Impact of meningococcal C conjugate vaccination programs with and without catch-up campaigns in adolescents: Lessons learned from Bahia, Brazil

ABSTRACT

The significant increase in the incidence rates and ongoing outbreaks of serogroup C meningococcal (MenC) disease, associated with the sequence type-103 complex, motivated the incorporation of the meningococcal C conjugate (MCC) vaccine in the routine immunization program in the State of Bahia, Brazil in early 2010, targeting children younger than 5 years of age. In its capital, Salvador, the program also included a catch-up campaign for individuals 10–24 years of age. We performed an observational, ecological study, analyzing data collected from 2007 to 2015, to compare the impact of these two immunization strategies on meningococcal disease incidence and mortality rates.

In Salvador, following the vaccination program, a dramatic early impact on MenC disease and mortality rates could be observed, with significant reductions in incidence rates of MenC disease in all age groups, including individuals that were too old to have been vaccinated, indicating the presence of herd protection. Compared to the pre-vaccine period, a virtual disappearance of MenC disease was observed in 2015. However, in the state of Bahia (excluding the city of Salvador), no herd protection could be observed, with significant impact only among vaccine-eligible children within 5 years of introduction of the MCC vaccination program.

These results highlight the importance of catch-up campaigns, including adolescents and young adults, to induce herd protection compared to immunization strategies restricted to infants and young children. This information is crucial for identifying optimal immunization policies and future strategies, focused on adolescents, to optimize the impact of MCC vaccination programs.

Introduction

Meningococcal disease (MD) is a leading cause of morbidity and mortality worldwide,¹ affecting 0.5–1.2 million people, each year. Even in treated patients, the case-fatality rate (CFR) of MD often exceeds 10%, with 50,000 to 135,000 deaths occurring annually.^2,3 In addition, a range of long term sequelae (including paralysis, hearing loss, mental impairment, amputations, and seizures) may affect up to 20% of survivors.⁴ Meningitis and septicemia are the most common clinical presentations of the disease.^3,5 Although MD affects individuals of all ages, the highest rates of disease are usually found among children aged <5 years, especially in infants. Peaks in incidence are also seen in adolescents as well as in elderly in some countries.^4-6 The incidence rates and serogroup distribution of MD varies by country and over time. However, 6 of the 12 recognized serogroups (A, B, C, W, X and Y) are known to cause virtually all cases of the disease worldwide.^4-6

Vaccination is considered the best control strategy for prevention of meningococcal disease. Even though meningococcal polysaccharide vaccines have played a pivotal role in disease prevention, controlling outbreaks and epidemics, they have several limitations. They do not generate an adequate immune response against serogroup C in young children, even in patients over 2 years of age the protection offered is of limited duration. Furthermore, they are capable of inducing hypo-responsiveness after subsequent doses. Meningococcal conjugate vaccines, in contrast, offer a number of advantages over polysaccharide formulations, inducing the production of high antibody levels, including in young infants, higher antibody avidity and increased serum bactericidal activity. They can also prevent acquisition of meningococci nasopharyngeal carriage among vaccinees, which proved to be of paramount importance in the success of the immunization programs with meningococcal C conjugate (MCC) vaccines.⁷ The United Kingdom was the first country to implement a national MCC immunization program in 1999, through routine infant immunization and a large catch-up campaign including all children, adolescents and young adults, resulting in significant and sustained disease reduction, with the induction of herd protection.^8,9 More recently, the introduction of MenAfriVac, a meningococcal group A conjugate vaccine, through mass immunization campaigns of 1–29-year-olds in the “meningitis belt” countries in Africa, also has had a dramatic effect on the incidence of meningitis cases.¹⁰

In Brazil, MD has been a major public health concern over time, providing opportunities for several vaccination programs, with important lessons learned from these experiences for an effective and successful control of the disease. In the 1970s, the spread of a serogroup A epidemic strain, with incidence rates as high as 180 cases/100,000 population in determined regions, provided the first major experience in the world with polysaccharide A and C vaccines on a large scale, resulting in the successful control of the epidemic. During the 1980s and 1990s, serogroup B meningococcal (MenB) disease became prevalent in Brazil, with practically no cases of serogroup A reported. Epidemics attributed to MenB in several locations around the country were observed, motivating a reactive vaccination campaign in the early 1990s with the Cuban serogroup B outer membrane vesicle vaccine targeting almost 3 million children in the State of Sao Paulo, with limited impact, if any, on the disease rates and a lack of effectiveness in young children.^11,12 From 2002 onward, a significant rise in the number and proportion of cases due to serogroup C, associated with the ST-103 complex, was registered, with several outbreaks affecting different regions across the country.^11,12

