PCV: what it delivered, and going cheaper

A modelling study of 112 low- and middle-income countries estimates that the pneumococcal conjugate vaccine PCV13 could prevent 697,000 child deaths between 2000 and 2030, at a cost of I$851 per disability-adjusted life year (DALY) averted ¹. The same study found that if PCV coverage had matched routine DTP coverage in 2020, a further 146,000 deaths would be prevented ¹. The vaccine works. The gap is in getting it to every child, and at a price every health ministry can pay.

The burden

Pneumococcus (Streptococcus pneumoniae) is still the single largest bacterial cause of pneumonia deaths. The Global Burden of Disease (GBD) 2023 study, published in Lancet Infectious Diseases in 2026, attributes 634,000 lower respiratory infection deaths in 2023 to pneumococcus, across all ages. That is 25.3% of all such deaths, more than any other pathogen ¹¹. Children under 5 and adults over 70 carry the highest burden ¹¹.

For young children the numbers have fallen a long way. WHO's 2025 position paper, citing earlier modelling, puts the toll before PCV at 14.5 million cases and 735,000 deaths a year in HIV-uninfected children under 5. By 2015 these figures had fallen to an estimated 3.7 million cases and 294,000 deaths ¹⁰. Between 2000 and 2015, pneumococcal deaths in children under 5 fell by an estimated 51%, largely because of PCV introduction ¹⁰. The 112-country model used a different burden estimate of 318,000 child deaths in 2015 ¹.

Where. Death from pneumonia is concentrated in sub-Saharan Africa. GBD 2023 found that deaths from lower respiratory infections in children under 5 fell 33.4% since 2010, to 94.8 per 100,000 in 2023. But sub-Saharan Africa was the region furthest from the global target of fewer than 60 deaths per 100,000 ¹¹.

The disease rate itself differs by more than tenfold between settings. Before PCV, reported invasive pneumococcal disease (IPD) in children under 2 was 44 per 100,000 in Europe and up to 167 per 100,000 in the United States. In Mozambique, rates as high as 797 per 100,000 were reported ¹⁰.

How it kills. About 81% of pneumococcal deaths come from pneumonia ¹⁰. Pneumococcal meningitis and bloodstream infection are less common but more deadly. In low- and middle-income countries, case fatality reaches up to 20% for septicaemia and 50% for meningitis. Survivors of meningitis in Africa and Asia face three times the risk of lasting harm, such as hearing loss and seizures, as survivors in Europe ¹⁰.

The trend. WHO projects that PCV will avert around 2 million deaths in children under 5 from 2021 through 2030 ¹⁰.

How it works

The germ. Pneumococcus lives in the back of the nose and throat, often in healthy young children, without causing symptoms. This "carriage" is how it spreads from person to person. Disease usually starts from carriage: the bacteria move into the lungs, the blood or the lining of the brain ¹⁰.

The target. Each pneumococcus is wrapped in a sugar coat, the polysaccharide capsule. It is the main thing that makes the germ dangerous. There are more than 100 capsule types, called serotypes, but only a small number cause most serious disease ¹⁰. Immunity against one serotype mostly does not protect against others ¹⁰. That is why vaccines are named by how many serotypes they cover: PCV10 covers 10, PCV13 covers 13.

The conjugate trick. Infants' immune systems respond poorly to sugar coats alone. A conjugate vaccine joins each capsule sugar to a carrier protein, which the infant immune system can recognise. PCV13 and Pneumosil use a non-toxic diphtheria protein called CRM197. GSK's PCV10 uses a protein from Haemophilus influenzae for most serotypes ¹⁰.

Two kinds of protection. A vaccinated child is protected directly. The vaccine also reduces carriage of vaccine serotypes in the nose. With less carriage, fewer germs pass to babies too young to be vaccinated, to older people, and to people with HIV. This is herd protection, and it is why carriage surveys are used to judge programmes ²¹⁰.

Schedules, in plain terms. Schedules are written as primary doses plus booster doses ¹⁰:

• 3p+0: three infant doses (for example at 6, 10 and 14 weeks), no booster.
• 2p+1: two infant doses at least 8 weeks apart, then a booster at 9–18 months.
• 1p+1: one infant dose from 6 weeks, then a booster from 9 months. This is the new reduced schedule.

