What Proportion of Children and Young People with Bacterial Meningitis Develop Physical

What proportion of children and young people with bacterial meningitis develop physical and psychological morbidity?

Bibliographic information / Study type & evidence / Study details / Number of patients / Characteristics of studies/patients / Data collection methods / Follow up & outcome / Results & effect size / Reviewer comment
Bedford
et al.,
2001 162 / Cohort study
EL 2+ / Aim: To describe seqelae at 5 years of meningitis in first year of life compared to matched controls.
Setting: England and Wales / Cases: N=1717 eligible
N=1485 recruited / Population: Survivors of acute meningitis in infancy (between 1985 and 1987 in England and Wales)
Study sample:
Cases: Parents and GPs of survivors of acute meningitis in infancy between 1985 and 1987
N=1584
Pathogens:
H. influenzae= 413 (26%)
Neisseria meningitidis= 402 (25%)
S. pneumoniae= 143 (9%)
E. coli= 70 (4%)
Echovirus= 31 (2%)
Other virus= 19 (1%)
Other Gram +ve bacteria= 29 (2%)
Other Gram –ve bacteria= 32 (2%)
L. monocytogenes= 16 (1%)
Other microorganism= 7 (0.5%)
No organism grown= 320 (20%)
Not known= 4 (0.2%)
Controls: Parents and GPs of age and sex matched children from same GP lists.
N=1391 / GPs asked to complete specially designed questionnaire including info on:
developmental problems in:
neuromotor development, learning, vision, hearing, speech and language, behaviour;
seizure disorder
Parents asked to complete questionnaire on child’s health, development and learning.
Info from GPs and parents combined. Where severity of a condition differed between GP and parent, most severe cited category chosen. / Disability at age 5 classified as:
severe if child unable to attend mainstream school; moderate if disability impaired functioning but not associated with severe intellectual or developmental impairment; mild disorder if child has condition prevalent among children of same age but not typically associated with meningitis; no disability if no evidence of developmental problem. / Cases vs controls:
Learning difficulties:
RR (95% CI) = 7.0 (4.1 – 11.8)
Neuromotor disabilities
RR (95% CI) = 8.6 (4.9 to 15.2)
Seizure disorders
RR (95% CI) = 2.7 (1.9 to 3.9)
Hearing problems
RR (95% CI) = 1.9 (1.6 to 2.2)
Sensorineural hearing loss
RR (95%CI) 22.8 (7.22 to 72.1)
Ocular or visual disorders
RR (95% CI) = 3.4 (2.6 to 4.6)
Speech or language problems, or both
RR (95% CI) = 3.5 (2.8 to 4.6)
Behavioural problems
RR (95% CI) = 3.6 (2.6 to 4.9)
Cerebral palsy:
Cases= 79 (5.3%), Controls= 2 (0.1%)
Hemophilus influenzae:
Severe disability= 14 (3.4%)
Moderate disability= 30 (7.3%)
Mild disorder= 134 (32.5%)
No disability= 235 (57.0%)
Neisseria meningitidis:
Severe disability= 12 (2.9%)
Moderate disability= 26 (6.5%)
Mild disorder= 120 (29.8%)
No disability= 244 (60.7%)
Streptococcus pneuomniae:
Severe disability= 14 (9.7%)
Moderate disability= 20 (13.9%)
Mild disorder= 37 (25.8%)
No disability= 72 (50.3%)
Escherichia coli:
Severe disability= 4 (5.7%)
Moderate disability= 13 (18.6%)
Mild disorder= 18 (25.7%)
No disability= 35 (50.0%)
Group B Streptococcus:
Severe disability= 13 (13.3%)
Moderate disability= 17 (17.3%)
Mild disorder= 18 (18.4%)
No disability= 50 (51.0%)
Other gram positive bacteria:
Severe disability= 6 (20.6%)
Moderate disability= 8 (27.6%)
Mild disorder= 5 (17.2%)
No disability= 10 (34.5%) / Pre Hib vaccine; a quarter of cases had meningitis due to Hib.
Subjects of this study were the survivors from a national incidence study of infantile meningitis in England and Wales between 1985 and 1987. The same subjects were used in Halket et al., (2003) 182 and de Louvois et al., (2007) 183.