During 2007–2009, the incidence rate of MenC disease increased substantially in the state of Bahia. In response, the state government initiated in 2010 an immunization campaign with the meningococcal C conjugate (MCC) vaccine, with a two-dose primary immunization series (at 2 and 4 months) followed by a booster dose at 12–15 months. Children between 1 and 5 years of age were eligible to receive one dose of MCC vaccine. In the capital, Salvador, a different strategy was implemented extending the vaccination to the population aged 10–24 years-old.¹³ Furthermore, in late 2010, the Brazilian Ministry of Health introduced the MCC vaccine for children <2 years-old into the routine schedule of the national immunization program.^12-14 The MCC schedule for infants included two doses at either 2 and 4 months or 3 and 5 months followed by a booster at 12–15 months of age. Unlike Salvador, the MCC vaccination program in the rest of the country was implemented without a catch-up campaign.^12-14

The provision of a surveillance system to estimate disease burden of MD plays a pivotal role in the development and implementation of the disease control and prevention strategies by monitoring changes in the epidemiology of MD across the time, such as disease rates, mortality rates, serogroup replacement, and detecting outbreaks. The aims of the present study were to analyze and compare the effects of two different vaccination strategies against MenC disease in the epidemiology of MD, one in the State of Bahia (excluding its capital, Salvador), that targeted only children younger than 5 years of age and the other in the city of Salvador, the only place in Brazil where the MCC vaccination program was implemented with an extended catch-up campaign, including adolescents and young adults from 10–24 years.

Results

From 2007 to 2015, a total of 1,264 cases of MD and 325 deaths were reported in the state of Bahia, of which 517 cases of MD and 128 deaths were reported in the city of Salvador. Males accounted for 58.1% of all cases and 57.5% of deaths.

Serogroup information was available for 715 cases of MD (56%). In the State of Bahia, a lower proportion of confirmed cases with serogroup information was observed compared to Salvador (47.5% vs 69.1%). Among the 715 cases of MD in the State of Bahia with serogroup information, 320 (44.7%) were from Salvador and 395 (55.3%) from the State of Bahia (without Salvador), with a clear predominance of serogroup C, responsible for 85.3% and 89.6% of the identified cases, respectively, in the city of Salvador and in the State of Bahia (without Salvador), followed by serogroup B (9.5% and 7.1%) and serogroup W (5.2% and 3.3%). No trends in the proportion of confirmed cases without serogroup information was observed over time during the study period.

In Salvador, a rapid and significant increase in MenC disease incidence rates was observed during the pre-vaccination period, from 0.49/100,000 in 2007 to a peak of 4.0/100,000 population in 2010 (Fig. 1). The highest incidence rates of MenC disease in the pre-vaccine period were observed in the 0–4 years age group (5.45/100,000 population), followed by the 5–9 years (4.06/100,000 population), 10–19 years (2.94/100,000 population), 20–39 years (1.92/100,000 population) and individuals > 40 years (0.96/100,000 population) (Table 1). In the State of Bahia (without Salvador) incidence rates of MenC disease also increased significantly, from 0.16/100,000 population in 2007 to 0.54/100,000 population in 2010. A similar pattern was observed in the age group distribution, with the highest incidence rates in the 0–4 years age group and a decreasing trend toward older age groups (Table 1).

Figure 1.

Incidence rates of serogroup C meningococcal disease, by age group, in the city of Salvador, 2007 – 2015.

Table 1.

Incidence rates of MenC disease by age group in the city of Salvador and in the state of Bahia (without Salvador), Brazil, in pre- and post-vaccine periods.

