Rolling out the first malaria vaccine meant building Africa's safety-reporting systems

When the world's first malaria vaccine, RTS,S/AS01, was rolled out in Ghana, Kenya and Malawi from 2019, the harder problem was not delivering the doses but detecting whether they were safe — because the systems that flag a vaccine side-effect barely existed in the districts doing the rollout. A programme built alongside the pilot trained roughly 6,000 health workers to recognise and report adverse events, and by its own account captured 90% of the safety signals it was looking for through dedicated studies rather than routine reporting [s2].

The context is a genuine public-health advance. The WHO-coordinated Malaria Vaccine Implementation Programme reached more than 2 million children across the three countries, and WHO recommended RTS,S for broad use in October 2021 on the strength of it [s1]. The pilot's evaluation found a vaccine-attributable 13% drop in mortality among children old enough to be vaccinated, alongside a substantial reduction in hospitalisations for severe malaria [s1]. Both RTS,S and the newer R21 vaccine cut malaria cases by more than 50% in the first year after vaccination, and by about 75% when doses are timed to a highly seasonal transmission peak [s1]. (Health Newspapers has covered the mortality evaluation separately; this piece is about the safety machinery underneath it.)

Why safety monitoring was the bottleneck

Pharmacovigilance — the routine detection, reporting and investigation of adverse events following immunisation — is the invisible infrastructure of any vaccination programme. In much of sub-Saharan Africa it is thin, and a brand-new vaccine raised the stakes: regulators and the public needed confidence that any adverse event of special interest would be caught and evaluated, not missed. To that end, several phase IV studies ran during the rollout, paired with a capacity-building push among frontline health workers [s2].

The training reached approximately 5,000 community healthcare professionals and 1,000 study staff, delivered mainly as tailored two-day in-person sessions for both medical and non-medical workers, in hospitals and in communities [s2]. It was backed by online courses, printed job aids, a tele-expertise platform, and a mobile alert system for reporting events [s2]. Knowledge assessments before and after showed clear gains: mean scores rose from 41% to 72% among medical professionals and from 38% to 68% among non-medical staff after the initial training, with similar improvements after refresher sessions (45% to 74% and 45% to 72%) [s2].

What the numbers do and do not show

The most consequential figure is that 90% of reported adverse events of special interest and adverse events following immunisation were documented through the phase IV studies [s2]. That is a measure of how much of the safety picture depended on purpose-built research rather than the routine national reporting system — a strength during a closely watched pilot, but also a flag. A signal-detection capacity that lives inside time-limited studies does not automatically transfer to a national programme once those studies end.

The authors are candid about the gaps. The digital tools built to speed reporting were underused, with workers often defaulting to paper [s2]. And a knowledge score is an input, not an outcome: a health worker scoring 72% on an assessment is better prepared to spot an adverse event, but the analysis does not establish that more real events were actually detected, reported and investigated as a result [s2]. The initiative's own framing is a set of lessons for the next rollout, not a claim that the problem is solved [s2].

Why it matters beyond malaria

The malaria vaccine is being introduced across Africa through Gavi-supported programmes, and R21 has expanded the supply that constrained RTS,S. Every one of those introductions inherits the same question the pilot exposed: can the receiving health system detect a safety signal on its own? The pilot's answer is that the capacity can be built quickly and cheaply through training, but that it defaults back to paper, concentrates inside studies, and has not been shown to survive the transition to routine practice [s2].

That is the thing to watch as the vaccine scales. The efficacy case for RTS,S and R21 is settled enough for WHO to recommend both [s1]. Whether the countries rolling them out can independently catch and act on the rare adverse event — the test that keeps public trust intact — is a separate question, and the pilot suggests it is one of investment in surveillance systems rather than of the vaccines themselves [s2].

This article is informational and is not medical advice.

Sources

• [s1] World Health Organization, "Questions and answers on RTS,S/AS01 malaria vaccine." https://www.who.int/news-room/questions-and-answers/item/q-a-on-rts-s-malaria-vaccine
• [s2] Drug Safety, "Enhancing Disease Surveillance and Pharmacovigilance Practices for RTS,S/AS01E Malaria Vaccine Rollout in Ghana, Kenya, and Malawi: Strategies, Impact, Challenges, and Lessons Learned," published 14 July 2026. https://doi.org/10.1007/s40264-026-01685-3