As a large-scale before-after study, PCVIS can answer important questions about the use of pneumococcal conjugate vaccine in Kenya.
How well does the vaccine protect children who have been immunized against pneumococcal disease?
By tracking the incidence of pneumococcal disease (pneumonia, meningitis and sepsis) before and after PCV was introduced, our study can determine how much disease has been prevented by the vaccine.

Cases of invasive pneumococcal disease (IPD) in children and adults admitted to Kilifi County Hospital (KCH).


PCV10 reduced hospital admissions for pneumonia in children under five by more than one quarter after four years of vaccine use.


PCV10 virtually eliminated vaccine-type invasive pneumococcal disease after four years of use.

How well does the vaccine prevent transmission of pneumococcal bacteria within the community?
An important component of pneumococcal disease transmission is the carriage of the bacteria in the nose of healthy children and adults - potential pneumococcal disease is commonly spread by asymptomatic carriers within communities. Because carriage of the bacteria is a necessary first step for a person to become sick with pneumococcal disease and/or spread the bacteria to others, reducing carriage is likely to reduce transmission of the bacteria and therefore the number of cases of disease. Our study is measuring community carriage of pneumococcus in Kilifi along with pneumonia and invasive pneumococcal disease to gain a more comprehensive understanding of vaccine impact, beyond the impact on disease alone.

From 2011 – 2016, PCV10 reduced the carriage of vaccine type bacteria in Kenya, but not to the very low levels seen in middle- and high-income countries.

How well does the vaccine provide indirect protection for people who haven't been vaccinated?
Although pneumococcal disease is a major cause of disease and death in children, it also affects the rest of the population, especially elderly adults and those who are immunocompromised. Because PCV works largely by reducing transmission via carriage in healthy people, we have also found a reduction in vaccine-type IPD in unvaccinated populations, showing that PCV10 provided substantial herd protection. In addition to measuring vaccine impact in vaccinated children, we also looked at the potential impact on disease and transmission in unvaccinated people.

The Gavi supported PCV10 program reduced vaccine-type IPD in unvaccinated populations (those under two months, or too young to be vaccinated, and those over five years or too old to be vaccinated), showing that PCV10 provided substantial population protection/herd immunity.

Does the use of PCV, which only offers protection against some strains of pneumococcal bacteria, result in an increase in transmission or disease caused by non-vaccine-type strains?
Evidence from early use of PCVs in developed country settings has shown that strains not included in the vaccines could begin replacing vaccine-type strains in terms of the amount if transmission and disease they cause. If the magnitude of replacement is high enough, it has the potential to significantly reduce the overall impact of vaccine use. As part of our analysis, we sought to identify and quantify this potential serotype replacement.

Whilst an increase in non-vaccine type pneumococcal bacteria, particularly serotype 19A, was observed across all age groups, evidence of serotype replacement disease was not found up to six years after the introduction of PCV10. Further research is needed to continue to monitor for incidence of serotype replacement disease and understand the persistent prevalence of vaccine-type pneumococcal carriage.

How good is the health system at ensuring children are vaccinated, and how accurate are vaccine coverage estimates?
Because PCVIS is embedded in a robust health and demographic surveillance system, we have been able to track the number of children who are vaccinated each year and capture information about the number of vaccine doses each child receives and the timeliness of each. This allows an in-depth understanding of vaccine coverage in the population, which can offer insights about the functionality of the health system as well as context for interpreting the measured vaccine impact on disease and transmission. It also enables us to compare the accuracy of different methods of vaccine coverage measurement, including routine health system vaccination records, against a ‘gold standard’ of vaccination coverage measurement. Gavi champions the Fully Immunized Child as a proxy measure of health system effectiveness, so it is important to understand the accuracy of immunization estimates. nnIn 2018 we transitioned from the Kilifi Vaccine Monitoring Study (KiVMS) led by PCVIS to the Kilifi Electronic Immunisation System (KEIR) which is run by Kilifi County Ministry of Health staff with technical support from the PCVIS team.