	Salvador					Bahia (without Salvador)
Age group (years)	Pre-vaccine period Rateα	Post-vaccine period Rateα	RR*	(95% CI)†	p value	Pre-vaccine period Rateα	Post-vaccine period Rateα	RR*	(95% CI)†	p value
0-4	5.45	0.73	0.11	(0.04-0.28)	<0.0001	0.78	0.24	0.30	(0.15-0.63)	0.001
5-9	3.91	1.11	0.28	(0.13-0.58)	0.006	0.48	0.27	0.57	(0.21-1.13)	0.116
10-19	3.05	0.70	0.23	(0.12-0.40)	0.006	0.55	0.42	0.76	(0.51-1.14)	0.203
20-39	1.92	0.78	0.39	(0.27-0.58)	0.001	0.23	0.19	0.84	(0.53-1.34)	0.480
≥ 40	1.01	0.45	0.44	(0.25-0.76)	0.004	0.15	0.11	0.76	(0.41-1.40)	0.389
Total	2.13	0.67	0.31	(0.24-0.39)	<0.0001	0.33	0.22	0.67	(0.53-0.85)	0.001

^*Relative risk;

^†Confidence interval;

^αAverage annual incidence rate per 100,000 population.

After the introduction of MCC vaccination in 2010, a significant decrease in incidence rates of MenC disease was observed in the city of Salvador. Comparing the post-vaccination period with the pre-vaccination period we observed a significant reduction of 69% (p < 0.0001) in overall MenC disease rates in Salvador (Table 1). In the first year after the MCC vaccination campaign, a 58% reduction in the incidence rates of MenC disease was already observed, decreasing from 4.0/100,000 population in 2010 to 1.67/100,000 population in 2011 (Fig. 1), followed by a reduction of 75% (1.0/100,000 population) in 2012, 86.2% (0.55/100,000 population) in 2013, 92.9% (0.14/100,000 population) in 2014 and 98% (0.03/100,000 population) in 2015. Table 1 clearly shows that reductions in incidence rates observed in the City of Salvador were significant in all age groups, including individuals aged ≥ 40 years, an age group that was not eligible to receive the vaccine. A significant reduction in MenC mortality rates of 57% (p = 0.005), from 0.46/100,000 population in the pre-vaccine period to 0.20/100,0000 population in the post-vaccine period, was also observed in Salvador (Table 2).

Table 2.

Mortality rates of Meningococcal disease by serogroup in the city of Salvador and in the state of Bahia (without Salvador), Brazil, in pre- and post-vaccine periods.

	Meningococcal disease
	Salvador					Bahia (without Salvador)
	Deaths	Mortalityπ	RR*	95% CI†	p-value	Deaths	Mortalityπ	RR*	95% CI†	p-value
All cases
Pre-vaccine periodα	83	0.78	ref			109	0.25	ref
Post-vaccine periodβ	45	0.34	0.43	(0.30-0.62)	<0.0001	88	0.16	0.64	(0.48-0.85)	0.002
MenC
Pre-vaccine period	50	0.46	ref			27	0.06	ref
Post-vaccine period	27	0.20	0.43	(0.27-0.68)	0.005	33	0.06	0.97	(0.58-1.62)	0.9351
MenW
Pre-vaccine period	0	0.00	ref			1	0.002	ref
Post-vaccine period	6	0.04	—	—	0.25	6	0.01	4.8	(3.10-7.71)	<0.0001
MenB
Pre-vaccine period	4	0.04	ref			1	0.002	ref
Post-vaccine period	2	0.00	0.00	(0.00-39.00)	0.25	4	0.007	3.2	(2.02-5.23)	<0.0001
Unidentified Serogroups
Pre-vaccine period	29	0.26	ref			80	0.18	ref
Post-vaccine period	10	0.07	0.29	(0.29-1.50)	0.050	44	0.08	0.45	(0.18-1.03)	0.021

^*Relative risk;

^†Confidence interval;

^α2007-2010;

^β2011-2015;

^πAverage annual mortality rate per 100,000 population.

The analysis of the impact on MenC disease rates in the State of Bahia (excluding the city of Salvador), where the vaccination program only targeted children < 5 years, showed that in the post-vaccination period a significant reduction of 33% (p = 0.001) in overall MenC disease rates was observed (Table 1). In the first year after the MCC vaccination campaign, a reduction of only 9% in the incidence of MenC disease was observed, decreasing from 0.55/100,000 population in 2010 to 0.50/100,000 population in 2011 (Fig. 2), followed by a reduction of 58% (0.23/100,000 population) in 2012, 65.4% (0.19/100,000 population) in 2013, 65.4% (0.19/100,000 population) in 2014 and 88% (0.06/100,000 population) in 2015. However, performing the analysis stratifying the impact on the different age groups, we could observe that reductions of MenC disease rates in the post-vaccination period reached statistical significance only for children 0–4 years, the age group that was targeted by the vaccination program, with no significant impact noted in other age groups (Table 1). Mortality rates of MenC disease in the post-vaccine period were identical to the pre-vaccine period in the State of Bahia (without Salvador) (Table 2).