WHO says there is no clear evidence that either 3-dose schedule (3p+0 or 2p+1) protects better than the other. Countries may choose based on local factors, such as which visits reach the most children ¹⁰. The single most effective step, in WHO's words, is to ensure high coverage of 3 doses in the routine schedule ¹⁰.

Antibiotic resistance. PCV also helps against drug-resistant infections. After PCV7 was introduced, IPD in children under 2 caused by strains not susceptible to penicillin fell by 82% in South Africa and 81% in the United States ¹⁰.

What the evidence shows

2014–2019: national impact and indirect protection

South Africa's 2014 surveillance report found that invasive pneumococcal disease in children under two fell from 54.8 to 17.0 cases per 100,000 person-years after vaccine introduction. Disease due to PCV7 serotypes fell by 89%, while non-vaccine-serotype disease increased by 33% among HIV-uninfected children ¹⁹. The 2024 follow-up below extends this same national surveillance programme; it is not a second independent experiment.

In Kilifi, Kenya, a 2019 longitudinal study compared periods before and after PCV10 introduction with a catch-up campaign. Vaccine-type invasive disease fell sharply in under-fives (adjusted incidence rate ratio 0.08, 95% CI 0.03–0.22), and also declined in older, unvaccinated age groups ²⁰. Non-vaccine-type carriage increased, but non-vaccine-type invasive disease did not significantly increase during follow-up ²⁰. These studies support direct and indirect protection while showing why disease surveillance, rather than carriage alone, must track serotype replacement. Neither is a randomised estimate of deaths averted.

Deaths and cost: the 112-country model

The team led by Chen at the Saw Swee Hock School of Public Health, National University of Singapore, updated its model with real-world effectiveness data from lower-income settings, WHO–UNICEF coverage estimates and a global burden study ¹. Their central estimate of 697,000 deaths averted has a wide uncertainty range (359,000 to 1,040,000). The model also counts 46.0 million DALYs and 131 million cases prevented ¹.

PCV was estimated to prevent 5.3% of pneumococcal deaths in children under 5 during 2000–2030 ¹. The authors conclude that delays in roll-out and low coverage "have cost many lives" ¹.

Fifteen years in South Africa

The National Institute for Communicable Diseases (NICD) runs one of Africa's largest laboratory surveillance systems for IPD. South Africa introduced PCV7 in 2009 and PCV13 in 2011, both on a 2+1 schedule ².

Across 52,957 cases from 2005 to 2019, IPD in children under 2 was 76.0% lower in 2019 than expected without the vaccine (95% CI 72.8–79.0). Disease from the PCV7 serotypes fell 95.5%, and from the extra PCV13 serotypes 93.8% ². In this age group, disease from non-vaccine serotypes did not change significantly ².

The vaccine also protected people who were not vaccinated. In adults aged 25 to 44, in a setting with high HIV prevalence, IPD fell 50.4% through herd protection ².

Meningitis in Mozambique

At the Instituto Nacional de Saúde (INS), Mozambique, sentinel surveillance in Maputo, Beira and Nampula central hospitals tested 4,075 cerebrospinal fluid samples from 2013 to 2023. It tracked pneumococcal meningitis in children under 5 ³.

It fell from 44.7 to 4.6 per 100,000, a reduction of 89.7% ³. Mozambique started with PCV10 on a 3p+0 schedule and later switched to PCV13 on 2p+1. In the PCV13 period (2020–2023), incidence was 51.2% lower than in the PCV10 period (IRR 0.49, 95% CI 0.4–0.6). No meningitis cases due to PCV13 serotypes were detected in 2020–2023 ³. After the switch, the proportion of meningitis cases that died fell 71.9% ³.

Pneumonia, the main target

Pneumonia is the leading killer that PCV targets, but good African effectiveness data on it are rare. The Manhiça Health Research Centre (CISM) in Mozambique ran a matched case-control study at Mavalane and Manhiça from 2014 to 2016, with 812 cases of X-ray-confirmed pneumonia and age-matched community controls ⁴.

Three doses of PCV10 gave an adjusted effectiveness of 47.2% (95% CI 13.7–67.7) ⁴. Coverage was already high: 94.4% of cases and 97.0% of controls had at least one dose, which makes effectiveness harder to measure ⁴.