Halket
et al., 2003 182 / Cohort study
EL 2+ / Aim: To assess how meningitis in first year of life affects teenage behaviour.
Setting: England and Wales / 739 cases, 480 controls / Cases: Survivors from a national incidence study of infantile meningitis (between 1985 and 1987 in England and Wales) aged 4 to 16.
Controls: age and sex matched from same GP lists.
Mean age: 13.3 years (SD 0.4)
The incidence of each strain of meningitis was not specified. Although the subjects of this study were from the same cohort as Bedford et al., (2001) 162 (who did report the incidences) and de Louvois et al., (2007) 183, not all of the cohort was used. / A postal questionnaire (the Strengths and Difficulties Questionnaire) was sent to parents and school teachers.
25 questions divided into categories on emotional symptoms, conduct problems, hyperactivity, peer problems and prosocial behaviour.
Also questions on the impact of the child’s behaviour on the family or classroom. / A total deviance score of 0-13 is considered ‘normal’. A total prosocial behaviour score of 10-6 is considered ‘normal’.
Impact score from 0 (not at all) to 2 (a great deal). Maximum burden scores are 10 for parents and 6 for teachers.
Complicated meningitis group: one or more of the following – meningitis diagnosed at <28 days, birth weight <2,000g, coma, convulsions, hydrocephalus, a temperature >40oC, ventriculitis, relapse. / Complicated meningitis vs. controls:
Parent’s response:
Not normal total deviance:
RR (95% CI) = 2.18 (1.77-2.68)
Not normal impact score:
RR (95% CI) = 3.48 (2.56-4.73)
Teacher’s response:
Not normal total deviance:
RR (95% CI) = 1.62 (1.27-2.08)
Not normal impact score:
RR (95% CI) = 1.59 (1.25-2.03)
Uncomplicated meningitis vs. controls:
Parent’s responses:
Not normal total deviance:
RR (95% CI) = 1.79 (1.44-2.22)
Not normal impact score:
RR (95% CI) = 2.46 (1.78-3.39)
Teacher’s responses:
Not normal total deviance:
RR (95% CI) = 1.45 (1.13-1.86)
Not normal impact score:
RR (95% CI) = 1.44 (1.13-1.84)
All meningitis cases vs. control; total deviance ratings:
Parent and teacher rate as normal:
RR (95% CI) = 0.73 (0.64-0.83)
Parents and teacher rate as not normal:
RR (95% CI) = 2.18 (1.56-3.04)
Parent rates as normal, teacher as not normal:
RR (95% CI) = 0.99 (0.68-1.45)
Parent rates as not normal, teacher rates as normal:
RR (95% CI) = 2.09 (1.43-3.06)
All meningitis cases vs. control; normal social skills score:
Parents:
RR (95% CI) = 0.82 (0.73-0.91)
Teachers:
RR (95% CI) = 0.88 (0.80 to 0.98) / Missing information:
Total deviance scores:
Complicated meningitis – 3 parents and
83 teachers
Uncomplicated meningitis - 85 teachers
Control – 129 teachers
Impact score:
Complicated meningitis – 15 parents and 92 teachers
Uncomplicated meningitis – 8 parents and 93 teachers
Control – 19 parents and 139 teachers
Subjects of this study were the survivors from a national incidence study of infantile meningitis in England and Wales between 1985 and 1987. The same subjects were used in Bedford et al., (2001) 162 and de Louvois et al., (2007) 183.
de Louvois
et al., 2007 183 / Cohort study
EL 2+ / Aim: To assess whether meningitis in first year of life adversely affects academic achievement at age 16. / 460 cases, 288 controls / Cases: 16 year olds who had had meningitis in infancy (between 1985 and 1987)
Controls: age and sex matched from same GP lists.