Figure 2.

Incidence rates of serogroup C meningococcal disease, by age group, in the state of Bahia (without Salvador), 2007–2015.

According to the descriptive data, no significant difference was observed in MenB disease rates in Salvador and also in the State of Bahia (without Salvador) in the post-vaccination period compared to the pre-vaccination period, although a decreasing incidence trend was observed (Table 3). Few cases of MenW were identified from 2011 in both regions, with increased incidence rates during the post-vaccine period in the state of Bahia (without Salvador).

Table 3.

Incidence rates of meningococcal disease by serogroup in the city of Salvador and in the state of Bahia (without Salvador), Brazil, in pre- and post-vaccine periods, 2007–2015.

	Salvador					Bahia (without Salvador)
Age group (years)	Pre-vaccine period Rateα	Post-vaccine period Rateα	RR*	(95% CI)†	p value	Pre-vaccine period Rateα	Post-vaccine period Rateα	RR*	(95% CI)†	p value
0-4	5.45	0.73	0.11	(0.04-0.28)	<0.0001	0.78	0.24	0.30	(0.15-0.63)	0.001
5-9	3.91	1.11	0.28	(0.13-0.58)	0.006	0.48	0.27	0.57	(0.21-1.13)	0.116
10-19	3.05	0.70	0.23	(0.12-0.40)	0.006	0.55	0.42	0.76	(0.51-1.14)	0.203
20-39	1.92	0.78	0.39	(0.27-0.58)	0.001	0.23	0.19	0.84	(0.53-1.34)	0.480
≥ 40	1.01	0.45	0.44	(0.25-0.76)	0.004	0.15	0.11	0.76	(0.41-1.40)	0.389

^*Relative risk;

^†Confidence interval;

^α2007-2010;

^β2011-2015;

^πAverage annual incidence rate per 100,000 population.

Discussion

Two different vaccination strategies were implemented to control MD in the State of Bahia, one is its capital, Salvador, with routine immunization of infants and a “one-time” catch-up campaign including children aged < 5 years and individuals 10–24 years of age and the other, in the rest of the state, with routine immunization of infants and a limited catch-up campaign only including children aged < 5 years. The significant number of cases associated with high case fatality rates reported among adolescents and young adults in the City of Salvador provided the rational for the decision to also include these age groups in the vaccination strategy implemented. These different approaches provided a unique opportunity to compare the effects of these vaccination strategies on MD burden.

Our results showed that in Salvador, after the vaccination program, a dramatic early impact on MenC disease and mortality rates could be observed, with significant reductions in incidence rates of MenC disease observed in all age groups, including individuals that were too old to have been vaccinated, indicating the presence of herd protection. Decreases in incidence rates were already observed in the first year after the implementation of the vaccination program, when a 58% reduction was demonstrated in overall MenC rates and were sustained for at least 5 years after implementation. Compared to the pre-vaccine period, a virtual disappearance of Men C disease was observed in 2015.

Overall, the effects of the vaccination program in Salvador were similar to those observed in UK and several other European countries, Australia and Canada, that also introduced MCC vaccination with catch-up campaigns targeting adolescents and young adults.^15-19 The success of the MCC vaccination program in these countries was attributed to the combined efficacy of the vaccine not only preventing disease but also preventing acquisition of carriage.^8,20 A striking feature of these MCC vaccination programs, which included catch-up campaigns achieving high coverage among adolescents and young adults, has been the additional decrease in disease incidence in unvaccinated individuals as a result of herd protection.²¹ Immunization of adolescents and young adults (the age groups usually responsible for the highest rates of colonization) in these catch-up campaigns prevents the acquisition of MenC carriage in the vaccinated age group, interrupting transmission of the organism in the community.^8,20,21

In Salvador, an 80% coverage was achieved among young adolescents 10–14 years and a 67% coverage among older adolescents 15–19 years. The importance of achieving high coverage rates among adolescents and young adults in catch-up campaigns to obtain herd effects is highlighted by the recent example of France, where MCC vaccination coverage rates in these age groups were too low, limiting the impact on disease rates in unvaccinated age groups.²²

