Streptococcus pneumoniae is estimated to cause the deaths of 826,000 children in the world each year.1 To reduce the burden of pneumococcal disease, several middle and high-income countries have implemented routine pediatric immunization with pneumococcal conjugate vaccines (PCV) and have experienced reduced rates of community-acquired pneumonia and pneumococcal pneumonia.2–6 As certain nonvaccine pneumococcal serotypes were found to increase in incidence in countries where the vaccine had been introduced, conjugate pneumococcal vaccines with additional serotypes have been developed, such as the 13-valent PCV (PCV-13). Currently, the World Health Organization supports the inclusion of PCV vaccines with additional serotypes to national pediatric immunization schedules worldwide.7
On December 12, 2010, Nicaragua became the first nation eligible for GAVI Alliance support to introduce routine immunization with PCV-13. Nicaraguan infants are offered the vaccine in a “3 + 0” dosing schedule, at 2, 4 and 6 months of age. During the first year of the immunization program, a single catch-up dose was also provided to children aged 12–24 months.
While substantial reductions in hospitalizations for pneumococcal disease have been observed in higher income countries following the introduction of pneumococcal immunization programs, less is known about the impact of PCV immunization programs in lower income countries. In these settings, the prevaccine burden of pneumococcal disease and distribution of circulating pneumococcal serotypes are often unknown. Also, the clinical trials of PCVs were primarily conducted in the United States and Europe; there may be differences in the host immune response to the vaccine in developing countries.
The goal of this study was to examine changes in pediatric hospitalizations and ambulatory (outpatient) visits for pneumonia before and after introduction of a PCV-13 immunization program in León, Nicaragua. To investigate whether there may have been differences in health care access during the years we studied, we planned to also report on health facility visits for an unrelated diagnosis, diarrhea. A secondary goal included examining changes in all-cause infant mortality before and after PCV-13 introduction.
MATERIALS AND METHODS
Nicaragua is a low-middle income country in Central America with a per capita gross domestic product of US$1020.8 This study was conducted in the Department of León in western Nicaragua. León is home to Nicaragua’s second largest city, the Municipality of León (2012 population: 193,598), 5 peri-urban municipalities and 4 rural municipalities, for a total population of 396,969 in 2012. The climate is tropical with distinct dry and rainy seasons; an annual peak in pneumonia cases occurs during the rainy season, July to October.9 Among the 56 typeable S. pneumoniae isolates from Nicaraguan children submitted to the SIREVA II surveillance network between 2000 and 2010, 71% had serotypes that are included in PCV-13.10
The National Immunization Program offered routine influenza immunization for children <5 years of age throughout all of the years studied (2008–2012), with separate H1N1 and seasonal influenza vaccines offered in 2010 and combined H1N1 and seasonal influenza vaccines offered in 2011 and 2012.
As we also analyzed diarrhea as a comparison diagnosis, there were no notable changes in diarrhea prevention activities or household water supply in the Department over the years studied. The pentavalent rotavirus vaccine had been added to the national immunization schedule in 2006, 2 years before the years studied.
Health Facilities and Data Collection
The study analyzed data collected between 2008 and 2012 at 107 public health facilities in the Department of León. The facilities included 93 health posts, 13 primary care health centers and the only public referral hospital [Hospital Escuela Oscar Danilo Rosales Argüello (HEODRA)] in the Department. Service is provided free-of-charge at all public health facilities. All public health facilities in Nicaragua are under a government mandate to report visits for pneumonia, diarrhea and infant deaths to their department epidemiology office. Apart from the public health care system, the local health ministry estimates that 10% of the population access private health facilities. Data were maintained in deidentified format, and the study was determined to be exempt from full Institutional Review Board review.
While data were available from the public health facilities since 2005, before data analysis, we planned to limit our analyses to 2008–2012. In 2007, a new government policy substantially reduced patients’ costs of accessing healthcare. Therefore, we focused our analysis on years without potential changes in healthcare access.
Assignment of Diagnosis Codes and Information Management
Hospital admission diagnoses are assigned at the time of admission by the medical team admitting the patient and are maintained by the hospital epidemiology office in an electronic database. Similarly, emergency department visit diagnoses are assigned by the emergency department provider treating the patient and are maintained in the electronic hospital epidemiology database. The definition of pneumonia used by hospital physicians is defined on page 81 of the Ministry of Health’s Guide to the Management of Common Infectious Diseases in Childhood and Malnutrition,11 as an infectious syndrome including constitutional and respiratory symptoms, present with physical exam findings of consolidation, with confirmation of infiltrate on chest radiograph. According to the Guide and HEODRA protocols, all patients admitted for pneumonia are required to receive a chest radiograph.
