Prospective study of bacterial meningitis in North East Thames region, 1991-3, during introduction of Haemophilus influenzae vaccine

BMJ 1994; 309 doi: (Published 26 November 1994)
Cite this as: BMJ 1994;309:1412
  1. Gillian Urwin, lecturer in medical microbiologya,
  2. Mei F Yuan, research assistanta,
  3. Roger A Feldman, professor of clinical epidemiologyb
  1. a Department of Medical Microbiology, The London Hospital Medical College, London E1 2AD
  2. b Department of Epidemiology and Medical Statistics, The London Hospital Medical College at QMW, London E1 4NS
  1. Correspondence to: Professor Feldman.
  • Accepted 11 October 1994


Objective: To describe the epidemiology of primary bacterial meningitis in the North East Thames region over a three year period before and during the introduction of the vaccine for Haemophilus influenzae type b.

Design: Analysis of information on cases of primary bacterial meningitis identified by microbiology laboratories in the region, with collection of case data by questionnaire.

Main outcome measures: Annual incidence rates for types of meningitis according to age and ethnic group.

Results: The annual incidence rates for the three major causes of bacterial meningitis in the general population were 1.9/100000 for Neisseria meningitidis, 1.6/100000 for Haemophilus influenzae before vaccination, and 1.0/100000 for Streptococcus pneumoniae. Higher rates of H influenzae meningitis were found in Asians compared with white people (3.6/100000 v 1.5/100000, P = 0.01). As a result of the vaccine programme introduced in October 1992 the number of cases of H influenzae meningitis in children under 5 years has fallen by 87%.

Conclusions: Bacterial meningitis is a serious problem especially in preschool children. There are differences in the incidence of some causes of bacterial meningitis in different ethnic groups; with H influenzae type b being significantly more common among black and Asian people than among white people. The immunisation programme for H influenzae type b in the North East Thames region has been successful in reducing the incidence of this type of meningitis in Asian and white populations. The numbers were too small to evaluate in the black population.

Epidemiological implications

  • Epidemiological implications

  • Before immunisation was introduced Haemophilus influenzae type b was the most common cause of bacterial meningitis in preschool children in the North East Thames region

  • The Asian population had significantly higher rates of H influenzae disease than white people

  • A substantial reduction in the incidence of H influenzae meningitis was observed with the introduction of immunisation of infants with conjugate vaccine

  • Neisseria meningitidis is now the most common cause of meningitis in preschool children in the North East Thames region


Bacterial meningitis remains an important cause of morbidity and mortality, especially in children, with three agents causing most reported illness. In children Haemophilus influenzae type b has been the major cause of bacterial meningitis.1 2 Vaccination to prevent disease induced by H influenzae type b has led to a substantial reduction in the incidence of invasive disease in children in the United States3 and virtual elimination from Finland.4

Although meningitis is a notifiable disease, available case data have been incomplete. Reporting of H influenzae infections to the Public Health Laboratory Service was recently found to be 24% incomplete.5 During an outbreak of meningococcal meningitis in Gloucester, only 57% of cases were formally notified.6 For any study of bacterial meningitis active surveillance is essential and can be organised with the help of microbiology laboratories, public health facilities, and hospital consultants.

We organised a prospective study to evaluate the effect of the new H influenzae vaccine in the North East Thames region, a population of over 3.7 million with a large proportion of people from different ethnic groups. Consultant microbiologists in the region agreed to participate to identify cases of primary bacterial meningitis. The 1991 census provided up to date figures on the regional population, including data on ethnicity. Early infant immunisation against disease due to H influenzae was introduced in the second year of the study and allowed documentation of the impact of the immunisation programme on occurrence of the disease.


The microbiology consultants in the 19 microbiology laboratories in the region identified all clinical cases of bacterial meningitis in which there was documented microbiological evidence of the pathogen responsible for the disease by at least one of the following investigations: culture of blood or cerebrospinal fluid, cerebrospinal fluid Gram stain, or antigen detection.

Not included in the study were cases of septicaemia when there were no symptoms or clinical signs of meningitis; cases of proved bacterial meningitis associated with HIV infection; and bacterial meningitis related to neurosurgery and neurosurgical devices. The microbiologist at each centre forwarded the laboratory results of culture of blood and cerebrospinal fluid, with the organism, to the coordinating hospital.

Clinical and epidemiological details were collected by a questionnaire completed by the microbiologist or clinician in charge of the case. When this information was not returned within three months, an additional letter was sent. When there was no response to this, the details were collected from the case notes with permission from the consultant clinician. Laboratories were contacted on a regular basis to ensure that reporting of cases was maintained throughout the study period. The questionnaire and laboratory data were entered on a computer file and analysed by using EpiInfo 5.1.