The incidence of each strain of meningitis was not specified. Although the subjects of this study were from the same cohort as Bedford et al., (2001) 162 (who did report the incidences) and Halket et al., (2003) 182, not the entire cohort was used. / Pupils were asked to list all the GCSE examinations they had taken, the number passed and the grades obtained through use of a standard questionnaire. / Passes in five key subjects (English language, English literature, mathematics, science and a modern foreign language) were scored (grade A*= 6 points, grade A= 5 points… grade E pass= 1 point). Five passes (C or above) at GCSE equated to fifteen points. / GCSE point score in five key subjects:
0:
Index= 67 (14.6%)
Control= 10 (3.5%)
1-14:
Index= 190 (41.3%)
Control= 75 (26.0%)
15-24:
Index= 130 (28.3%)
Control= 123 (42.7%)
>24:
Index= 73 (15.9%)
Control= 80 (27.8%)
Number of GCSE passes:
0:
Index= 117 (25.4%)
Control= 19 (4.13%)
1-4:
Index= 105 (22.8%)
Control= 41 (14.2%)
5-10:
Index= 198 (43.0%)
Control= 189 (65.6%)
>10:
Index= 40 (8.7%)
Control= 39 (13.5%)
There was a significant difference in mean number of GCSE passes between index and control in comprehensive schools (5.05 SD 4.1 vs. 6.88 SD 3.5, p<0.0001).
There was no significant difference in mean number of GCSE passes between index and control in grammar or independent schools. / The study reports 461 cases and 289 controls, but the analyses does not include 1 case and 1 control as they did not attend a comprehensive, independent, grammar or special school (coded as ‘other’).
Study done prior to Hib vaccine – almost a third of the 16 year olds who attended special schools (total n= 36) had had meningitis due to H influenzae
Legood et al., 2008 166 / Cohort study
EL 2+ / Aim: to estimate the overall long-term health related quality of life implications of an episode of pneumococcal meningitis in childhood. / 70 cases
66 controls / Cases: Children aged 5 and over who had S. pneumoniae disease in Oxford and North East Thames regions of the UK.
Controls: 61 controls were siblings of cases, 5 were neighbourhood controls of similar age and same sex.
Mean age:
Cases= 9.9 years (5.4-20.4)
Controls= 11.3 years (5.3-28.8)
Pathogens: S. pneumoniae= 70 (100%) / HUI-3 :Health utility index that measures health related quality of life.
Measures health related quality of life.
Parent completed questionnaire if child was under 11. / HUI score for vision, hearing, speech, ambulation, dexterity, emotion, cognition and pain. / Mean HUI in cases vs. controls:
Vision:
0.981 (0.952-1.009) vs. 0.992 (0.988-0.997)
p= 0.434
Hearing:
0.930 (0.886-0.974) vs. 0.996 (0.987-1.004)
p= 0.005
Speech:
0.976 (0.945-1.006) vs. 0.995 (0.985-1.005)
p= 0.248
Ambulation:
0.986 (0.957-1.014) vs. 1.000 (1.000-1.000)
p= 0.333
Dexterity:
1.000 (1.000-1.000) vs. 1.000 (1.000-1.000)
P= 1.000
Emotion:
0.915 (0.877-0.954) vs. 0.942 (0.911-0.972)
p= 0.297
Cognitive:
0.871 (0.822-0.921) vs. 0.916 (0.877-0.955)
p= 0.167
Pain*:
0.952 (0.927-0.978) vs. 0.972 (0.956-0.987)
p= 0.203
Overall score*:
0.774 (0.711-0.837) vs. 0.866 (0.824-0.907)
p= 0.019 / *= data is missing for one case
Univariate analyses were conducted for each attribute with no correction for multiple comparisons (e.g. Bonferroni correction), and so the significance levels reported may over estimate the true effects. Also, the significance of the overall score probably reflects the effect of hearing.
This study was done in the UK. Funding was received from the Meningitis Research Foundation.
Anderson
et al., 2004 167 / Prospective cohort study
EL 2+ / Aim: to investigate long-term neurobehavioural outcome from childhood bacterial meningitis. / 7 year followup: 130 cases, 130 controls
12 year followup:
109 cases,
96 controls / Cases: children aged 3 months to 14 years admitted to Royal Children’s Hospital, Melbourne, with bacterial meningitis from October 1983 to October 1986.
Median age at illness was 18 months.
Children with documented pre-existing neurologic and developmental deficits, immunodeficiency states, previous CNS surgery or meningitis secondary to cranial trauma or shunt infections were excluded.
Controls: Grade and sex matched controls recruited from the classroom of each case child (one control was from neighbouring school). Controls for cases at special schools were taken from another school in the same region.
