Background In 1999, invasive meningococcal disease was hyperendemic in Ireland at 14.75/100 000 population, with 60% group B and 30% group C diseases. National sepsis guidelines and meningococcal C vaccines were introduced in 2000. Despite a spontaneous decline in group B infection, invasive meningococcal disease remains a leading cause of sepsis. This study characterises the epidemiology of invasive meningococcal disease in children in Ireland since the introduction of meningococcal C vaccine and reviews its clinical presentation, hospital course and outcome in anticipation of meningococcal B vaccine introduction.
Methods National surveillance data were obtained from the Health Protection Surveillance Centre. A retrospective study of all meningococcal cases at two tertiary paediatric hospitals was conducted from 2001 to 2011. Records were reviewed using a standardised assessment tool. A study of 407 meningococcal cases published in 2002 provided comparative data.
Results Of 1820 cases <19 years of age notified nationally, 382 (21%) cases attended a study hospital; 94% group B, 3% group C, 225 (59%) male, median age 5 years (range 0.1–18). Fever was absent at presentation in 18%. Fifteen patients (3.6%) died. 221 (61%) were admitted to paediatric intensive care units (PICU). Permanent sequelae occurred in 9.4%. Compared with the historical cohort, there were differences in presentation, an increase in PICU interventions, but no significant decline in morbidity or mortality.
Conclusions Despite the meningococcal C vaccination campaign, invasive meningococcal disease continues to cause serious morbidity and claim lives. Group B infections remain dominant. As children who die often present with fulminant disease, preventive strategies including use of meningococcal B vaccine are needed to avert death and sequelae.
- Invasive Meningococcal Disease
- Infectious Diseases
Statistics from Altmetric.com
What is already known on this topic?
Invasive meningococcal disease is a leading cause of paediatric sepsis and was hyperendemic in Ireland in the 1990s.
Meningococcal C vaccine was included in the routine immunisation schedule in Ireland in 2000, almost eradicating serogroup C disease.
There has been a spontaneous decline in serogroup B disease in Ireland since 2000.
What this study adds?
Despite the decline in incidence, morbidity and mortality of invasive meningococcal disease have not improved significantly.
94% of cases were due to capsular group B meningococcus.
Children who die are critically ill at presentation, highlighting the need for meningococcal B vaccination.
Invasive meningococcal disease (IMD) was hyperendemic in Ireland in 1999, with an annual incidence of 14.75/100 000 population. Serogroup C infections accounted for 30% of cases at that time.1
Ireland was the second country to introduce conjugate meningococcal C vaccines (MCV) in October 2000. MCV was included in the routine childhood primary immunisation schedule, and a catch-up programme was also implemented, targeting all those <23 years old. National uptake of MCV was 70% and rose to 90% in 2011 at 12 months of age.2
Similar to international experience,3–5 an immediate and statistically significant effect on IMD rates was noted, with year on year decreases from 2001. This fall in serogroup C disease coincided with a spontaneous decline in serogroup B disease. IMD surveillance has been carried out since 1997 by the Irish Meningococcal and Meningitis Reference Laboratory (IMMRL) in conjunction with the Health Protection Surveillance Centre (HSPC). The annual IMD incidence dropped from 14.75/100 000 in 1999 to 2/100 000 in 2011, with 98% of disease caused by serogroup B and a national case fatality rate (CFR) of 3.6%.6
The turn of the millennium also saw an emphasis on the appropriate and timely management of paediatric sepsis with the publication of national and international guidelines. The Department of Health and Children in Ireland first published guidelines for the management of suspected meningococcal disease in 1999,7 and this coincided with a personal practice statement in the UK which would help shape further national guidelines.8 ,9 The Surviving Sepsis Campaign was initiated in 2002, publishing a number of evidence-based guidelines.10 All of these guidelines highlighted the importance of early recognition of sepsis with aggressive resuscitation and appropriate transfer to a specialist paediatric intensive care unit (PICU).