At the primary care centers and health posts, visit diagnoses are assigned by the medical provider treating the patient and are reported to the department epidemiology office, where reports are maintained in an electronic database. Both the hospital and department databases include information on the patient’s sex, age group, date of the visit, municipality of origin and diagnosis code. In the primary care centers and health posts, medical providers use the same clinical definition of pneumonia as above for the hospital setting, but due to lack of access to radiographic imaging, patients rarely receive chest radiography.
Infant deaths of any cause occurring either in the hospital or primary care centers are reported to the department epidemiology office on a weekly basis. Deaths occurring outside any health facility are investigated by a team including a Ministry of Health physician, the department epidemiologist and the local medical provider before assignment of the infant mortality code. The comparison diagnosis used in this study, diarrhea, was defined as an increase in stool frequency to 3 or more stools within a 24 hour period with increased stool volume.11
The national census is performed every 10 years by the Nicaraguan Institute of Development Information; the last census occurred in 2005. Official population estimates by age group for each municipality are provided for intervening years by the Ministry of Health.
PCV-13 Coverage by Municipality
The National Immunization Program maintains data on immunization coverage by year and age group for each routinely administered vaccine. By the end of 2011, 63% of infants in León had received 3 doses of PCV-13 (range: 49–71% by municipality) and 87% of 1-year olds had received 1 dose of PCV-13 (range: 68–100% by municipality). By the end of 2012, 97% of infants in León had received 3 doses of PCV-13 (range: 80–100% by municipality) and 98% of 1-year olds had received 1 dose of PCV-13 (range: 80–100% by municipality).
The primary outcomes examined were the incidence rates of pneumonia hospitalizations and ambulatory visits for pneumonia among infants (<12 months of age) and 1-year olds (12–23 months) before and after PCV-13 introduction (2008–2010 vs. 2011–2012). These age groups were chosen as the primary focus both due to their high burden of pneumococcal disease and eligibility for PCV-13 immunization during the years studied. Secondary outcomes examined included incidence rates of pneumonia hospitalizations and ambulatory visits for pneumonia among children aged 2–4 and 5–14 years and all-cause infant mortality. As a comparison, we also examined incidence rates of health facility visits for diarrhea over the years studied, including both hospitalizations and ambulatory visits.
Numbers of all health facility visits for pneumonia were extracted from the Department database between January 2005 and December 2012 and plotted by week. For further analysis, the period was limited to 2008–2012. We defined 2 periods: “prevaccine,” January 2008 to December 2010, and “vaccine,” January 2011 to December 2012. Counts of hospital admissions for pneumonia were obtained from the hospital epidemiology database. In this analysis, emergency department visits that did not result in hospital admission were categorized as ambulatory visits, along with visits to primary care centers or health posts.
Incidence rates of pneumonia hospitalizations, ambulatory visits for pneumonia, all-cause infant mortality and corresponding 95% confidence intervals (CI) in the vaccine and prevaccine periods were estimated using generalized estimating equations, allowing for correlation of individuals within municipality. In this analysis, exposure time was estimated by official municipality population estimates for each calendar year. Generalized estimating equations were also used to estimate incidence rate ratios and 95% CI for the vaccine period when compared with the prevaccine period, stratified by age group. In these analyses, we additionally controlled for municipality as categorized into 1 of 3 groups: (1) León municipality (urban area where the hospital is located), (2) municipalities sharing a border with León municipality (peri-urban) and (3) municipalities not sharing a border with León municipality (rural), to account for distance from a patient’s home municipality to the hospital and potential differences in care seeking by municipality category.
All Health Facility Visits for Pneumonia
The total numbers of health facility visits for pneumonia by week between 2005 and 2012 for infants and children aged 12– 23 months are shown in Figure 1.
