Haemophilus influenzae Type b Disease among Children in Rural Mozambique: Impact of Vaccine Introduction

Abstract

Objective

Haemophilus influenzae type b (Hib) conjugate vaccine has dramatically reduced invasive Hib disease worldwide. Yet, data on protection against pneumonia and among children with HIV are limited. We evaluated the impact of Hib conjugate vaccine introduction in 2009 in a rural, high-HIV prevalence area in Mozambique.

Study design

From 2006-2011, we conducted hospital-based surveillance for invasive Hib disease and clinical pneumonia (classified as severe and very severe) among children <5 years of age. Incidences calculated using population denominators were compared between baseline (2006-2008) and post-Hib conjugate vaccine (2010-2011) periods. Surveillance data for radiologically-confirmed pneumonia among children <2 years of age in 2011 were compared with baseline data from 2004-2006.

Results

Among 50 cases of invasive Hib disease, 5 occurred after Hib conjugate vaccine introduction; 1 case-patient was age-eligible for Hib conjugate vaccine (and had received 3 doses). Four post-Hib conjugate vaccine case-patients (including Hib conjugate vaccine failure) had HIV. Among children <1 and <5 years of age, significant reductions occurred in rates of invasive Hib disease (91% and 85%, respectively) and very severe pneumonia (29% and 34%, respectively). Radiologically-confirmed pneumonia incidence fell significantly (33%) in children <2 years of age. Severe pneumonia incidence did not decline.

Conclusions

We demonstrate important reductions in invasive disease and pneumonia following Hib conjugate vaccine introduction in a high-HIV area. Continued surveillance is needed to monitor long-term Hib conjugate vaccine effects, particularly among children with HIV.

A highly effective vaccine against Haemophilus influenzae type b (Hib) has been available since the 1990s and has led to dramatic declines in Hib disease in settings where it has been incorporated into the routine infant immunization program. Although uptake of the Hib conjugate vaccine was initially concentrated in high-income countries,¹ in 2006 the World Health Organization recommended that Hib conjugate vaccine be included in the routine immunization schedule in all countries.² Since that time, increasing numbers of middle- and low-income countries have introduced Hib conjugate vaccine, and several studies have demonstrated the impact and effectiveness of the vaccine against Hib disease in those settings.^3–12

The incidence of the most severe manifestations of Hib disease (meningitis and bacteremia/sepsis) has been reduced to extremely low levels in a variety of resource-poor settings.³ However, the impact of Hib conjugate vaccine on nonbacteremic pneumonia has been more challenging to measure. Although extremely important from a policy perspective because of the enormous disease burden, pneumonia is a nonspecific endpoint that poses methodological challenges for measuring the impact of vaccine introduction. In addition, questions remain about the impact and effectiveness in populations with a high HIV prevalence. The burden of Hib bacteremic pneumonia has been found to be more than 20 times higher among children with HIV compared with those without HIV.¹³ Immunogenicity studies have found lower immune responses to Hib conjugate vaccine among children with HIV.¹⁴ Limited evidence from low- and middle-income countries with high HIV prevalence suggests that Hib conjugate vaccine does lead to an important reduction in Hib disease when used in the routine infant immunization program; however, the effectiveness appears to be lower in children with HIV compared with those without HIV.¹⁴

Manhiça District is located in southern Mozambique and is a rural area with high prevalence of HIV of 39.9% among adult population between 18 and 49 years of age¹⁵; the HIV prevalence among children has not been systematically measured. Since 2001, population-based surveillance for invasive bacterial disease and pediatric morbidity (including pneumonia) has been conducted by the Centro de Investigacão em Saúde de Manhiça (CISM), a Health Research Center located in Manhiça. Hib conjugate vaccine was introduced in the routine Expanded Program on Immunization (EPI) 2009. The study was initially planned to evaluate the vaccine effectiveness and impact by doing a case-control study after vaccine introduction as well as disease surveillance. The effectiveness study was not feasible due to sample size issues. Here, we analyzed data from the surveillance system to evaluate the impact of Hib conjugate vaccine against invasive Hib disease and pneumonia among children <5 years old.

Methods

Since 1996, CISM has continuously maintained a demographic surveillance system (DSS) in a well-defined geographical area in Manhiça District, southern Mozambique. In 2009, the total population in the entire District was 167 000 inhabitants of whom 84 000 (50%) resided in the DSS area. Within the DSS area, 17% of the population is <5 years of age. In addition to the DSS, a passive surveillance system for pediatric morbidity was established at the Manhiça District Hospital and 5 surrounding health centers within the DSS area. Each person residing in the DSS area has a unique identification number, which permits linking of demographic and morbidity data. Clinical and laboratory data are routinely collected by the research team using standardized forms.