The aim of this study was to characterise IMD epidemiology in children in Ireland since MCV introduction and to review the clinical presentation, hospital course and outcome of children with confirmed IMD admitted to the two national tertiary paediatric centres in Ireland in anticipation of future introduction of MenB vaccination.
The HSPC publish annual reports on IMD, and these were accessed to ascertain national incidence and mortality rates.6
All cases of laboratory confirmed IMD from Dublin's two tertiary referral paediatric hospitals, Our Lady's Children's Hospital, Crumlin and the Children's University Hospital, Temple Street, over the 11-year period 2001–2011, were identified from the IMMRL database. These two centres are the national centres for PICU care and serve the local catchment area, with approximately 84 000 children (collectively) attending their emergency departments (ED) per annum. The IMMRL database includes all cases of sterile site culture and/or PCR-confirmed IMD in Ireland.
Specific meningococcal PCR testing of blood and cerebrospinal fluid (CSF) samples and typing of isolates were performed at the IMMRL. A PCR, specific to Neisseria meningitidis, was used followed by a capsular grouping PCR. Initially, an in-house PCR assay was used, and this was subsequently replaced by a real-time PCR (RT-PCR) assay; the target of this first stage PCR was the citrA gene. All positives were then subjected to a second capsular grouping PCR. Cases were thus categorised based on either traditional serotyping or capsular grouping and will simply be referred to as either group B or group C disease for the remainder of the article.
To reduce the potential for interobserver bias, all cases were reviewed by a single observer in each hospital (CÓM and RJD), using a standardised assessment tool. Clinical presentation data were obtained from ED records, the first hospital note or the interhospital transfer letter, as appropriate. Laboratory data were recorded from the first available sample on the hospital system or from results accompanying the patient on transfer.
A previous study in children by Healy et al11 of 407 IMD cases from the prevaccine era of 1995–2000 provided comparative data with regard to clinical presentation, course and outcome. To facilitate comparisons, similar markers of illness severity and therapies administered in PICU were recorded in the current study.
Study approval was obtained from the medical research ethics committees of each hospital.
A case of IMD was defined as isolation of N. meningitidis from a normally sterile site or detection of specific meningococcal nucleic acid by PCR in blood, CSF or joint fluid. For the purposes of analysis, bloodstream infection (BSI) was defined as a positive blood culture and/or positive blood PCR test. Meningitis was defined as a positive CSF culture and/or positive CSF PCR, or presence of CSF pleocytosis with an abnormal white to red cell count ratio of >1:700, or if gram-negative diplococci were seen on Gram stain.11
Residual morbidity was defined as a persisting significant deficit at the last recorded outpatient visit, including neurological deficit, sensorineural hearing loss, renal failure, permanent scarring or amputation.
Fever was defined as any temperature >38°C prior to admission or at triage in the ED. Rash was classified as blanching, petechial or purpuric.
For purposes of comparison, as for the previous study, patients were deemed to be in clinical shock if they required >10 mL/kg fluid resuscitation in the first hour following arrival into the ED, if tachycardia was present and capillary refill time was >3 s and/or if they were documented as ‘in shock’ by the admitting doctor.
A Mann-Whitney U test was used to analyse differences in age between patients with IMD before and after the implementation of the national MCV immunisation programme. A two-tailed z test was used to determine significance between the two groups for symptoms, treatment and outcome. A p value <0.05 was taken as statistically significant.
From 2001 to 2011 inclusive, 1823 IMD cases in patients under 19 years of age were notified to the HSPC, of which 382 (21%) were identified as being admitted to one of the two study hospitals and form the basis for this study.
In the study cohort, there were 225 (59%) male patients, with a male to female ratio of 1.43:1. The median age was 1.5 years (range 0.1–18 years). Group B accounted for 360 (94%) cases, and group C 12 (3%). There were five group W, two group Y, and in three cases, the group could not be determined. Age and gender distribution of patients were similar to the historic cohort.
