Streptococcus pneumoniae and Neisseria meningitidis remain the most common vaccine-preventable childhood invasive bacterial infections in New Zealand.1 Before routine Haemophilus influenzae type B (Hib) vaccination, Hib was a substantial cause of invasive bacterial disease in childhood.2 The clinical presentation of these infections is broad and includes meningitis, septicemia and septic shock.3–5
Since the early 1990s, vaccines for Hib, N. meningitidis and S. pneumoniae have been sequentially introduced onto the national immunization schedule in New Zealand.6 The Hib conjugate vaccine was introduced in January 1994 in a 3 + 1 schedule.7 A strain-specific meningococcal outer membrane vesicle vaccine (MeNZB) was introduced in July 2004 and discontinued in 2008, in response to a group B meningococcal disease epidemic where 86% of cases were because of a single subtype (P1.7b4).8 It was originally available in 3 doses to all New Zealanders under 20 years of age. In 2006, administration was restricted to infants, in 3 doses given 6 weeks apart, before being discontinued in 2008. High-risk groups were still able to receive the vaccine until 2011.9–13 The 7-valent pneumococcal conjugate vaccine (PCV7) was introduced in New Zealand in June 2008 using a 3 + 1 schedule. In 2011, this was changed to a 10-valent vaccine (PCV10), with the 13-valent vaccine (PCV13) available for high-risk groups. In 2014, PCV13 was made available to all children, with PCV10 no longer funded. In July 2017, the switch back to PCV10 was made, with PCV13 available only to children considered at high-risk of pneumococcal disease.7
The Institute of Environmental Science and Research (ESR) collects national surveillance data on notifiable diseases in New Zealand. Laboratories must send all invasive isolates of Hib, N. meningitidis and S. pneumoniae to ESR for verification and typing. ESR surveillance data have shown a reduction in Hib meningitis and epiglottitis and meningococcal meningitis notifications after the introduction of the Hib and MeNZB vaccines, respectively.9–11,14,15 However, long-term trends in hospital discharge rates have not been reported. Defining changes in the epidemiology of invasive bacterial infections is important to inform future vaccine research priorities and policy decisions about national immunization programs. Analysis of multiple data sources is important wherever possible to strengthen evidence of any changes that occur.
The aim of this study was to analyze long-term trends in notifications and hospital discharge rates for invasive bacterial infections caused by S. pneumonia (IPD), N. meningitidis and Hib in children from New Zealand <15 years of age, describe the impact of vaccines on rates of these diseases and compare rates in New Zealand to data from England.
In this population-based observational study, 2 datasets were analyzed to determine the rates of invasive bacterial infections in children from New Zealand (Table, Supplemental Digital Content 1, http://links.lww.com/INF/D160).
The National Minimum Dataset (NMDS) includes routinely collected data on hospital discharges from all private and public hospitals in New Zealand (termed hospitalizations). An extract from the NMDS was used that included diseases defined according to their International Classification of Diseases-9-CMA-II codes (Table, Supplemental Digital Content 1, http://links.lww.com/INF/D160) and equivalent codes for different International Classification of Diseases revisions.
The NMDS included ethnicity, gender, age at discharge, the year and month of discharge and encrypted national health index number, which is unique for each person. Only the first discharge was counted for patients with multiple discharges for the same disease if the events occurred within 1 year of each other.
EpiSurv is New Zealand’s national database for notifiable disease surveillance, operated by ESR on behalf of the Ministry of Health1 (termed notifications). EpiSurv collects data on laboratory-confirmed notified cases independently of NMDS data (listed in Table, Supplemental Digital Content 1, http://links.lww.com/INF/D160). No data were available on EpiSurv for pneumococcal septicemia.