Coverage with at least two PCV10 doses in children aged 2–11 months was 84% in 2016, five years after vaccine introduction; coverage with at least one dose in children aged 12–59 months was 87% in 2016.


Coverage surveys in 12–23 month old children overestimate protection while administrative methods underestimate full immunisation coverage.

Is the PCV vaccine program cost-effective in this middle-income setting?
The cost of introducing new vaccines is substantially subsidized for Gavi-eligible countries, often making the decision to introduce a Gavi-supported PCV program an obvious choice. However, as countries like Kenya approach Gavi graduation and the subsequent requirement to fully self-finance their vaccine programs, it is important that the costs and benefits of PCV, which is an expensive product, are understood. We performed a cost-effectiveness evaluation of the PCV program to inform evidence-based decisions about its continuation once Gavi support ends.

The PCV program in Kenya is cost-effective. If Kenya continues the PCV program beyond the start of Gavi transition in 2022, every $1 million invested will result in 6,536 additional healthy years of life compared to discontinuing the vaccine. If children born after 2022 no longer receive PCV, IPD incidence would return to pre-vaccine levels within a decade.

Why does vaccine-type pneumococcal carriage persist and what causes residual transmission?
In Kilifi, 5 years post vaccine introduction, 10% of children still carried vaccine-type pneumococci at the back of their noses. This means that unvaccinated individuals are still being exposed to vaccine-type pneumococci and remain at risk of disease. The Pneumococcal Transmission Study set out to assess whether people in other parts of Kenya carry more or less vaccine-type pneumococci compared to Kilifi, to identify the cause of residual transmission and to propose interventions to eliminate or reduce the presence of vaccine-type pneumococci.

There was high residual VT-pneumococcal carriage in children across Kenya six-years post-PCV10 introduction, especially in rural areas. Analysis of the causes of residual transmission is ongoing.

Does a urine antigen test improve our ability to detect cases of pneumococcal pneumonia among adults?
Pneumonia caused by pneumococcal bacteria is usually detected by culturing (growing) pneumococcal bacteria from blood samples collected from patients. However, culturing has low sensitivity, meaning that the test can be negative for some patients with pneumococcal pneumonia, giving a false negative result. A urine antigen test, which detects pneumococcal bacteria in the urine of patients with pneumococcal pneumonia, may be more sensitive than blood culture for diagnosis. The urine antigen test has been used in some high-income settings to diagnose pneumococcal pneumonia. We are conducting analyses to determine the utility of the urine antigen test for diagnosing pneumococcal pneumonia in Kenya. To be useful, the urine antigen test will need to have high specificity. In other words, it should return a negative test in nearly, if not all, patients without pneumococcal pneumonia and so should produce few to no false positive results. The urine antigen test will also need to have better sensitivity than culture for it to be considered as a viable alternative.
How well would a PCV schedule with a booster dose prevent transmission of vaccine-type strains of pneumococcal bacteria?
The World Health Organization (WHO) recommends administration of PCV as three primary doses with no booster dose (3+0) or two primary doses with a booster dose (2+1). Primary doses are usually given 4-8 weeks apart. A booster dose is typically given 6 months after the last primary dose, usually at 9 months of age or older. In Kenya, the National Vaccine Immunization Program recommends a 3+0 schedule with the first dose of PCV at age 6 weeks, the second at age 10 weeks and the third at age 14 weeks. Although use of PCV has substantially reduced pneumococcal disease and transmission of pneumococcal bacteria in Kenya, there remains a low level of transmission of vaccine-type strains. In high income settings which use a PCV schedule with a booster dose, there is virtually no transmission of vaccine-type strains. We are developing a series of studies to evaluate whether use of 2+1 PCV schedule in Kenya can reduce transmission of vaccine-type strains.
Does the use of PCV induce a high level of population immunity?
PCV induces antibodies against strains of the bacteria included in the vaccine which protect vaccinated individuals from pneumococcal disease and also prevent transmission of vaccine-type strains of pneumococcal bacteria. We are conducting studies to assess what proportion of young and older children have protective antibody levels and whether these levels of population immunity change over time.