e r a l perfusion. Moribund children were defined as being unarousable with poor p e r i p h e r a l perfusion and unstable vital signs. After the
first 100 patients were enrolled, an Infant Observation Scale score also was included for patients 24 months or younger.6 Nuchal rigidity was considered present if neck stiffness was noted with active and/or passive neck flexion. The Brudzinski sign was positive if passive neck flexion resulted in flexion of the legs (hips and knees). The Kerning sign was positive if complete extension of the leg was not possible or produced significant discomfort. After recording the clinical information, a l u m b a r puncture was performed. Final diagnoses were assigned at the resolution of the illness. A diagnosis of bacterial meningitis was made if cerebrospinal fluid latex agglutination or Gram stain was positive or if pathogenic bacteria grew from the cultured cerebrospinal fluid. A diagnosis of aseptic meningitis was made if the cerebrospinal fluid WBC count was 10 cells/ram3 or more in a neonate o r f i v e cells/mm 3 or more in a child older than 1 month, in the absence of a positive cerebrospinal Gram stain and/or latex agglutination and the absence of growth of pathogenic bacteria from the cerebrospinal fluid. To determine reliability of meningeal signs in meningitis relative to age, children with bacterial meningitis and aseptic meningitis were initially compared in four age groups: 0 to 6 months, 7 to 12 months, 13 to 18 months, and 19 months or older. F o r further analysis of both meningeal signs and other clinical variables, children were separated into two age groups: 0 to 12 months and older than 12 months. Comparison between groups was made using the Z2 statistic.
RESULTS F r o m August 1985 to F e b r u a r y 1988, 547 children underwent l u m b a r puncture: 341 (62%) were 0 to 12 months and 206 (38%) were older than 12 months. In infants 12 months or younger, the indication for l u m b a r puncture in more than haft (57%) was a fever without obvious source. I n this age group, positive meningeal signs and a full or bulging fontanel were indications for l u m b a r puncture in only 12% and 10%, respectively. By contrast, in children older than 12 months of age, 50% of patients underwent
Table 1. Clinical variables in meningitis by age
Bacterial Meningitis Months (11) (%)
7 to 12 Months (14) (%)
13 to 18 Months (8) (%)
55 27 18
33 71 50
NA 87 50
NA 95 75
36
93
62
45 45 73
93 36 86
87 50 75
0 to 6
Clinical Variables Bulging fontanel Nuchal rigidity Positive Kernig's sign Positive Brudzinski's sign One third positive (nuchaf rigidity, Kernig's sign, Brudzinski's sign) Toxic/moribund Lethargic/cematose
Aseptic Meningitis Older Than 18 Months (20) (%)
P*
6 to 6 Months (64) (%)
Months (9) (%)
13 to 18 Months (3) (%)
Older Than 18 Months (43) (%)
NS < .0001 < .003
14 3 6
0 22 11
NA 0 0
NA 79 30
NS
meningitis, aseptic; meningitis, bacterial
Clinical Predictors of Bacterial Versus Aseptic Meningitis in Childhood From the Department of Pediatrics, Medical College of Wisconsin, Children's Hospital of Wisconsin, Milwaukee. Received for publication May 13, 1991. Revisions received November 12, 1991, and January 2, 1992. Acceptedfor
publication February 3, 1992.
Christine Walsh-Kelly, MD David B Nelson, MD, MSc Douglas S Smith, MD, MSc Joseph D Losek, MD, FACEP Marlene Melzer-Lange, MD Halim M Hennes, MD Peter W Glaeser, MD
Study objective: Toassess the reliability of meningeal signs and other physical findings in predicting bacterial and aseptic meningitis at various ages. Design: Children requiring lumbar puncture were evaluated prospectively for meningeal signs and other physical parameters before lumbar puncture. Setting: Emergency department of Children's Hespital of Wisconsin. Participants: One hundred seventy-two children, aged 1 week to 17 years, with meningitis (53 bacterial and 119 aseptic). Measurements and main results: Nuchal rigidity was present in 27% of infants aged 0 to 6 months with bacterial meningitis versus 95% of patients 19 months or older (P= .0001). Three percent of infants 0 to 6 months old with aseptic meningitis had nuchal rigidity versus 79% of patients 19 months or older (P= .0005). Seventy-two percent of infants 12 months of age or younger with bacterial meningitis had at least one positive meningeal sign versus 17% of infants with aseptic meningitis (P= .0001). Eighty-five percent of children older than 12 months with meningitis had at least one positive meningeal sign, 93% with bacterial meningitis, and 82% with aseptic meningitis. Conclusion: Despite a lack of meningeal signs, a high index of suspicion for meningitis is essential when evaluating the febrile infant 12 months of age or younger. [Walsh-Kelly C, Nelson DB, Smith DS, Losek JD, Melzer-Lange M, Hennes HM, Glaeser PW: Clinical predictors of bacterial versus aseptic meningitis in childhood. Ann EmergMed August 1992;21:910-914.]