Pathogens:
Haemophilus influenzae type B= 85 (78%)
Staphylococcus pneumoniae= 12 (11%)
Neisseria meningitidis= 6 (5.5%)
Other= 6 (5.5%) / Wechsler Intelligence Scale for Children-III for children 16 years and younger, Wechsler Adult Intelligence Scale-III for children aged 17 and 18.
Full Scale Intellectual Quotient
Wide Range Achievement Test-3 / Outcomes: verbal comprehension, perceptual organisation and freedom from distractibility, reading (word decoding), spelling and arithmetic ability. / Full scale IQ, mean (SE):
Case= 97.2 (1.1)
Control= 101.6 (1.2)
Difference: 4.3 (95% CI; 1.1-7.6) p=0.10
Verbal comprehension, mean (SE):
Case= 95.0 (1.1)
Control= 99.4 (1.2)
Difference: 4.3 (95% CI; 1.1-7.6) p=0.009
Perceptual organisation, mean (SE):
Case= 99.4 (1.3)
Control= 103.6 (1.4)
Difference: 4.1 (95% CI; 0.4-7.9) p= 0.029
Freedom from distractibility, mean (SE):
Case= 97.7 (1.4)
Control= 99.7 (1.5)
Difference: 2.0 (95% CI; -2.0 to 6.0) p=0.323
Reading ability, mean (SE):
Case= 99.0 (1.3)
Control= 104.3 (1.4)
Difference: 5.2 (95% CI; 1.4-9.1)
P= 0.007
Spelling, mean (SE):
Case= 95.4 (1.3)
Control= 101.3 (1.3)
Difference: 5.9 (95% CI; 2.2-9.5) p= 0.002
Arithmetic, mean (SE):
Case= 95.0 (1.2)
Control= 97.4 (1.2)
Difference: 2.5 (95% CI; -0.9 to 5.8) p=0.146
The age at which children developed meningitis was not a significant predictor of long-term health related quality of life, although meningitis before 12 months of age was significantly related to poorer performance on tasks requiring language and executive skills. / *= n= 107 as 2 cases were unable to complete the tests
This study was done in Australia and was funded by the Government Employee’s Medical Research Fund and the Murdoch Childrens Research Institute.
Oostenbrink
et al., 2002 168 / Retrospective cohort study
2+ / Aim: to evaluate the neurological outcome of bacterial meningitis in children
Setting: One hospital in Rotterdam, The Netherlands, between 1988 and 1998. / 103 children with bacterial meningitis / 1 month to 15 years old.
51 male (50%)
Pathogens:
N. Meningitidis = 51 (50%)
Steptococcus pneumoniae (SP) = 10 (10%)
H. influenzae type B= 34 (33%)
No pathogen identified= 8 (8%) / Clinical records / Follow-up:
Median= 6.7 months
Outcomes:
Neurological and audiological sequelae / Death= 2 children (2%)
Number of children whose persistent neurological sequelae were assessed during follow-up= 13
Number of children whose hearing function was assessed= 83
Hearing loss= 7
Neurological sequelae= 7
Total sequelae= 13
Deafness= 1
Mild hearing loss= 6
Mental retardation= 5
Persistant palsy of the abducens nerve= 3
Locomotion deficits= 3
Epilepsy= 1 / This study was done in the Netherlands; no source of funding is reported.
Ritchi et al.,
2008 169 / Cohort study
2+ / Aim: To examine behaviour problems, personality, self-perceived competence and academic deficits in children who had recovered from non-Haemophilus influenzae type b bacterial meningitis without obvious medical sequelae. / 674 children with certified diagnoses of bacterial meningitis
Cohort was divided by whether parents were concerned about school achievement and health status. From both groups, samples of children with (n= 100) and without (n= 101) were drawn for further assessment.
After checking exclusion criteria, final cohort of children with (n= 84) and without (n= 98) academic or behavioural problems (total n= 182). / 57% male
Average age at onset of meningitis= 2.4 years (1 mo to 9.4 years)
Mean age: 10 years (5-14yrs)
Pathogens:
Streptococcus pneumoniae= 26 (14%)
Neisseria meningitidis= 146 (80%)
Other= 10 (5%) / Part II of the Child Behaviour Checklist (completed by parents)
Personality Questionnaire for children (completed by parents)
Dutch adaptation of the Self-Perception Profile for children (completed by children)
Academic Achievement Test (completed by children, involves writing to dictation, copying sentences, reading aloud and solving written arithmetic problems) / Follow-up:
4-10 yrs after surviving meningitis
Outcomes:
Ratings for behaviour, personality and self-perceived confidence.