This study was approved by the Mozambique Ministry of Health Ethical Review Committee and the Institutional Review Board of the Clinic Hospital of Barcelona; it was considered to be a public health evaluation and not human subjects research by the US Centers for Disease Control and Prevention.

Hib Conjugate Vaccine Introduction and Coverage

Hib conjugate vaccine was introduced into the EPI with support from the Global Alliance for Vaccine and Immunization in the form of a pentavalent formulation (diphtheria/tetanus/whole-cell-pertussis [DTP]/hepatitis B [Hep B] virus vaccine /tetanus-toxoid conjugate Hib vaccine). Starting in early 2009, the vaccine was distributed progressively from north to south of the country to use the old existing stocks of DTP/Hep B vaccine and reached Manhiça district in August 2009. According to the national EPI schedule, pentavalent vaccine was targeted at children aged 2, 3, and 4 months of age. Children who had not completed the primary series of DTP/Hep B vaccine received the remaining doses using the pentavalent vaccine. There was no catch-up campaign.

In Manhiça District, coverage for the third dose of pentavalent in infants in 2010 was 79%; countrywide coverage was 60% for the same period (unpublished data, Ministry of Health). Of note, in April-May 2010, there was an interruption in the pentavalent vaccine supply that affected coverage nationwide. During the study period, no pneumococcal vaccine was provided by the EPI.

Surveillance Procedures

Invasive Hib Disease Surveillance.

Surveillance for invasive bacterial disease has been on-going at Manhiça District Hospital since 2001. Surveillance was strengthened starting in 2006 with more precise case definitions for suspect meningitis and criteria for lumbar puncture together with routine use of latex agglutination for testing cerebrospinal fluid (CSF).^16,17 All pediatric patients admitted into the hospital except for those with elective admission or trauma had a blood sample collected for culture. CSF samples are collected from suspected cases of meningitis. Blood and CSF cultures are performed using conventional methods, as described elsewhere.^17,18

Bacterial isolates were identified by colony morphologic analysis and growth requirements. Haemophilus species were identified by colony morphology, Gram stain, X and V factor dependence, and all H influenzae were serotyped with a commercial slide agglutination test. Additionally, to enhance detection of Hib in CSF samples, a real time polymerase chain reaction (PCR) targeting bcsB gene¹⁹ was performed retrospectively for all CSF samples during the study period. Internal quality control for all microbiology procedures was performed weekly and external quality control for microbiological laboratory standards on the Hib and pneumococcal isolates were provided by US Centers for Disease Control and Prevention, Atlanta, Georgia.

An episode of invasive Hib disease was defined as isolation of Hib from blood or CSF samples in an admitted child, positive Hib in the latex agglutination, or PCR performed on those samples. Cases with Hib detected in CSF were classified as Hib meningitis.

Pneumonia Surveillance.

Ongoing morbidity surveillance at Manhiça District Hospital routinely captures clinical and radiologic data on hospitalizations for clinical severe pneumonia and has been ongoing since 2004. Cases of clinical pneumonia are defined as a history of cough or difficulty breathing and increased respiratory rate according to the age group.²⁰ Clinical severe pneumonia is defined by the presence of chest wall indrawing in a child with clinical pneumonia. Very severe pneumonia includes the presence of any danger sign (inability to drink or feed convulsions, lethargy, and unconsciousness) in a child with severe clinical pneumonia.

In addition, starting in January 2011, chest radiographs (CXR) have been performed for all children <2 years old admitted with severe or very severe pneumonia. Digital CXR images (scanned films or obtained digital CXR) are routinely interpreted by 2 study readers according to World Health Organization guidelines for standardized interpretation.²¹ Episodes with evidence of consolidation or pleural effusion are defined as radiologically-confirmed pneumonia. Baseline (pre-Hib conjugate vaccine) data for radiologically-confirmed severe pneumonia were available for comparison from studies conducted in 2004-2006 that used the same methodology for identifying and diagnosing cases of radiologically-confirmed pneumonia.²²

HIV Testing.

HIV-1 testing was not routinely performed in the early years of surveillance among the Hib cases, starting in January 2010, HIV status was determined for children with confirmed invasive Hib disease (in blood or CSF) or radiologically confirmed pneumonia. For HIV testing, rapid tests (Determined and confirmed by Unigold [Trinity Biotech, Bray, Ireland]) were used to detect HIV antibodies for children ≥18 months of age and blood spots collected to performed DNA- PCR for HIV-1 among children <18 months of age.