24/910
ANNALS OF EMERGENCY MEDICINE 2 1 : 8
AUGUST1992
MENINGITIS Walsh-KeUy et al
I
INTRODUCTION Bacterial meningitis is a serious pediatric infection requiring expedient diagnosis and treatment to minimize morbidity and mortality. Opinions differ as to the age at which specific locahzing signs and symptoms of meningitis are commonly manifested. Lorber and Sunderland I stated that very few children more than 6 months old have bacterial meningitis without abnormal meningeal signs or evidence of serious illness. j However, Rutter and Smales 2 suggested that children less than 18 months old may not manifest meningeal signs or bulging fontanel in the presence of meningitis. Results of various studies conducted to determine the frequency of positive meningeal signs in childhood meningitis are difficult to i n t e r p r e t because of age differences in the population, indications for l u m b a r puncture, and etiology of meningitis.l-s This study was u n d e r t a k e n to determine the age at which meningeal signs are likely to be positive in both bacterial and aseptic meningitis. In addition, other variables such as altered mental status, toxic a p p e a r a n c e , a n d presence of shock were studied to determine their use in establishing the diagnosis of bacterial and aseptic meningitis~
MATERIALS AND METHODS From August 1985 through F e b r u a r y 1988, clinical data were recorded prospectively for all children undergoing lumbar puncture after examination by one of six pediatric emergency attending physicians in the emergency d e p a r t m e n t at Children's Hospital of Wisconsin, Milwaukee. The child's degree of illness was classified as well, mildly ill, toxic, and moribund. Mildly ill children were defined as having stable vital signs, decreased activity, or increased irritability but were responsive and consolable. Toxic children were defined as being lethargic, inconsolable, and uninterested in their environment and showing significant alterations in respiratory or heart rates or decreased p e r i p h e r a l perfusion. Moribund children were defined as being unarousable with poor p e r i p h e r a l perfusion and unstable vital signs. After the
first 100 patients were enrolled, an Infant Observation Scale score also was included for patients 24 months or younger.6 Nuchal rigidity was considered present if neck stiffness was noted with active and/or passive neck flexion. The Brudzinski sign was positive if passive neck flexion resulted in flexion of the legs (hips and knees). The Kerning sign was positive if complete extension of the leg was not possible or produced significant discomfort. After recording the clinical information, a l u m b a r puncture was performed. Final diagnoses were assigned at the resolution of the illness. A diagnosis of bacterial meningitis was made if cerebrospinal fluid latex agglutination or Gram stain was positive or if pathogenic bacteria grew from the cultured cerebrospinal fluid. A diagnosis of aseptic meningitis was made if the cerebrospinal fluid WBC count was 10 cells/ram3 or more in a neonate o r f i v e cells/mm 3 or more in a child older than 1 month, in the absence of a positive cerebrospinal Gram stain and/or latex agglutination and the absence of growth of pathogenic bacteria from the cerebrospinal fluid. To determine reliability of meningeal signs in meningitis relative to age, children with bacterial meningitis and aseptic meningitis were initially compared in four age groups: 0 to 6 months, 7 to 12 months, 13 to 18 months, and 19 months or older. F o r further analysis of both meningeal signs and other clinical variables, children were separated into two age groups: 0 to 12 months and older than 12 months. Comparison between groups was made using the Z2 statistic.
RESULTS F r o m August 1985 to F e b r u a r y 1988, 547 children underwent l u m b a r puncture: 341 (62%) were 0 to 12 months and 206 (38%) were older than 12 months. In infants 12 months or younger, the indication for l u m b a r puncture in more than haft (57%) was a fever without obvious source. I n this age group, positive meningeal signs and a full or bulging fontanel were indications for l u m b a r puncture in only 12% and 10%, respectively. By contrast, in children older than 12 months of age, 50% of patients underwent
Table 1. Clinical variables in meningitis by age
Bacterial Meningitis Months (11) (%)
7 to 12 Months (14) (%)
13 to 18 Months (8) (%)
55 27 18
33 71 50
NA 87 50
NA 95 75
36
93
62
45 45 73
93 36 86
87 50 75
0 to 6
Clinical Variables Bulging fontanel Nuchal rigidity Positive Kernig's sign Positive Brudzinski's sign One third positive (nuchaf rigidity, Kernig's sign, Brudzinski's sign) Toxic/moribund Lethargic/cematose
Aseptic Meningitis Older Than 18 Months (20) (%)
P*
6 to 6 Months (64) (%)
Months (9) (%)
13 to 18 Months (3) (%)
Older Than 18 Months (43) (%)
NS < .0001 < .003
14 3 6
0 22 11
NA 0 0
NA 79 30
NS