Academic deficits. / Total behavioural problem score (n= 61, 9%): d=0.52* (95% CI: 0.30-0.74, p<0.001)
Estimated percentage of children with behaviour problems after surviving bacterial meningitis= 9%
Deviation on at least 2 of 4 academic deficit tasks: n= 184 (27%)
Deficit in writing to dictation: n= 258 (38%)
Deficit in reading aloud: n= 159 (24%)
Deficit in copying sentences: n= 116 (17%)
Deficit in written arithmetic: n= 222 (33%) / *= effect size of 0.2 is considered to be a small deviation, 0.5 a moderate deviation, and 0.8 a large deviation from the norm.
Mean effect sizes in the total cohort (n= 674) were estimated from effect sizes and percentages in a cohort with (n= 84) and without (n= 98) academic or behavioural problems.
This study was done in the Netherlands. Funding was provided by the Health Research and Development Council of the Netherlands and from Fonds Bevordering Neuropsychologisch Onderzoek bij Kinderen (Fund for the advancement of neuropsychological research in children).
Douglas et al.,
2008 170 / Retrospective cohort study
2+ / Aim: to demonstrate whether one causative agent of meningitis is more likely to cause profound hearing loss and labyrinthitis ossificans
Setting: One territory of Australia between 1984-2005 / 35 of 59 patients who received cochlear implants after meningitis and for whom a causative agents could be confirmed from medical records.
70 ears (11 bilateral cochlear implantations) / 22 females, 37 males
Mean age at time of deafness: 2yrs 9mos (4mos – 30 yrs 9 mos)
51 of 59 patients were under 3 yrs old.
Pathogens: Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae. / Notifiable Diseases Database System of the New South Wales health department
Australian National Centre for Immunisation Research and Surveillance of Vaccine Preventable Diseases
Sydney Cochlear Implant Centre / Outcome:
Incidence of cochlear implant by causative agent of meningitis. / Total cases of meningitis= 1568
Total number of cochlear implant patients= 80
Cause of meningitis (total n= 1568):
Streptococcus pneumoniae= 645 (41.1%)
Neisseria meningitidis= 892 (56.9%)
Haemophilus influenzae= 31 (1.9%)
Cochlear implant by causative agent (total n= 59, n from Streptococcus pnuemoniae, Neisseria meningitidis or Hib= 35):
Streptococcus pneumoniae= 30 (85.7%)
Neisseria meningitidis= 4 (11.4%)
Haemophilus influenzae= 1 (2.9%)
Incidence of cochlear implant for each causative agent (total n= 1568):
Streptococcus pneumoniae= 30 (4.7%)
Neisseria meningitidis= 4 (0.4%)
Haemophilus influenzae= 1 (3.2%) / This study was done in Australia. Funding was provided by the Graham Fraser foundation (UK).