The global picture: PSERENADE

The PSERENADE analysis, led from Johns Hopkins Bloomberg School of Public Health with WHO funding, pooled 47 surveillance sites in 30 countries: 32 PCV13 sites (488,758 cases) and 15 PCV10 sites (46,386 cases) ⁵.

Six years after introduction, all-serotype IPD in children under 5 fell 58–74% with either PCV10 or PCV13. Disease from PCV10-type serotypes fell 83–99% ⁵. At PCV13 sites, serotype 19A disease in young children fell 61–79% ⁵. Most of these sites are in high-income countries, so the results are strongest for those settings.

Going cheaper: fewer doses in Johannesburg

The SAMRC Wits Vaccines and Infectious Diseases Analytics unit (VIDA) in Johannesburg randomised 600 HIV-unexposed infants in 2017 to a 1+1 schedule (one primary dose at 6 or 14 weeks, plus a booster at 9 months) or the standard 2+1, using PCV10 or PCV13 (NCT04275284) ⁶.

They followed 354 children to age 5 and measured antibody levels at ages 3, 4 and 5. Both PCV13 1+1 schedules, primed at 6 or 14 weeks, stayed non-inferior to 2+1 up to age 5 ⁶. A companion study from the same group found that carriage of vaccine-type pneumococci stayed suppressed under 1+1 at ages 3 to 5 ⁶. One fewer dose means lower vaccine cost and one fewer clinic visit.

Going cheaper: fewer doses in the Gambia

The Pneumococcal Vaccine Schedules (PVS) trial, run by the MRC Unit The Gambia at LSHTM, is the largest real-world test of 1+1 in Africa. It compared 1p+1 (6 weeks, 9 months) with the Gambia's standard 3p+0 (6, 10, 14 weeks) using PCV13, in 68 geographic clusters (ISRCTN15056916) ¹⁵. WHO reports that 33,001 children were enrolled ¹⁰.

Its primary outcome was vaccine-type carriage in children with clinical pneumonia in year 4. The preliminary results, as summarised by WHO, showed non-inferiority ¹⁰:

• Vaccine-type carriage in children with pneumonia was 8.0% under 1p+1 and 8.4% under 3p+0 (adjusted prevalence ratio 0.95; 95% CI 0.80–1.13).
• Community carriage at all ages was 11.0% and 11.3% (adjusted prevalence ratio 0.97).
• X-ray-confirmed pneumonia in children under 5 did not differ significantly (adjusted IRR 1.07; 95% CI 0.81–1.41).

These are preliminary figures reported in the WHO position paper; the full peer-reviewed report has not yet been published.

Going cheaper: fractional doses in Niger

Epicentre and Médecins Sans Frontières ran a cluster-randomised trial in 63 village clusters in Niger with Pneumosil, a low-cost PCV made by the Serum Institute of India (NCT05175014) ⁷. Mass catch-up campaigns in 2022 targeted children aged 1–9 years with either a full dose, a one-fifth dose, or no campaign. Randomisation was 3:3:1, so each survey sampled about 950 people in each campaign arm and about 320 in the control arm ⁷.

Before vaccination, vaccine-type carriage was 15.6–18.8% across arms. After the campaign, it was 4.6% in the full-dose arm, 8.0% in the fractional-dose arm and 16.5% in the control arm ⁷. Full dose against control gave a risk difference of −16.2% (95% CI −28.6 to −3.0). Fractional doses met the non-inferiority test. No vaccine-related adverse events occurred ⁷. This matters for low-coverage and humanitarian settings, where a quick campaign can cut transmission.

India's low-cost PCV

A cross-sectional study of 700 children in rural Pune, in India's Universal Immunization Programme, found vaccine-type carriage of 23.4% in children who got Pneumosil and 30.9% in PCV-naive children (adjusted OR 0.68, 0.49–0.96) ⁸. The Serum Institute of India, the manufacturer, carried out this study. The effect is modest, and it is the first real-world signal for this vaccine in the country with the most child pneumonia deaths.