Vaccination Status.

Vaccination status and dates of administration of vaccines were obtained from vaccination cards and recorded in a standardized form for invasive Hib disease cases in the postvaccination period.

Data Management and Statistical Analyses

Standardized forms were double entered in FoxPro (v. 2.6, Microsoft Corporation, Redmond, Washington) at CISM, and discrepancies in data entry were resolved by referring to the original forms.

We calculated the incidence rates for episodes of invasive Hib disease, Hib meningitis, clinical severe pneumonia, and very severe pneumonia among admitted children <1 and <5 years of age from the DSS area; the incidence of radiologically-confirmed severe pneumonia was calculated for children <1 and <2 years of age (data on older children were not available) and compared the pre- and post-Hib conjugate vaccine rates. To estimate the incidences, person-time of follow-up for children in the demographic surveillance area was calculated using dates of birth and death, excluding periods of outmigration. Children were considered not exposed during the 15 days after each episode (an arbitrary lag period). Recurrences observed during the lag period were considered the same episode.

To evaluate the impact of Hib conjugate vaccine on the incidence rates of the outcomes mentioned above, the prevaccine baseline period was defined as January 1, 2006-December 31, 2008, and the postvaccine period as January 1, 2010-December 31, 2011. The year of Hib conjugate vaccine introduction (2009) was not included in the comparative analysis of 2 pre-/postintroduction periods. For radiologically-confirmed pneumonia analysis, we designated March 18, 2004 to March 17, 2006 the pre-vaccine period and January 1-December 31, 2012 the postintroduction period. The year of Hib conjugate vaccine introduction (2009), was not included in the comparative analysis of 2 pre-/postintroduction periods.

Negative binomial regression models were estimated to compare incidence rates between age-groups or pre-/postvaccine introduction. Models were estimated with random intercept to take into account repeated measures because children can belong to several age categories or to several calendar years (ie, pre- and postvaccine introduction) during the follow-up. Overall P values for age were calculated using the Likelihoodratio test.

Statistical analysis was performed using STATA software (v. 12.0, STATA Corporation, College Station, Texas).

Results

From January 1, 2006 to December 31, 2011, 8067 children <5 years of age from the DSS area were admitted to the hospital. Children <1 year of age accounted for 35% (2837 of 8067) of all admissions among those <5 years of age. Blood and CSF samples were collected for bacterial culture from 6800 (84%) and 1098 (14%) of admitted children, respectively. A total of 50 cases of laboratory-confirmed invasive Hib were detected among children <5 years of age. There were 31 cases of bacteremia and 19 cases of meningitis (including 11 cases of meningitis that also had positive blood cultures).

Among the 50 episodes of invasive Hib disease, 45 (90%) cases occurred before Hib conjugate vaccine introduction. Prior to Hib conjugate vaccine, cases of invasive Hib disease occurred in children from 4-52 months of age, with a median age of 13.6 months. HIV testing results were not available for any case prior to Hib conjugate vaccine introduction. Following Hib conjugate vaccine introduction, only 5 cases of invasive Hib disease were detected. All 5 cases presented as bacteremia and none were meningitis. The 5 children with invasive Hib disease ranged from 10-57 months of age, with a median of 28 months. Four of the five children (80%) were too old to have received any dose of Hib conjugate vaccine and 3 out of 4 had HIV. The youngest child had received 3 doses of Hib conjugate vaccine and had HIV. The case fatality ratio for children hospitalized with invasive Hib disease in the pre-Hib conjugate vaccine period was 13.2% (5 of 38) compared with 0% (0 of 5) during the 2 years post-Hib conjugate vaccine period.

Figure 1 shows the trends of annual incidence rates of invasive Hib among children <1 year and <5 years of age from 2006-2011. Declines were observed among children in both age groups over the course of the study period; however, the most marked reduction was among <1-year-olds following Hib conjugate vaccine introduction. Figure 2 summarizes the trends of annual incidence rates of severe pneumonia among children <1 year and <5 years of age from 2006-2011. Very severe pneumonia decreased in both age groups (from 10.7 per 100 000 person-years at risk in 2008 to 8.1 in 2011 among children <1 year of age and from 6.1 per 100 000 person-year at risk in 2008 to 4.1 in 2011 among children <5 years of age) after Hib vaccine introduction. However, severe pneumonia did not decrease during the same period.