During the years studied (2008–2012) the numbers of all-cause hospitalizations at HEODRA remained relatively stable with a low of 21,627 in 2008 and a high of 23,155 in 2010. During the period before the program’s implementation (January 2008 to December 2010), there were 1578 pneumonia hospitalizations reported among infants in León within a total of 24,491 years of exposure and 602 pneumonia hospitalizations reported among 1-year olds within 24,193 years of exposure. During the vaccine period (January 2011 to December 2012), there were 716 pneumonia hospitalizations reported among infants within 16,676 years of exposure and 296 pneumonia hospitalizations reported among 1-year olds within 16,077 years of exposure. Incidence rates of pneumonia hospitalizations by year for children <5 years of age are shown in the upper panel of Figure 2. The estimated adjusted incidence rate ratios (IRRa) and corresponding CIs for the period after (2011–2012) versus before (2008–2010) introduction of PCV-13 are shown in Table 1; the estimated IRRa for pneumonia hospitalizations was 0.67 (0.59–0.75) among infants and 0.74 (0.67–0.81) among 1-year olds.
Ambulatory Visits for Pneumonia
In the prevaccine period, there were 12,301 ambulatory visits for pneumonia reported among infants in León within a total of 24,491 years of exposure, and 10,441 ambulatory visits among 1-year olds within a total of 24,193 years of exposure. In the vaccine period, there were 7291 ambulatory visits for pneumonia reported among infants within 16,676 years of exposure and 5817 visits among 1-year olds within a total of 16,077 years of exposure. Incidence rates of ambulatory visits for pneumonia by year for children <5 years of age are shown in the lower panel of Figure 2. The estimated IRRa for the period after (2011–2012) versus before (2008–2010) introduction of PCV-13 by age group are shown in Table 1; the estimated IRRa for ambulatory visits for pneumonia was 0.87 (95% CI: 0.75–1.01) among infants and 0.84 (95% CI: 0.74–0.95) among 1-year olds.
All-cause Infant Mortality
In the prevaccine period, there were 339 total infant deaths in León within 24,491 years of exposure. Of these infant deaths, 263 were among neonates and 76 were among infants aged 1–11 months. In the vaccine period, there were 155 infant deaths within 16,676 years of exposure. Of these infant deaths, 105 were among neonates and 50 were among infants aged 1–11 months. The infant mortality rate was 138 infant per 10,000 child-years (95% CI: 133–144) in the prevaccine period and 93 per 10,000 child-years (95% CI: 79–109) in the vaccine period. The IRRa of infant mortality in the vaccine versus prevaccine period was 0.67 (95% CI: 0.57–0.80).
Health Facility Visits for Diarrhea
For comparison, the IRRa of all health facility visits for diarrhea are shown in Table 1. The incidence rates of health facility visits for diarrhea were not significantly different during 2011–2012 when compared with 2008–2010 among children <24 months of age. There was a significant increase in the incidence rate of diarrhea visits among children 2 years of age and older over the 2 periods.
We observed lower rates of hospitalizations and ambulatory visits for pneumonia among children eligible for immunization during the first 2 years of a 3+0 PCV-13 immunization program in this developing world setting. Among infants, we also found a surprising 33% (95% CI: 20–43) lower all-cause infant mortality rate in the 2 years following PCV-13 introduction. Among 1-year olds, we observed lower rates of health facility visits for pneumonia during the second year of the program when compared with the first year (Fig. 2). This corresponds with the introduction schedule, which offered a single “catch-up” dose of the vaccine to 1-year olds during the first year of the program, but by the second year, 1-year olds had been offered the full course of vaccine. Rapid attainment of high PCV-13 coverage in this setting may have contributed to the observed changes in health facility visits and infant mortality soon after program implementation.
Lower rates of hospitalizations for pneumonia were observed in children under 2 years of age soon after PCV introduction in other areas, such as the United States (32% reduction in the second to seventh year of program with 3+1 PCV-7 schedule),12 Australia (38% reduction in first 30 months of program with 3+0 PCV-7 schedule)3 and Uruguay (approximately 50% reduction in second year of the program with 2+1 PCV-7 schedule).5 As such, our findings from a developing nation contribute to the growing body of evidence showing a substantial reduction in pneumonia hospitalizations in populations where PCVs have been introduced. We also observed decreases in ambulatory visits for pneumonia, although the magnitude of the reduction was not as great as for pneumonia hospitalizations. Reports from the United States after PCV introduction are mixed with 1 study showing decreases in ambulatory visits for pneumonia,13 but 2 other studies showing no changes.14,15 It may be more difficult to observe an effect of PCV immunization programs on rates of ambulatory visits, as pneumonia diagnosis in the ambulatory setting is less likely to be specific for pneumococcal pneumonia; in the ambulatory setting, pneumonia diagnosis is typically based on clinical signs and symptoms rather than blood culture or chest radiography. Alternately, PCVs may be more effective at reducing severe disease that would require hospitalization when compared with milder disease that would result in ambulatory treatment.