Summary of key results

LimitsWhat didn't work (or the limits)

PCV10 on a reduced schedule. In the Johannesburg trial, the PCV10 1+1 schedules were non-inferior at age 3 but failed the test at ages 4 and 5 ⁶. PCV13 outperformed PCV10 across most serotypes in the 1+1 groups ⁶. Countries using a PCV10 product should not assume a dose can be dropped.

No 1+1 evidence yet for Pneumosil. WHO notes that the evidence for 1p+1 comes from GSK's PCV10 and Pfizer's PCV13. There is no direct evidence for a 1p+1 schedule with Pneumosil. WHO says it would likely work, but asks countries that try it to measure its effect on carriage or disease ¹⁰.

Two doses were not enough against pneumonia in Mozambique. Two doses of PCV10 gave an effectiveness of 23.5%, with a confidence interval that included zero (−31.9 to 55.6). The IPD analyses in the same study did not converge, so they gave no usable estimate ⁴.

Pneumonia protection is smaller than IPD protection. Pneumonia has many causes, and the vaccine only prevents the share caused by vaccine serotypes. A systematic review cited by WHO found a pooled efficacy of 27% (95% CI 15–36) against all-cause X-ray-confirmed pneumonia for early PCVs in children under 2 ¹⁰. The 47.2% in Mozambique should be read against that background ⁴.

Serotype replacement. As vaccine serotypes decline, others fill the space. In PSERENADE, IPD from non-vaccine serotypes rose 2.3- to 3.3-fold in children under 5. Serotype 19A rose 1.6- to 2.3-fold at PCV10 sites ⁵. In Mozambique, meningitis from serotypes not in PCV10 or PCV13 rose 76% (IRR 1.76, 95% CI 1.2–2.6) ³. In South Africa, non-vaccine serotype IPD rose 78.5% in adults aged 25 to 44 and 234.9% in adults over 64 ². In children the overall trend is still down, because the fall in vaccine serotypes is bigger than the rise in others ¹⁰.

Transmission continues in the Gambia. The MRC Unit The Gambia at LSHTM surveyed 4,087 people in the Central and Upper River Regions 10 years after PCV13 arrived. Vaccine-type carriage was still 6.4% overall and 13.6% in children aged 5 to 9. Carriage was similar in fully vaccinated and unvaccinated children under 10 ⁹. Residual transmission at this level supports the case for a booster or catch-up campaign.

Carriage is a stand-in. The Niger, India and Gambia results measure carriage, and the Johannesburg trial measures antibodies, not disease or death ⁶⁷⁸⁹.

Model uncertainty. The 697,000 figure is a model estimate with a wide credible interval ¹. It is best read as an order of magnitude.

Cost and scale

Coverage. By the end of 2025, 169 WHO Member States had introduced PCV. Global coverage with the last dose was estimated at 71% ¹². In 2024, 162 of 194 Member States offered PCV to their whole birth cohort ¹⁰. Among the 57 Gavi-eligible countries, 50 had introduced PCV by 2024, and last-dose coverage in that group was 75% ¹⁰.

Gavi and the Advance Market Commitment. Gavi launched the Pneumococcal Advance Market Commitment (AMC) in 2009 to bring down prices for lower-income countries ¹⁴. According to Gavi, donors contributed more than US$1.3 billion, the AMC procured 161 million doses, and Gavi provided US$3.3 billion to support pneumococcal vaccine costs between 2009 and 2020. More than 215 million children had been immunised with Gavi support by the end of 2019 ¹⁴.

Price per dose. UNICEF Supply Division publishes the prices it has agreed with each maker ¹³:

• Serum Institute of India (Pneumosil, PCV10, 5-dose vial): US$1.50–2.00 per dose for Gavi countries in recent years, and US$2.90–4.00 for middle-income countries.
• Pfizer (PCV13, 4-dose vial): US$2.75–2.90 per dose for Gavi countries, against US$17.90–18.05 for middle-income countries.
• Pfizer (PCV13, single-dose vial): a US$3.30 country or Gavi share under the AMC; US$13.50–25.00 for middle-income countries.
• GSK (PCV10, 4-dose vial): US$2.90 per dose for Gavi countries in recent years.
• Biological E / BE Vaccines (PCV14, 5-dose vial): US$2.06 per dose for Gavi countries in forecast years.

For a Gavi country, the gap between the cheapest and dearest option is now small.