Data were compared with a recent study from the United Kingdom, which evaluated long-term trends in notifications and hospital discharges.12 The United Kingdom has a similar public health system, with good mechanisms for data collection and a comparable infant vaccination schedule. In contrast to New Zealand, the United Kingdom has included vaccines for meningococcal disease in the national schedule since the 1999 introduction of MenC vaccination and have recently introduced a MenB vaccine.13
Annual age-specific and age-standardized hospitalization rates and notification rates were analyzed in children <15 years of age. Diagnoses in the EpiSurv dataset were separated into “meningitis,” “septicemia” or “epiglottitis.” Notified cases that were recorded as unknown for “meningitis,” “septicemia” or “epiglottitis” were not counted in the analysis for these diseases but were included in the analysis for total invasive disease for the 3 bacteria. The NMDS dataset did not include data for Hib disease, meningococcal disease or IPD. The rates for Hib diseases were calculated by adding the number of cases of haemophilus meningitis, epiglottitis and septicemia; the rates for meningococcal disease were calculated by adding the number of cases of meningococcal meningitis and meningococcemia; and the rates of IPD were calculated by adding the number cases of pneumococcal meningitis and septicemia.
The New Zealand estimated resident population at June 30 of each year was used, derived from Census night counts by Statistics New Zealand,16 as our denominator to calculate rates of disease. Annual estimated resident populations were only available from 1991 onward, so rates could not be calculated on data before this. Rates of hospital admissions and notifications and 95% confidence intervals (CIs) were calculated using the direct method of standardization17 (in 5-year age groups, confined to people younger than 15 years of age), and the 1976 European Standard Population, to enable direct comparison with UK data.12
We identified Maori children (the indigenous people of New Zealand) from the NMDS data by the prioritized ethnic code 21 and calculated hospital admission rates for Maori children younger than 15 years of age from this dataset using the Māori estimated resident population at June 30 of each year as our denominator. We were unable to establish reliable rates of disease for Pacific Island children because Statistics from New Zealand does not provide annual population estimates for this population. The small number of patients with unknown ethnicity (EpiSurv = 145, NMDS = 68) was excluded from Māori specific rates but included in total rates.
Unadjusted rates of IPD were also calculated in infants <1 year of age to analyze the burden of disease in infants and the impact of PCV in this age group, when compared with unadjusted rates for all children <15 years of age.
Statistical analysis was performed using R statistical software with the package epitools.18 This study was approved by the University of Otago Human Ethics Committee (Health) HD15/043.
Annual age-standardized hospitalization rates for Hib (any type) disease in New Zealand varied between 16.45 and 0.12 per 100,000 children from 1991 to 2014. From 1993, hospitalization rates declined by 84% within 2 years (from 13.53/100,000 in 1993 to 2.19/100,000 in 1995). Hospitalization rates were similar to notification rates for Hib (Fig. 1).
A reduction in hospitalization rates for meningitis and epiglottitis caused by Hib also followed the introduction of Hib vaccine (Fig. 1). During the 2 years 1993–1995, there was an 82% reduction in meningitis rates (8.63/100,000 in 1993 to 1.58/100,000 in 1995) and an 87% reduction for epiglottitis (3.67/100,000 in 1993 to 0.48/100,000 in 1995). Hospitalization rates and notification rates were similar for meningitis and epiglottitis. Hospitalization rates for Hib septicemia showed a similar trend.
New Zealand experienced a prolonged national epidemic of meningococcal disease beginning in the early 1990s. Between 1991 and 1997, hospitalization rates increased 14-fold from a baseline of 4.11 (95% CI, 2.83–5.78) to 55.86 per 100,000 children <15 years of age (95% CI, 50.88–61.20). After the introduction of MeNZB vaccine in 2004, rates declined by 73% within 3 years: from 36.68 (95% CI, 32.65–41.08) per 100,000 in 2003 to 10.05 (95% CI, 8.00–12.48) per 100,000 in 2006. After the MeNZB mass vaccination campaign ended in 2006, hospitalization rates continued to decline to 2.47 (95% CI, 1.53–3.78) per 100,000 in 2014 (Fig. 2). Hospitalization rates and notification rates for meningococcal meningitis and meningococcemia were similar during the epidemic.