Three hundred and eighty-two patients were identified; 331 (87%) presented through the ED of a study hospital, and 51 (13%) were transferred from other hospitals. Complete data were available for 353 patients. Charts were missing or the data incomplete in 29, of whom four were dead on arrival at the hospital or resuscitation in the ED failed (table 2).
Rash was petechial in 197/353 (56%), purpuric in 89/353 (25%) or macular and blanching in 21/353 (6%). Three patients had a documented rash with no description recorded. A higher proportion of patients with a purpuric rash (79%) was admitted to PICU, compared with patients with petechial rash (55%).
Of the 128/353 (36%) patients reviewed by a general practitioner (GP) prior to presentation, 41 (32%) received intramuscular penicillin. Two hundred and seventy-seven (78%) patients met the criteria for shock similar to 74% in the prevaccine era.
Site of infection/diagnosis
All 382 patients had information regarding at least one site of infection available. BSI was documented in 359 (94%) patients; 244 (64%) were diagnosed by PCR alone. Of the 115 patients with positive blood cultures, 106 also had PCR testing, of which three were negative. Meningitis was diagnosed in 130 patients; however, only 180 (47%) patients had a lumbar puncture compared with 70% of the patients in the historic cohort. BSI alone was diagnosed in 252 (66%) patients, meningitis alone in 23 (6%) patients, and both BSI and meningitis in 107 (28%). No cases of septic arthritis were identified.
Of the 331 patients admitted through the ED of a study hospital, 172 (52%) were admitted to PICU. All transferred patients were admitted to PICU. Median PICU length of stay was 20 hours (range, 1–736 hours). Median time to resumption of normal feeding was 24 hours (range, 0–1464 hours). The PICU interventions used, compared with the historic cohort, are listed in table 2. Fifty-eight (16%) patients received dexamethasone.
There were 58 IMD deaths in children <19 years of age reported to the HSPC during the study period (CFR 3.2%); 15 attended a study hospital, 4 of whom died in the ED or were dead on arrival (CFR 3.9%). Of the 331 direct admissions through the ED, 10 died or were dead on arrival (CFR 3%). Five of the fifty-one (10%) interhospital transfers died.
Of 15 deaths in the study hospitals, the median age was 1.5 years (range, 0.2–5), 9 were male, 14 were group B and 1 group Y. Inpatient data were available for 10 patients. The median length PICU stay was 5 hours (range, 1–217). Four of the ten children had been seen by a GP prior to arrival in hospital, and one received intramuscular penicillin.
Of 367 survivors, follow-up was documented in 340 (93%) children, of whom 32 (9%)––30 group B and 2 group C––had permanent sequelae (table 2).
The study cohort accounts for 21% of nationally notified IMD in <19-year-old patients during the study period. The results highlight the success of the conjugate MCV programme in Ireland, with no cases of group C disease reported nationally in 2011 and only two cases in our institutions from 2004 to 2011. Group B disease however, though reduced, remains a public health concern, accounting for almost 90% of reported IMD nationally in 2011. Introduction of an effective MenB vaccine will be a vital component of the strategy to eradicate IMD.
Thus far, the initial concerns regarding capsule switching or group replacement following MCV introduction have not been borne out.12 Only 1.6% of IMD in our study was caused by non group B or C strains. Improvements in PCR techniques have also ensured that the group is now identified in the majority of cases. The infecting group was not determined in only three cases, compared with 37 of the 447 cases originally identified in the prevaccine study. The impact of intramuscular penicillin on culture positivity is reviewed elsewhere.13
The current data reveal a change in the clinical presentation of IMD. Almost one-fifth of children with IMD had no documented fever prior to arrival at hospital. This presents a challenge to the clinician to ensure a high index of suspicion for IMD is maintained even in the absence of fever. The reason for this difference is not apparent, but could reflect changes in the use of antipyretics by parents.14 It is unlikely to be related to change in group representation within the cohort, as rates of fever for group B and C diseases were not significantly different in the prevaccine era.11 While it could represent recording bias, presence or absence of fever and rash was consistently documented throughout both study periods.