Over the three year period a total of 518 cases of bacterial meningitis were identified (table I). Neisseria meningitidis, H influenzae, and Streptococcus pneumoniae were responsible for 87.6% of cases. These three major causes of bacterial meningitis were most common in children under 5 years, and especially in infants over 1 month of age (figure). For N meningitidis there are two peaks of disease, the first in children under 5 and a second smaller peak in 15-20 year olds. Only 6.5% of H influenzae meningitis occurred in those over the age of 15 years. S pneumoniae was most prevalent in children under the age of 5 years but continued as a cause of meningitis throughout adult life.

View this table:

Cases of bacterial meningitis by causative agent and year, North East Thames region, 1991-3

In the third year of the study there was a striking decline in the number of cases of H influenzae in children under 5 (table II). There were over 50 cases in 1991 and 1992. In 1993 there were seven cases in this age group, an 87% decline.


Cases of bacterial meningitis in 5 year age groups, North East Thames region, 1991-3

View this table:

Age specific annual incidence of bacterial meningitis for major causative agents, North East Thames region, 1991-3

There were no significant differences in the rates of meningitis due to N meningitidis among white people, Asians, and black people. In contrast, rates of H influenzae meningitis were significantly greater among Asians compared with white people (P=0.01, table III). Although there were higher rates of H influenzae meningitis in children under 1 year of age in both black and Asian children than in white children, the difference was not significant (table III).

View this table:

Incidence of bacterial meningitis by ethnic group and causative agent for all cases and those in children aged under 1 year, North East Thames region, 1991 and 1992


Although every effort was made to collect data on all cases of bacterial meningitis, the number of cases is an underestimate; patients presenting to hospitals outside the region and those dying before being admitted to hospital have not been included. In addition, clinical cases of pyogenic meningitis, when there is no microbiological evidence of the pathogen involved, have not been included in the study.

Data reported to the meningococcal reference laboratory gave annual incidence rates for the region of 1.0-2.8/100000.7 The incidence of 1.8-2.1/100000 in our study did not include cases of meningococcal septicaemia or other disease due to N meningitidis, which could have added 15% to the total. There was no significant difference in the annual number of cases of meningococcal meningitis identified in each of the three years. Age specific attack rates for meningococcal meningitis show an increased risk of bacterial meningitis in preschool children, with those under the age of 1 year at greatest risk. Prospective surveillance studies from the United States have found the highest attack rates to be in infants under 4 months (26.4/100000), with 29% of cases of meningococcal disease in children under 1 year.10

Meningitis accounts for 60% of invasive H influenzae type b disease,7 8 which in our region represents an annual incidence of 34.2 cases/100000 in children under 5. In Scotland, passive reporting gave an incidence of 22-25/100000/year in the under 5 age group,11 12 Wales had an incidence of 34.6/100000/ year, Oxford 33.4/100000/year, and five regions of England 20-33/100000/year.*RF 7-9, 13*

Denominator data concerning ethnicity were obtained from the 1991 census, which undercounted ethnic minorities more than other groups and undercounted urban areas, especially in London, more than other areas.14 The largest undercounting was in young adults. In comparing rates of H influenzae type b meningitis in various ethnic groups in infants, when undercounting is minimal, the rates were highest in the black and Asian groups, but the numbers were too small to show significance. In the population as a whole in the North East Thames region there was a significantly increased risk in the Asian population of H influenzae type b meningitis compared with that in white groups but no difference in the rates of meningococcal meningitis. Even if there were undercounting of 10% overall in the Asian subgroup the difference would be significant (P=0.02). Native American Indian and Innuit children under 5 years have high rates (254-282/100000/year) of H influenzae type b disease.15 In the United states, black and Hispanic people have a higher rate of this disease than white populations.1 Factors such as poverty, overcrowding, and use of health care16 may account for the increase in incidence in the Asian population but do not seem to affect the incidence of meningococcal meningitis in children under 1 year of age in this region.

The data from 1993 show the success of the H influenzae vaccine programme in the region. Uptake of immunisation in children at 2, 3, and 4 months has been 87-89%. In older children, the catch up programme has achieved uptake rates of 34% in children born in 1989, and 77% in children born in 1990.17 The seven cases of meningitis in children under 5 years is a significantly lower incidence than in previous years. This study highlights the success of the immunisation programme and the varying incidence of H influenzae disease in the different ethnic groups of the region.

We are most grateful to our colleagues in the North East Thames region, without whose help this study would not have been possible. We thank Dr D R Jones for allowing access to the typing results of the meningococcal isolates from the region; Professor J D Williams and Dr M Powell for assistance in development of the project; and Dr N Banatvala and Mr A McIntosh for their help with statistical analysis and data management. The project is funded by a grant from the Wolfson Foundation.