Hospitalization rates for IPD in children from New Zealand <15 years of age ranged from a peak of 9.59 per 100,000 in 2001 to 2.17 per 100,000 in 2010. Within 2 years of the introduction of PCV7, rates decreased by 62%, from 7.80 per 100,000 in 2007 (95% CI, 6.01–9.95) to 2.98 (95% CI, 1.94–4.37) per 100,000 in 2009. Unlike for Hib and meningococcal disease, notification rates were 3 to 6 times higher than hospitalization rates (Fig. 3). Notification rates ranged from a peak of 17.99 (95% CI, 15.29–21.04) in 2009 (the first full year in which laboratory notification was mandatory) to 7.48 (95% CI, 5.79–9.53) in 2013. Compared with all children <15 years of age, unadjusted rates for IPD in those <1 year of age were much greater and experienced the greatest absolute decrease (Fig. 4).
Hospitalization rates for pneumococcal meningitis and septicemia showed similar trends to those for IPD. Hospitalization rates of pneumococcal meningitis ranged from a peak of 4.38 (95% CI, 3.03–6.12) in 2001 to 0.81 (95% CI, 0.32–1.67) in 2010. Both hospitalization and notification rates for pneumococcal meningitis were low. After the introduction of PCV7, hospitalization rates for pneumococcal septicemia declined by 45% from 2.51 (95% CI, 1.55–3.85) in 2007 to 0.81 (95% CI, 0.32–1.67) in 2010. From 2009 to 2014, notification rates were similar to hospitalization rates. Hospitalization rates for pneumococcal septicemia peaked in 2007 at 5.29 per 100,000 (95% CI, 3.84–7.11) and then reduced by 74% to 1.37 (95% CI, 0.71–2.39) in 2010 (Fig. 3).
Hospitalization rates for Māori children for Hib disease were similar to those of non-Māori children. Hospitalization rates for Māori children with meningococcal disease (all types) have historically been higher than those for the total population. However, during the epidemic, Māori children had substantially higher rates of hospitalization (Fig. 2). Absolute differences in incidence between Māori and non-Māori children have decreased since the vaccine was introduced and have almost disappeared. However, there continues to be a relatively higher rate for Maori children (rate ratio 2.16 in 2013). For IPD, Māori rates were generally higher than total rates between 2001 and 2013. The rate ratio between Māori and total rates of IPD has decreased, reducing from 1.7 times higher in 2006 (10.85 compared with 6.33) to 1.3 times higher in 2013 (3.43 compared with 2.61). For pneumococcal meningitis, Māori rates were 1.65 times higher in 2001 (7.26 compared with 4.40), whereas in 2013, they were lower than total rates (1.71 compared with 1.75; Fig. 3).
New Zealand hospitalization rates for invasive bacterial disease were generally higher than those reported in England (Table 1).12 Rates for meningococcal disease in New Zealand were almost double those of England from 1997 to 2006 but were 24% lower from 2007 to 2011. New Zealand rates of pneumococcal septicemia were 48%–100% higher than the English rates from 1997 to 2006, whereas rates of pneumococcal meningitis were similar in the 2 countries after 2001. The difference between rates of Hib meningitis and septicemia in New Zealand and England showed greater fluctuations. For example, rates of Hib septicemia were 40%–60% lower in New Zealand than in England between 1997 and 2006 and 70% higher between 2007 and 2011. This may be because rates of Hib disease in both countries were very low in these years.
The introduction of protein-conjugate vaccines has substantially reduced the incidence of invasive bacterial infections in New Zealand children. The absolute reductions in disease incidence have been greatest in Māori children but disparities still exist between Māori and non-Māori children.19
Consistent with several other countries,12,20–22 Hib disease almost disappeared after the introduction of Hib conjugate vaccine to New Zealand. This occurred despite low vaccination coverage at the time23 and demonstrates the effectiveness of Hib vaccine in reducing nasopharyngeal carriage of the organism.24
New Zealand’s meningococcal epidemic in the 1990s was caused predominantly by a single strain of Group B N. meningitidis.25 At the time, no meningococcal group B vaccines were available and a strain-specific outer membrane vesicle vaccine (MeNZB) was designed specifically for the epidemic. One study estimated the vaccine efficacy of the MeNZB vaccine to be 68% during the epidemic.14 Rates of disease fell rapidly after the introduction of MeNZB vaccine, although it is uncertain how much this was vaccine related or the natural course of the epidemic. The P1.7b4 strain remains the predominant strain causing invasive meningococcal disease in New Zealand, yet despite this, rates of disease in New Zealand have remained at low levels after withdrawal of the vaccine, similar to the rates before the meningococcal epidemic.1
One of the successes of New Zealand’s national immunization program has been its impact on ethnic disparities in the incidence of these diseases. Māori children in New Zealand have consistently higher hospitalization rates for infections, including gastroenteritis, skin infections, acute rheumatic fever and meningococcal disease, than non-Māori children.13 During the peak of the New Zealand meningococcal epidemic, rates of invasive disease for Māori, in particular meningitis, were more than 3 times those in non-Māori.26 During the epidemic, the coverage rates (those completing 3 doses) for Maori children <18 years of age were estimated at 78%, only slightly below non-Maori (83%).27 While rates in Māori children still remain higher than those in non-Māori, it appears that the MeNZB program played a part in reducing disparities in disease incidence, with the greatest reduction in rates of invasive meningococcal disease occurring in Māori.