Figure 1.

Trends of incidence rate of invasive Hib disease by different age groups before and after Hib conjugate vaccine introduction (introduced in August 2009), in Manhiça District, Southern Mozambique. HibCV, Haemophilus influenzae type b conjugate vaccine.

Figure 2.

Incidence rate of severe and very severe pneumonia by year among children <1 year and <5 years of age, from the study area admitted at Manhiça District Hospital in 2006-2011.

The impact of Hib conjugate vaccine introduction on the incidence of invasive Hib disease, clinical pneumonia (severe and very severe), and radiologically-confirmed pneumonia is presented in the Table. Overall invasive Hib disease declined significantly in children <1 year (from 173-15 cases per 100 000 person-years at risk) and <5 years of age (from 85-13 cases per 100 000 person-years at risk); there were no cases of Hib meningitis following Hib conjugate vaccine introduction. Very severe pneumonia also declined among both age groups (from 1183-838 cases per 100 000 person-years at risk) among children <1 year and from 623-409 cases per 100 000 person-years at risk among children <5 years of age. Radiologically-confirmed pneumonia declined significantly among children <2 years of age, and, even though a decline was observed among children <1 year of age, it was not statistically significant. No significant declines were observed in severe pneumonia.

Table.

Comparison of baseline incidence (2006-2008) and postvaccine introduction incidence (2010-2011) of invasive Hib-confirmed meningitis, pneumonia, and radiologically-confirmed pneumonia among children <1 year and <5 years in Manhiça, Mozambique

Outcome	Age	Pre-Hib vaccine introduction (2006-2008)		Post-Hib vaccine introduction (2010-2011)		Incidence rate ratio‡ (95% CI)	P value
		Episodes (TAR)*	Incidence rate†	Episodes (TAR)*	Incidence rate†
Invasive Hib disease	<1 y	17 (9811)	173	1 (6562)	15	0.09 (0.01, 0.66)	.0007
	<5 y	38 (44 945)	85	4 (31 541)	13	0.15 (0.05, 0.42)	<.0001
Confirmed bacterial meningitis	<1 y	8 (9812)	82	0 (6562)	0
	<5 y	17 (44 945)	38	0 (31 541)	0
Severe pneumonia	<1 y	497 (9793)	5075	314 (6549)	4794	0.95 (0.82, 1.10)	.4751
	<5 y	1076 (44 904)	2396	695 (31 513)	2205	0.91 (0.82, 1.01)	.0826
Very severe pneumonia	<1 y	116 (9808)	1183	55 (6560)	838	0.71 (0.51, 0.98)	.0338
	<5 y	280 (44 936)	623	129 (31 536)	409	0.64 (0.52, 0.80)	.0004
Radiologically-confirmed pneumonia§	<1 y	113 (6155)	1836	47 (3305)	1422	0.77 (0.53, 1.11)	.1510
	<2 y	188 (11 829)	1589	69 (6498)	1062	0.68 (0.50, 0.94)	.0144

TAR, time at risk.

^*TAR calculated as person-years at risk.

^†Incidence rate per 100 000 person-years at risk.

^‡Incidence rate ratio: incidence rate (post-Hib vaccine introduction) vs incidence rate (pre-Hib vaccine introduction).

^§Pre-Hib vaccine introduction 2004-2006 and post-Hib vaccine introduction 2011 only.

Incidence rates of invasive pneumococcal disease (IPD) among children <1 year and <5 years of age were examined to assess whether changes in the surveillance system might have influenced observed trends invasive Hib disease. The incidence of IPD in children <1 year of age did not differ significantly in postvaccine period (381 of 100 000) compared with the baseline period (479 of 100 000), P = .461 (incidence rate risk 0.75, 95% CI 0.45-1.23). However, among children <5 years of age, the rate of IPD in the post-Hib vaccine period (168 of 100 000) was significantly lower than during the baseline period (254/100 000), P = .008 (incidence rate risk 0.63, 95% CI 0.45-0.89).