Bibliographic information / Study type & evidence level / Study details / Number of patients / Patient Characteristics / Intervention & Comparison / Follow up & outcome / Results & Effect size / Reviewer comment
Wellman
et al., 2003 177 / Retrospective case series
EL= 3 / Aim: to establish the proportion of children who develop sensorineural hearing loss after bacterial meningitis and to correlate such loss with patient factors / N= 79, of which 68 had an audiological assessment and were included in the final analysis / Age:
<2 years= 58/79 (73.4%)
2-5 years= 13/79 (16.4%)
>5 years= 8/79 (10.2%)
Male= 57/79 (72%)
Cases were identified by searching medical records
Children from 1 day to 18 years old with a confirmed diagnosis of bacterial meningitis who underwent hearing assessment either as inpatients (n= 42, 61.7%) or as outpatients (n= 26, 38.3%)
Causative agent:
S. pneumoniae= 29/79 (36.7%)
N. meningitidis= 13/79 (16.5%)
Group B Strep.= 12/79 (15.2%)
H. influenzae= 11/79 (13.9%)
E. coli= 7/79 (8.87%)
Excluded if a bacterial etiology could not be confirmed (203 cases), death (9 cases) or had a pre-existing neurologic or sensorineual hearing loss (14 cases). / Audiological assessment
Levels of hearing:
Normal: < 30 dB
Mild: 30 to 55 dB
Moderate: 55 to 70 dB
Severe: 70 to 90 dB
Profound: > 90 dB / Mean number of days between admission and assessment (inpatients, n= 42)= 13.2 days (±7.25)
Mean number of days between discharge and assessment (outpatients, n= 26)= 74.3 days (±13.8) / Some degree of loss= 22/68 children (32.3%)
Permanent sensorineual hearing loss= 11/17 of the children who were followed up (64.7% of those followed up, 16.1% of the total number of children undergoing audiological assessment)
A statistically significant association between Streptococcus pneumoniae meningitis and sensorineural hearing loss was found (p<0.001)
No statistically significant differences between sensorineural hearing loss and patients’ age, sex, duration of illness before admission, and pathogens other than S. pneumoniae were found. / Bacterial meningitis: presence of a positive CSF culture, an accompanying positive CSF gram stain or a consistent CSF profile defined as CSF glucose level less than 2.12 mmol/L, CSF protein level > 1490 mg/L, and polymorphonuclear cell count > 2500 cells/ μL
This study was done in Canada. No source of funding is cited.
Grimwood
Et ol 1995 171 / Prospective cohort study
EL= 2+ / Aim: to determine the outcomes of bacterial meningitis in school-age survivors / 158 survivors of bacterial meningitis, with 130 completing followup (131 cases as one child had meningitis twice)
Grade and sex matched controls – next same sex student on the class roll or from equivalent school in same area. / Male= 83 (53%)
3 months to 14 years old
Median age at admission= 17 months
Bacterial meningitis: clinical presentation and either 1) the isolation of bacteria from the CSF by culture, 2) CSF leukocytosis (>100x106/L) and abnormal CSF biochemistry (glucose <45mg/dL [<2.5 mmol/L] and protein>50 mg/dL [>0.5g/L]) with isolation of bacteria from blood culture or CSF antigen detection by latex agglutination or 3) CSF leukocytosis (>1500 x 106/L with >75% neutrophils) and abnormal CSF biochemistry.
Excluded: children with known pre-existing neurologic or developmental abnormalities, immunodeficiency states, previous central nervous system surgery, or meningitis secondary to cranial trauma or CSF shunt infections. Also excluded at follow up those who died during the illness.
Hib= 100/131 (76%)
S. pneumoniae= 18/131 (14%)
N. meningitidis= 6/131 (5%) / Medical examination, neuropsychological evaluation, audiologic assessment, behaviour and social adjustment assessment, socio-demographic assessment. / A questionnaire for demographic data, presenting history and examination findings, laboratory and treatment details, and clinical course was completed during hospitalisation and at the discharge neurologic examination.
If a change in auditory responsiveness was suspected during recovery, an initial audiological assessment was made in the hospital. All other patients received their first audiologic assessments within 4 to 6 weeks from discharge. / Full scale IQ <70:
Survivors= 11/130 (8.5%)
Controls= 0/130 (0%)
P < 0.001
Spasticity:
Survivors= 3/127 (2%)
Controls= 0/129 (0%)
P not reported
Blind:
Survivors= 1/127 (1%)
Controls= 0/129 (0%)
P not reported
Deaf (severe to profound):
Survivors= 3/126 (3%)
Controls= 0/129 (0%)
P not reported
Epilepsy:
Survivors= 5/127 (5%)
Controls= 0/129 (0%)
P not reported
VP shunt:
Survivors= 2 (2%)
Controls= 0 (0%)
P not reported
No problems:
Survivors= 95 (73%)
Controls= 116 (89%)
P not reported
One minor problem (e.g. IQ 70-80, mild to moderate deafness):
Survivors= 16 (12%)
Controls= 14 (11%)
P not reported
More than one minor problem or at least one major problem (e.g. IQ<70, blind, severe to profound deafness):
Survivors= 20 (15%)
Controls= 0 (0%)
P not reported / All patients were initially treated with intravenous benzylpenicillin G (180 mg/kg per day) and chloramphenicol (100 mg/kg per day).
28 children were lost to followup (26 were unable to be found, 1 refused to participate and 1 had died from unrelated causes)