For decades, the rates of IPD in Māori children had been proportionally higher than those for non-Māori, and these differences have persisted after the introduction of the vaccine. However, the greatest reductions in rates of IPD appear to have occurred in Māori between 2001 and 2013. In 2014, the absolute differences in rates of IPD, septicemia and pneumococcal meningitis between Māori and non-Māori, although still present, were no longer statistically significant.
The United Kingdom has a similar public health system to New Zealand and high-quality data collection systems for invasive bacterial infections. Their immunization program is similar to that in New Zealand, with the major difference being the inclusion of meningococcal vaccines (meningococcal C since 1999 and men B since 2016). Compared with England,12 New Zealand has historically had higher rates of invasive bacterial disease. In the period 1997 to 2001, the rates of meningococcal disease in New Zealand were almost double those in England, almost 1.5 times higher for IPD and 20% higher for Hib meningitis. Interestingly, the rates of meningococcal disease in New Zealand dropped well below those in England from 2007 to 2011 time, despite New Zealand not having a meningococcal vaccine on its immunization schedule. New Zealand has not introduced a meningococcal C vaccine because the predominant strains in the community continue to be group B, meaning the impact of a Men C vaccine is likely to be limited.28 The United Kingdom introduced the Men C vaccine in 1999, leading to a rapid decline in the rates of disease to very low levels.29
It is unclear why hospitalizations for IPD in New Zealand have been substantially higher than in England over several decades. Differences in the procedures for the coding of hospitalizations may be a contributing factor. This is supported by the rates of hospitalization for pneumococcal meningitis, which our study found are more closely correlated with notifications, being similar in both countries.
ESR data on serotype-specific rates of IPD supports the notion that the reductions in rates of IPD are predominantly as a result of vaccine introduction in New Zealand. Before the introduction of PCV7 in 2008, PCV7 serotypes made up 79.5% of all serotypes causing IPD in children <5 years of age. Three years after the vaccine introduction, not only had overall rates of disease fallen substantially, but also the proportion of PCV7 serotypes causing disease was 22%.
This is an ecologic study, and any associations between the introduction of vaccines and reductions in the rates of disease must be viewed with caution. Māori disease rates are likely to be underestimated as it is known that inaccurate coding of Māori ethnicity occurs in many health datasets. We have also assumed that accuracy of ethnicity coding has remained constant over time, and this may not be the case. Data for Pacific Island children were not examined, another very high-risk group for many different invasive infections.19 Such information is essential if we are to accurately track disparities in ethnic-specific disease rates.
The introduction of vaccines to prevent invasive meningococcal, IPD and Hib disease in New Zealand has had a significant impact on the rates of severe infection. This study suggests that the impact of these vaccines has been greatest in the Māori population, reducing longstanding disparities in health outcomes. New Zealand currently has no plans to introduce meningococcal vaccines into the national immunization schedule and in July 2017 changed from PCV13 back to using PCV10. Ongoing accurate surveillance of invasive bacterial diseases in children remains essential.
1. ESR Public Health Surveillance. surv.esr.cri.nz
. Available at: https://surv.esr.cri.nz/surveillance/Meningococcal_disease.php
. Accessed September 22, 2016.
2. Lennon D, Reid S, Stewart J, et al.Reducing inequalities with vaccines: New Zealand’s MeNZB vaccine initiative to control an epidemic. J Paediatr Child Health. 2012;48:193201.