Comparing to Salvador, the effects of the vaccination program in the state of Bahia (excluding Salvador), however, were limited. Despite the dramatic decrease in the incidence rates of MD among children < 5 years, the age group that was vaccinated, no significant impact was observed in other age groups within 5 years of introduction of the MCC vaccination program. In Spain, where the MCC vaccination program initially targeted only children up to the age of six years, without immunizing adolescents and young adults, herd effects were also limited, with no apparent impact on unvaccinated children.¹⁸ Interestingly, other studies that evaluated the impact of the MCC vaccination program in Brazil, where vaccination was restricted to infants and toddlers, without catch-up in adolescents and young adults, found similar results to those observed in the state of Bahia, excluding the city of Salvador. In a preliminary analysis of the early impact of the routine infant MCC vaccination program in Brazil, the authors could observe significant decreases in incidence rates only among age groups that were vaccinated.²³ More recently, other studies that analyzed the effects of the MCC vaccination after three²⁴ and four²⁵ years of its introduction into the Brazilian NIP, also concluded that the impact on disease rates in Brazil could be observed only among vaccine-eligible age groups.

Moving forward with the MCC vaccination program in Brazil, to achieve the same impact that was observed in the city of Salvador, a key strategy to maximize control of MD and further reduce serogroup C disease rates and interrupt meningococcal transmission, beyond the levels observed so far, would be to induce herd protection in populations where it is currently lacking.

Despite the lack of reliable baseline data collected prior to vaccine introduction in Brazil, a study performed 4 years after the implementation of the MCC vaccination program in Salvador showed that MenC carriage rates among a representative sample of adolescents were very low.²⁶ However, in a similar study performed in adolescents during the post-vaccination period in the state of Sao Paulo, where the immunization program was restricted to infants and toddlers, the authors found an unusually high dominance of serogroup C, associated with the virulent clone belonging to the ST-103 complex, almost eight times higher than the prevalence rate of MenC carriage found in Salvador.²⁷ Interestingly, comparing to data from European and North-American countries, where carriage is most frequent in young adults,²⁸ these trials in Brazil demonstrated that carriage rates were highest at younger ages.

Taking all this evidence into account the Ministry of Health in Brazil decided to change the National Immunization Program, including a limited phased catch-up campaign with the MCC vaccine during adolescence.²⁹ In the next four years, two cohorts will be vaccinated each year, starting in 2017 with adolescents 12–13 years of age, followed by 11–12 years of age in 2018, 10–11 years of age in 2019 and 9–10 years of age in 2020. After this phased catch-up campaign, a booster dose of MCC vaccine will be routinely recommended at 9 years of age together with the HPV vaccine, to optimize the coverage of these two vaccines. The routine adolescent vaccination, inducing high levels of serum bactericidal antibodies, will be of paramount importance not only to provide direct protection against disease, since there is evidence suggesting that infant/toddler vaccination is not sufficient for long-term protection,^29-30 but also to maintain the herd protection that is expected to occur after the phased catch-up campaign.^31-33

This is an observational, ecological study, and thus subject to potential limitations. First, the burden of the disease may have been underestimated since we are analyzing secondary data that are not updated and underreporting is a common event. Second, rates were analyzed without redistribution of cases of unidentified serogroups; therefore serogroup-specific incidence was probably underestimated as well. A further limitation is that we may have different characteristics, in terms of density and carriage rates, in the two populations (from Salvador, an urban metropolis; and in the rest of the State, where urban and rural areas are present), that might have influenced the evaluated outcomes. However, all these potential limitations should not detract from the demonstration of the importance of catch-up campaigns, as already shown with meningococcal vaccination programs elsewhere.^9-10

Finally, although there is no clear evidence to date that MCC vaccination may lead to vaccine-driven serogroup replacement,^34-36 continuous surveillance is mandatory after implementation of mass immunization programs. Our results showed that MenB disease rates remained stable during the study period, without significant differences between pre- and post-vaccine periods, which is in line with the experience reported in other countries that introduced the MCC vaccine in mass immunization programs.^9,37–41 Although limited to few cases, we observed a rise in the incidence and mortality rates of serogroup W disease in the post-vaccination period. The emergence of the MenW hypervirulent ST-11 clonal complex, which is associated with high case-fatality rates and a propensity to cause outbreaks and epidemics, has also been observed in the southern region in Brazil, Argentina, Chile, Australia, UK as well as other European countries^42-48 requiring close monitoring and continuous surveillance.