In addition to overall reductions in health facility visits, we also observed decreased seasonality of pneumonia cases after PCV-13 introduction (Fig. 1). Pneumonia has been shown to have a seasonal pattern in the tropics with a peak during the rainy season,16,17 which may be related to seasonal patterns in respiratory virus transmission, such as influenza, or RSV in infants.18,19 In our study, seasonality was observed before PCV-13 introduction, but pneumonia cases were more evenly distributed after PCV-13 introduction. This suggests that the immunization program may be preventing postviral pneumococcal pneumonia. If pneumonia cases continue to be more evenly distributed, this may be beneficial in terms of health care resources, as health facilities do not need to be equipped to treat a high volume of patients at once.
This study provides an evaluation from a developing nation using the 3+0 immunization schedule, including 3 priming doses in infancy, but no booster dose at 1 year of age. Because it is possible that vaccine-elicited immunity will wane over time without the booster dose, it will be important to continue to monitor the impact of the immunization program in Nicaragua in children aged 1 year and older. In a clinical trial in Ghana, where the 9-valent PCV was given on a 3+0 dosing schedule, pneumococcal antibody titers remained high among immunized children up to 4 years after immunization.20 Also, after the national 3+0 PCV-7 immunization program in Australia, reductions in hospitalizations for pneumonia were observed among older age groups of children.21 Currently, there are no data to support immunogenicity or efficacy of the 2+1 immunization schedule over the 3+0 immunization schedule,22 and it is possible that coverage for the third dose of PCV-13 would be lower if it is delayed until 12 months of age.
To our knowledge, this is the first postlicensure study in a developing nation to show a decline in all-cause infant mortality in a population after PCV introduction. However, these results should be interpreted with caution, as we cannot rule out that factors other than PCV-13 introduction were responsible for this change. To further investigate the causality of these findings, we examined hospital records for causes of child deaths during the years studied. During the prevaccine period among children <24 months of age, there were 6 deaths per year among infants categorized as pneumonia related and during the vaccine period, there were 5 pneumonia-related deaths per year. Because of the small numbers of overall infant deaths categorized as pneumonia related, it was not possible to assess that a reduction in pneumonia-related deaths contributed to the reduction in all-cause infant mortality. Still, it is conceivable that PCV-13 introduction was responsible for the change in infant mortality, as a proportion of infant deaths in León occur outside the hospital setting, where the cause of death is not easily identified. Also, PCV-13 may cause a decrease in mortality through its effect on invasive pneumococcal disease, including sepsis and meningitis. As bacterial culture is not routinely practiced in this low-resource setting, the cause of many infant deaths remains unidentified.
Children in age groups not eligible for PCV-13 also experienced lower rates of health facility visits for pneumonia in the vaccine period when compared with the prevaccine period. This potential indirect effect (herd protection) of PCVs on unimmunized children has also been observed soon after PCV introduction in other populations with high PCV coverage, such as in Australia and Uruguay, but not during the early years after PCV introduction in the United States.3,5,12 The presence of an early indirect effect may depend on high vaccine coverage; coverage for 3 doses of PCV during the second year of PCV immunization programs in Australia, Uruguay and Nicaragua (León) were 91%, 82% and 97%, respectively, when compared with 30% in the United States.3,5,23
As this was an observational study, we cannot prove that the reductions observed were caused by PCV-13 introduction and not by other factors, such as differences in health care access during the years studied. To address this concern, we also examined incidence rates of health facility visits for an unrelated diagnosis, diarrhea. We did not observe reductions in visits for diarrhea, arguing against health care access as a cause for the decline in pneumonia-related health facility visits. Also, annual differences in influenza transmission intensity may influence all-cause pneumonia incidence. While this variability is diminished by our examination of multiple years, more years of surveillance in the vaccine period would be needed to show a sustained impact of the program. Another limitation of this study is that we did not collect blood samples or other laboratory samples to prove the bacterial etiology of the pneumonia case. While culture-positive pneumococcal pneumonia would provide a more specific diagnosis, S. pneumoniae is isolated from blood culture in only a minority of pneumococcal pneumonia cases, so depending on laboratory-confirmed pneumococcal pneumonia alone would exclude a substantial proportion of disease. Also, as S. pneumoniae is thought to be the most common cause of community-acquired pneumonia, accounting for between 20 and 60% of cases,24 a reduction in this most common pneumonia etiology would likely reduce overall rates of pneumonia.