Seventy-one per cent of the patients had the classical petechial or purpuric rash on arrival to ED. In driving for earlier recognition of IMD, it is important to emphasise that the ‘classical’ haemorrhagic rashes are often a late manifestation. Although 10% of children had documented neck stiffness compared with 5% in the prevaccine cohort, lumbar puncture was performed in significantly fewer children, reflecting changes in clinical practice. No children presented with septic arthritis compared with 4/407 in the prevaccine cohort, three of whom had group C infection.
The proportion of patients (62%) admitted to PICU is similar to that in the prevaccine era cohort. There were however significantly more interventions in the postvaccine era group with an increase in ventilation, inotropic support and haemofiltration. The duration of time spent in the ICU and time to resumption of normal feeding were similar in both groups, suggesting that the increase in intervention more likely reflects the more aggressive pre-emptive approach to resuscitation rather than changed disease severity.
Consistent with international experience, 9% of survivors had permanent sequelae,15–18 with no significant difference in overall morbidity between the prevaccination and postvaccination era cohorts. Sensorineural hearing loss was significantly less common and amputation significantly more common in the postvaccine era; however, the small numbers prevent firm conclusions to be drawn based on these changes. Dexamethasone administration was not recorded in Healy's study, and as only 16% of our cohort received dexamethasone, its use is unlikely to have accounted for the decline in hearing loss.
The duration of follow-up was extremely variable. The 2012 meningococcal outcomes in adolescents and children study19 raises important questions about the adequacy of current follow-up for patients with IMD.
The CFR of 3.9% compares favourably with international data,20 ,21 and though it is a downward trend from the prevaccine era CFR of 4.6%, this change does not reach statistical significance. Ten of 15 children who died were dead on arrival or within 6 hours of admission to PICU. The higher mortality noted in the population referred from other hospitals likely reflects disease severity, consequent to which the patients were selected for transfer, and may also reflect a delay in availability of expert paediatric intensive care. The burden of disease falls on the young. The age-specific annual incidence for those <1 year of age stood at 41.4/100 000 population at the end of the study period.6
This study has a number of limitations. We acknowledge that our focus on tertiary centres could lead to over-representation of severe disease. Healy's study collected data from two regional hospitals as well as the two study hospitals, which may influence the dataset. As with all retrospective studies, the quality of the contemporaneous clinical records was variable. Approximately, 10% of the data were inaccessible due to missing charts or poor documentation. As 14% of patients were referred from regional centres, subsequent follow-up information for these patients was often incomplete. Even for those with documented follow-up, whether or not hearing was screened was unclear, and duration of follow-up was inconsistent, leading to possible underestimation of sequelae.
There has been a marked reduction in IMD in Ireland from 2001 to 2011, with both a natural decline in group B disease and the dramatic success of the MCV programme. Despite improvements in diagnostics and more active intervention in the ICU setting, disease outcome is relatively unchanged with group B disease accounting for very high rates of infection in the youngest children and significant mortality and morbidity.
Differences were noted in the presenting features of children with IMD in the post-MenC vaccine era. Absence of fever in 18% and absence of non-blanching rash in 29% highlight the importance of having a low threshold for suspecting IMD. That most children who die are dead before or within 6 hours of admission to PICU, despite aggressive management strategies, leads us to conclude that the only effective way to prevent deaths from meningococcal infection is to prevent the disease. MenB vaccine will be an important tool towards IMD eradication. The data presented support its inclusion in the national immunisation programme in Ireland.
The authors thank the staff at the IMMRL, particularly Ms Sandra Morgan, and the medical records departments of both hospitals for their assistance throughout this study. They also thank Professor Mary Healy for her valuable assistance in providing data from the prevaccine era and thank the HSPC for their invaluable data.
Contributors KB, MC and PG contributed to the design of the study; RJD and CÓM carried out the study and analysed the data; and all authors contributed to the manuscript and are accountable for all aspects of the work.
Competing interests None declared.
Ethics approval The Ethics Committees of the Children's University Hospital, Temple Street and Our Lady's Childnren's Hospital, Crumlin.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.