In summary, the evidence gathered during this study provides robust evidence on the importance of achieving herd protection with MCC vaccination programs in order to substantially reduce morbidity and mortality rates due to serogroup C MD. It is important to emphasize that this is the first evidence in the literature showing the ability of MCC vaccines to provide herd protection in a place with increased incidence of MenC disease associated with the spread of the hypervirulent clone belonging to the ST-103 complex, highly unrelated to ST-11 complex strains. This information is vital for a better understanding of the MD epidemiology in Brazil and to anticipate the potential impact of different immunization strategies with MCC vaccines on disease burden. Continuous surveillance is required to provide information for evaluating meningococcal conjugate immunization schedules and to guide future vaccination strategies.

Patients and methods

Study design and population

This is an observational, ecological study conducted in the state of Bahia, located in the northeast region of Brazil. We investigated the incidence and distribution patterns of the confirmed cases of MD occurred before (2007-2010) and after (2011-2015) the immunization campaign with MCC vaccine.

The MCC vaccination was implemented in the routine infant immunization program in the State of Bahia (14,016,906 inhabitants in 2010) in February 2010. The recommended schedule was 2 doses, at 3 and 5 months of age, followed by a booster dose at 12–15 months. A limited catch-up vaccination was performed, only including children between one and five years, that received one dose of the vaccine.

In the capital of the State, Salvador (population of 2,675,656 inhabitants in 2010), the vaccination program also targeted children aged < 5years, but included an extended catch-up, targeting individuals from 10–24 years of age performed in three phases, between May and August 2010. Individuals from 10–14 years were the first group vaccinated, followed by the group from 15–19 years and the group from 20–24 years. Children from 5–9 years were not vaccinated.

Both MenC-TT (Neisvac-C®, Baxter Vaccines) and MenC-CRM (Menjugate®, Novartis Vaccines) vaccines were used during the campaign. Initially, in February 2010, MenC-TT was the only vaccine used in the State of Bahia, replaced by the use of MenC-CRM vaccine, once MCC vaccination was implemented in the whole country, from August 2010.

Coverage rates were calculated using the information system of the National Immunization Program (PNI, http://tabnet.datasus.gov.br/cgi/deftohtm.exe?pni/cnv/cpniuf.def). Coverage rates with at least one dose of the vaccine in 2011, the first year after the implementation of the vaccination program, for children < 5 years of age were estimated at 93% both in Salvador and in the State of Bahia. For the age groups included in the catch-up campaign in Salvador, coverage reached 80% for the group from 10–14 years, 67% for the group 15–19 years and 40% for the group 20–24 years at the end of 2010.

Data source

Information was based on the national Notifiable Diseases Information System (SINAN) sponsored by the Informatics Department of the National Health System (DATASUS), Health Ministry, Brazil.⁴⁹

In Brazil, the notification of cases of MD is mandatory. A confirmed case is defined by the presence of at least one of the following: isolation of Neisseria meningitidis from a normally sterile site, detection of bacterial DNA by polymerase chain reaction (PCR), antigen detection, clinical-epidemiological criteria (case of close contact with a laboratory confirmed case), gram-staining, or clinical criteria (patient with suggestive symptoms and petechial or purpuric rash).⁵⁰ In this study, all confirmed cases for meningococcal meningitis, meningococcemia and/or meningococcal meningitis plus meningococcemia were considered.

Data analysis

The incidence of MD was assessed year by year, comparing aggregated data for the pre- and post-vaccination period according to sex, age group (0-4, 5–9, 10–19, 20–39 and ≥40 years) and serogroup of Neisseria meningitidis. The incidence and mortality rates were calculated by serogroup and time period. Rates were not adjusted for the proportion of confirmed MD cases with unidentified serogroup. The annual number of deaths of MD was assessed for 2007–2015 and analyzed by serogroup and sex. Death was recorded as the outcome reported at the DATASUS database. We obtained population estimates for the city of Salvador and for the State of Bahia from the Brazilian Institute of Geography and Statistics (IBGE) and used the 2010 national census.⁵¹

For the morbidity and mortality rates, we calculated serogroup-specific relative risk (RR) to compare pre-vaccine to post-vaccine periods with the respective confidence intervals (95% CI). The significance level was set at p≤0.05. Statistical analysis was performed using STATA version 12 (STATA Corp., Texas, USA).

Disclosure of potential conflicts of interest

M.A.P.S. has received grants to support research projects and consultancy fee from GSK, Pfizer and Sanofi Pasteur; the other authors have nothing to disclose.