3. Sáez-Llorens X, McCracken GH Jr.Bacterial meningitis in children. Lancet. 2003;361:21392148.
4. Gorelick MH, Baker MDEpiglottitis in children, 1979 through 1992. Effects of Haemophilus influenzae type b immunization. Arch Pediatr Adolesc Med. 1994;148:4750.
5. Shah RK, Roberson DW, Jones DTEpiglottitis in the Hemophilus influenzae type B vaccine era: changing trends. Laryngoscope 2004;114:557560.
6. Reid SThe further and future evolution of the New Zealand Immunisation Schedule. N Z Med J. 2012;125:8699.
8. Lennon D, Jackson C, Wong S, et al.Fast tracking the vaccine licensure process to control an epidemic of serogroup B meningococcal disease in New Zealand. Clin Infect Dis. 2009;49:597605.
9. Wilson N, Wenger J, Mansoor O, et al.The beneficial impact of Hib vaccine on disease rates in New Zealand children. N Z Med J. 2002;115:U122.
10. Loring BJ, Turner N, Petousis-Harris HMeNZB vaccine and epidemic control: when do you stop vaccinating? Vaccine. 2008;26:58995904.
11. Sivertsen LM, Pattemore PK, Abbot GDThe changing face of epiglottitis in Canterbury 1970-1996. N Z Med J. 1998;111:208210.
12. Martin NG, Sadarangani M, Pollard AJ, et al.Hospital admission rates for meningitis and septicaemia caused by Haemophilus influenzae, Neisseria meningitidis, and Streptococcus pneumoniae in children in England over five decades: a population-based observational study. Lancet Infect Dis. 2014;14:397405.
14. Arnold R, Galloway Y, McNicholas A, et al.Effectiveness of a vaccination programme for an epidemic of meningococcal B in New Zealand. Vaccine. 2011;29:71007106.
15. Harrison LH, Trotter CL, Ramsay MEGlobal epidemiology
of meningococcal disease. Vaccine. 2009;27(suppl 2):B51B63.
16. Standard for population terms. Available at: m.stats.govt.nz/browse_for_stats/population/standard-pop-terms.aspx
. Accessed November 10, 2017.
17. Rothman KJ, Greenland S, Lash TLModern Epidemiology
. 2008.Philadelphia, PA: Walters Kluwer;
18. Tomas AAragon Developer: Epitools. Available at: https://CRAN.R-project.org/package=epitools
. Accessed November 10, 2017.
19. Baker MG, Barnard LT, Kvalsvig A, et al.Increasing incidence of serious infectious diseases and inequalities in New Zealand: a national epidemiological study. Lancet. 2012;379:11121119.
20. Wang H, Deeks S, Glasswell A, et al.Trends in invasive Haemophilus influenzae type B disease in Australia, 1995-2005. Commun Dis Intell Q Rep. 2008;32:316325.
21. Peltola HWorldwide Haemophilus influenzae type b disease at the beginning of the 21st
century: global analysis of the disease burden 25 years after the use of the polysaccharide vaccine and a decade after the advent of conjugates. Clin Microbiol Rev. 2000;13:302317.
22. Heath PT, McVernon JThe UK Hib vaccine experience. Arch Dis Child. 2002;86:396399.
24. Mohle-Boetani JC, Ajello G, Breneman E, et al.Carriage of Haemophilus influenzae type b in children after widespread vaccination with conjugate Haemophilus influenzae type b vaccines. Pediatr Infect Dis J. 1993;12:589593.
25. Baker MG, Martin DR, Kieft CE, et al.A 10-year serogroup B meningococcal disease epidemic in New Zealand: descriptive epidemiology
, 1991-2000. J Paediatr Child Health. 2001;37:S13S19.
26. Wilson N, Baker M, Martin D, et al.Meningococcal disease epidemiology
and control in New Zealand. N Z Med J. 1995;108:437442.
28. PHARMAC Immunisation Subcommitte minutes. Available at: https://www.pharmac.govt.nz/about/committees/ptac/ptac-subcommittees/
. Accessed May 22, 2018.