Discussion

These results demonstrate that, in an area with high HIV prevalence, Hib conjugate vaccine significantly reduced incidence rates of invasive Hib disease, Hib meningitis, very severe clinical pneumonia, and radiologically-confirmed pneumonia among children in Mozambique. The significant decline in invasive Hib disease, bacterial meningitis, and mortality are consistent with the impact of Hib disease in other low-income and sub-Saharan Africa settings.^3,5,11,12 The impact was observed in both children <1 year old, who have the highest rates of Hib disease and were age-eligible for vaccination during the study period, and in children <5 years of age, which included some children who were too old to have received vaccine. The decline in Hib disease in older children likely reflects herd immunity which has been observed in other settings following Hib conjugate vaccine introduction.²³

Interestingly, the rates of invasive Hib disease were declining somewhat in <5-year-olds and <1-year-olds prior to Hib conjugate vaccine introduction. In addition, rates of IPD declined during the study period despite the lack of pneumococcal conjugate vaccine. The invasive bacterial surveillance system was stable during this time period, so we do not believe the pre-Hib conjugate vaccine declines in Hib disease or the observed decline in IPD to be an artifact. Multiple factors such as improved nutritional status, more effective prevention of mother to child transmission of HIV, and better access to HIV treatment may have contributed to the downward trends in the incidence rate of IPD and Hib among older children. However, following Hib conjugate vaccine introduction, the rapid reduction in invasive Hib disease was clear and dramatic, suggesting that Hib conjugate vaccine use caused much of the observed reduction in disease incidence between 2009 and 2010.

The impact of Hib conjugate vaccine on pneumonia was less evident. There was a significant decline in very severe pneumonia and radiologically-confirmed pneumonia following Hib conjugate vaccine. However, no impact was seen on clinical severe pneumonia. Pneumonia is a nonspecific end-point that can be caused by multiple etiologies.^24,25 Because Hib is considered a leading cause of severe pneumonia, it is not surprising that the impact of the vaccine would be more easily measured among the most severe end of the pneumonia disease spectrum (very severe pneumonia). The lack of a decline in clinical severe pneumonia may be due to the lower specificity of that outcome. It is also possible that suboptimal vaccine coverage (79% for the third dose of Hib conjugate vaccine in 2010 in Manhiça District) blunted the observed impact against pneumonia. Additional years of surveillance, with higher levels of Hib conjugate vaccine coverage, may be necessary to be able to evaluate the impact of Hib conjugate vaccine against pneumonia.

These surveillance data could not provide an estimate of the impact of Hib conjugate vaccine specifically among children with HIV. The HIV prevalence among children in the surveillance areas is not well characterized, therefore, it was not possible to calculate incidence among children with and without HIV separately. There is uncertainty about trends in HIV infection among children in the study area because of the increasing prevalence among pregnant women, but also increasing access to drugs to prevent mother-to-child HIV transmission and improved survival among children with HIV. Thus, trends in pediatric HIV may have masked or increased the apparent impact of Hib conjugate vaccine. Nonetheless, these data demonstrate that use of Hib conjugate vaccine in a routine infant immunization program has tremendous impact, even in an area with a high HIV prevalence among pregnant women.²⁶ Although the case numbers are very small, these data suggest that following the introduction of Hib conjugate vaccine, children with HIV continue to be an important risk group for invasive Hib disease, because a significant proportion of invasive Hib disease occurred among children with HIV, and the only Hib conjugate-vaccinated child had HIV. Hib disease in children with HIV can occur at older ages than typically seen for children without HIV; therefore, it is important to continue to monitor Hib disease among children with HIV as the vaccine-eligible cohort ages. In addition, in South Africa—another country with high HIV burden—initial declines in Hib disease following Hib conjugate vaccine introduction were followed several years later by an increase in Hib cases among fully vaccinated children.²⁷ Although several factors may affect the risk for Hib conjugate vaccine failures (including underlying medical conditions, concurrently administered vaccines, and whether a booster dose is used),²⁸ the findings from South Africa highlight the need for continued surveillance for Hib disease in areas with a high HIV prevalence.

Our study had limitations, including the presence of concurrent factors that affected trends, challenges with measuring impact for nonspecific outcomes, and uncertainty about pediatric HIV infection in the source population. Even so, we were able to show an important impact of Hib conjugate vaccine on life-threatening childhood illnesses, including Hib meningitis, bacteremia Hib pneumonia, very severe pneumonia, and radiologically-confirmed pneumonia among children in Mozambique. The results of this study complement previous results from developing countries and show how Hib conjugate vaccine can help reduce child morbidity and mortality in low-income settings.

Glossary

CISM

Centro de Investigacão em Saúde de Manhiça

CSF

Cerebrospinal fluid

CXR

Chest radiographs

DSS

Demographic surveillance system

DTP

Diphtheria/tetanus/whole-cell-pertussis

EPI

Expanded Program on Immunization

Hep B

Hepatitis B

Hib

Haemophilus influenzae type b

IPD

Invasive pneumococcal disease

PCR

Polymerase chain reaction