In summary, the findings of this postlicensure study support World Health Organization recommendations to introduce PCVs to protect children from pneumonia. PCV immunization programs may also assist developing nations in the successful attainment of Millennium Development Goal Four, the reduction of infant mortality. Public health policy makers in other developing countries, especially those eligible for GAVI support, should work towards providing PCVs to their pediatric populations.
The authors would like to acknowledge the staff of the Sistemas Locales de Atención Integral a la Salud, León (SILAIS-León), the HEODRA and Erica Lloyd for their contributions to the study. This research was supported by an Investigator-initiated Research Award from Pfizer. The sponsors of the study had no role in the study design, data collection, statistical analysis, interpretation or writing of the article. S.B.-D. was responsible for the content of the article and the decision to submit for publication.
1. O’Brien KL, Wolfson LJ, Watt JP, et al. Hib and Pneumococcal Global Burden of Disease Study Team Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet. 2009; 374:893–902
2. Grijalva CG, Nuorti JP, Arbogast PG, et al. Decline in pneumonia admissions after routine childhood immunisation with pneumococcal conjugate vaccine in the USA: a time-series analysis. Lancet. 2007; 369:1179–1186
3. Jardine A, Menzies RI, McIntyre PB. Reduction in hospitalizations for pneumonia associated with the introduction of a pneumococcal conjugate vaccination schedule without a booster dose in Australia. Pediatr Infect Dis J. 2010; 29:607–612
4. Koshy E, Murray J, Bottle A, et al. Impact of the seven-valent pneumococcal conjugate vaccination (PCV7) programme on childhood hospital admissions for bacterial pneumonia and empyema in England: national time-trends study, 1997–2008. Thorax. 2010; 65:770–774
5. Pírez MC, Algorta G, Cedrés A, et al. Impact of universal pneumococcal vaccination on hospitalizations for pneumonia and meningitis in children in Montevideo, Uruguay. Pediatr Infect Dis J. 2011; 30:669–674
6. Ho PL, Chiu SS, Chow FK, et al. Pediatric hospitalization for pneumococcal diseases preventable by 7-valent pneumococcal conjugate vaccine in Hong Kong. Vaccine. 2007; 25:6837–6841
12. Griffin MR, Zhu Y, Moore MR, et al. U.S. hospitalizations for pneumonia after a decade of pneumococcal vaccination. N Engl J Med. 2013; 369:155–163
13. Zhou F, Kyaw MH, Shefer A, et al. Health care utilization for pneumonia in young children after routine pneumococcal conjugate vaccine use in the United States. Arch Pediatr Adolesc Med. 2007; 161:1162–1168
14. Kronman MP, Hersh AL, Feng R, et al. Ambulatory visit rates and antibiotic prescribing for children with pneumonia, 1994–2007. Pediatrics. 2011; 127:411–418
15. Grijalva CG, Poehling KA, Nuorti JP, et al. National impact of universal childhood immunization with pneumococcal conjugate vaccine on outpatient medical care visits in the United States. Pediatrics. 2006; 118:865–873
16. Paynter S, Ware RS, Lucero MG, et al. ARIVAC Consortium Poor growth and pneumonia seasonality in infants in the Philippines: cohort and time series studies. PLoS One. 2013; 8:e67528
17. Enwere G, Cheung YB, Zaman SM, et al. Epidemiology and clinical features of pneumonia according to radiographic findings in Gambian children. Trop Med Int Health. 2007; 12:1377–1385
18. Shek LP, Lee BW. Epidemiology and seasonality of respiratory tract virus infections in the tropics. Paediatr Respir Rev. 2003; 4:105–111
19. Weber MW, Mulholland EK, Greenwood BM. Respiratory syncytial virus infection in tropical and developing countries. Trop Med Int Health. 1998; 3:268–280
20. Akinsola AK, Ota MO, Enwere GC, et al. Pneumococcal antibody concentrations and carriage of pneumococci more than 3 years after infant immunization with a pneumococcal conjugate vaccine. PLoS One. 2012; 7:e31050
21. Jardine A, Menzies RI, McIntyre PB. Reduction in hospitalizations for pneumonia associated with the introduction of a pneumococcal conjugate vaccination schedule without a booster dose in Australia. Pediatr Infect Dis J. 2010; 29:607–612
24. Bartlett JG. Diagnostic tests for agents of community-acquired pneumonia. Clin Infect Dis. 2011; 52:(suppl 4)S296–S304