Manufacturers. Three products are WHO-prequalified: GSK's Synflorix (PCV10), Pfizer's Prevenar 13 (PCV13) and the Serum Institute of India's Pneumosil (PCV10), prequalified in 2019 ¹⁰. Biological E's 14-valent Pneubevax had been submitted for prequalification at the time of the 2025 position paper.

What to watch

WHO's new rules are in place. In September 2025 WHO published a new position paper, based on SAGE's March 2025 recommendations, replacing the 2019 paper ¹⁰. It allows two cost-saving options, both off-label, and both only in mature programmes ¹⁰:

• Switch to 1p+1 if average third-dose coverage was at least 80% over the previous 5 years (or a recent catch-up campaign reached at least 80% of under-5s, or surveillance shows low vaccine-type carriage), and if the country reaches at least 80% of children with a vaccine given at 9–18 months.
• Use fractional doses (at least 40%) of PCV13 under the same conditions.

WHO also says that if carriage or disease rises, or coverage falls well below 80% for more than a year, countries should run a catch-up campaign or go back to 3 doses. It warns against changing several things at once, such as cutting doses and switching product together ¹⁰. Watch which countries adopt 1p+1 first, and what their surveillance shows.

Catch-up campaigns. WHO now recommends multi-age catch-up campaigns with Pneumosil where routine coverage is persistently below 50% or in humanitarian emergencies. A 20% fractional dose may be used in exceptional situations to stretch supply ¹⁰. This follows directly from the Niger trial ⁷.

Trials to follow.

• PVS, the Gambia (ISRCTN15056916). The full peer-reviewed results of the 1+1 versus 3+0 trial are still to come ¹⁵. They will be the main African evidence behind 1p+1.
• Somaliland displacement camp (NCT04945681). LSHTM, Save the Children, Murdoch Children's Research Institute and Somaliland's Ministry of Health Development ran a PCV campaign for children aged 6 months to 4 years in a camp for displaced people near Hargeisa. The study enrolled 2,882 people and measures vaccine-type carriage in children under 2; primary data collection ended in January 2025 ¹⁶. Results are pending.
• Malnourished children, Timor-Leste (NCT06817421). The OPTIMAL trial, with Menzies School of Health Research and Timor-Leste's Ministry of Health, will test whether a dose of Pneumosil given to children hospitalised with severe acute malnutrition produces a good antibody response. It plans 214 children, with primary completion expected in March 2028 ¹⁷. WHO notes that evidence in malnourished children is limited ¹⁰.
• Higher-valency, lower-cost products. Inventprise's 25-valent vaccine completed a phase 2 infant trial on the WHO 3+0 schedule in the Philippines, with 421 infants (NCT06790290) ¹⁸. Results have not been published.

What it means

The core finding is consistent across South Africa, Mozambique and global surveillance: PCV removes most vaccine-type disease in young children, and it protects unvaccinated adults too ²³⁵. The largest remaining gain is coverage. Bringing PCV up to DTP coverage is worth an estimated 146,000 lives ¹.

For health ministries. The cost frontier is now clearer:

• A PCV13 1+1 schedule held to age 5 in a mature programme ⁶, and 1p+1 matched 3p+0 in the Gambia's preliminary results ¹⁰.
• A PCV10 1+1 schedule did not hold to age 5, so product choice matters ⁶.
• Fractional-dose Pneumosil campaigns cut carriage quickly in Niger ⁷.
• WHO now sets clear entry criteria for these options: 80% coverage for 5 years, a strong second-year visit, and surveillance to watch the results ¹⁰. Countries below that line should put coverage first.

For funders. The strongest case is still reaching the children who are missed. Low-cost products have brought the Gavi price below US$2 per dose for some products ¹³. Support for serotype surveillance, such as the NICD network in South Africa and the INS sentinel system in Mozambique, is what lets countries switch schedules safely ²³¹⁰.

For clinicians and health workers. Completing the schedule counts: three doses outperformed two against pneumonia in Mozambique ⁴. Check that children receive the booster or final dose on time.

The caveats set the next agenda. Serotype replacement and persistent carriage in older children, seen in the Gambia, mean that surveillance must continue, and that boosters, catch-up campaigns and broader-valency products deserve attention ⁵⁹.