Skip to content
A nurse drawing a vaccine dose from a vial into a syringe
Yellow fever · 2024

Yellow fever and vaccine coverage

Does low vaccine coverage put a state at higher risk of a yellow fever outbreak?

The background

Yellow fever is a viral disease spread by mosquitoes. There's a safe, effective vaccine, and the World Health Organization says at least 80% of people need to be vaccinated to keep outbreaks away.

In Nigeria, coverage is very uneven from one state to the next. I wanted to see what that gap means in practice: how much more likely is an outbreak where fewer people are protected?

What I did

I built a mathematical model that splits people into groups: those who could catch the disease, those who are vaccinated, those who are infected and those who have recovered. A second part of the model does the same for mosquitoes.

I fed it three years of real vaccination data from DHIS2, Nigeria's national health information system, plus population figures from the National Bureau of Statistics. Then I compared Bayelsa, where coverage was low, with Jigawa, where it was high.

For each state I worked out R₀, the number of people one case is likely to infect, and tested which parts of the model mattered most.

The model

How the SVIR-SI model is built

People and mosquitoes each move through their own set of boxes. Mosquito bites link the two.

People
Susceptible
Vaccinated
Infected
Recovered
Mosquitoes
Susceptible
Infected
What I found

The numbers that matter

33.1%

average vaccine coverage in Bayelsa

4.1

R₀ in Bayelsa: one case could infect about four people

84.1%

average vaccine coverage in Jigawa

2.6

R₀ in Jigawa: lower, but still above 1

In Bayelsa, infections were trending upwards and the risk of spread was high.

In Jigawa, infections were trending downwards. But even with coverage above the WHO target, R₀ stayed above 1, so the disease could still spread if a case came in.

Why it matters

Bayelsa needs an urgent push to raise coverage from about 33% to above 80%, along with stronger surveillance to catch cases early.

Jigawa shows that hitting 80% isn't always the finish line. My recommendation was to aim higher, towards 95%, and keep surveillance strong.

From the original work

Slides and figures

These come straight from my presentations and model outputs. Tap any one to see it bigger.

Work with me

Got data that needs a second look?

I'm open to consulting work, research partnerships and full-time roles. Send me an email and tell me what you're working on.