Why Good Vaccines Still Miss Children, and What Implementation Science Has To Do With It

Let's start with something that should be good news, but somehow keeps turning into a puzzle. We have vaccines that work. We've had them, in some cases, for decades. They are among the most cost-effective things humanity has ever invented. And yet, in 2024, around 14.3 million children still received not a single dose of any routine vaccine.

If you work anywhere near an immunization programme, you already feel the weight of that number. It isn't a science problem. Nobody is standing in a district health office wondering whether the pentavalent vaccine works. We know it works. The question that keeps people up at night is quieter and far more frustrating. We know what to do, so why isn't it happening?

That gap between what we know and what actually gets done has a name. People call it the "know-do gap," and closing it is more or less the entire job of a field called implementation science. This article is a friendly introduction to both: the gap, and the discipline that exists to close it. No jargon walls, I promise. Just the ideas, and why they matter for anyone trying to get vaccines to the children who need them.

The thing nobody tells you in vaccine training

When most of us first learn about immunization, we learn the science. Cold chain. Schedules. Antigens. Doses and intervals. All essential, all important. But here's the part that often goes unsaid: knowing the schedule perfectly does almost nothing to guarantee a child in a hard-to-reach community actually gets vaccinated.

Picture a district. I'm going to keep this hypothetical, because the truth is some version of it exists almost everywhere. The vaccines are in the fridge. The health workers are trained. The national schedule is clear. On paper, everything is in place. And yet coverage in one particular ward stubbornly refuses to climb past 50%.

Why? Maybe the outreach sessions keep getting cancelled because the fuel budget ran dry. Maybe the one motorbike that serves three villages broke down in March and nobody has fixed it. Maybe mothers are willing but the clinic only vaccinates on Tuesdays, and Tuesday is market day. Maybe the data the district sees is three months old, so by the time anyone notices the dip, a whole cohort of infants has aged out.

None of those are vaccine problems. Every single one is an implementation problem. And that distinction is the whole point.

So what exactly is implementation science?

Here's the least intimidating definition I can offer. Implementation science is the study of how to take something that works in theory and make it actually work in the real world, reliably, at scale, in messy and very different contexts.

The more formal way people describe it is as the scientific study of methods to promote the uptake of proven interventions into routine practice. Notice the word proven. Implementation science isn't about discovering new vaccines or new treatments. It assumes you already have something that works. Its obsession is the next, harder question: how do you get it delivered, used, and sustained, especially in the places where delivery is hardest?

If classical medical research asks "does this intervention work?", implementation science asks "now that we know it works, why isn't it reaching everyone, and what can we actually do about that?" That's a different question entirely. And for immunization, it's often the more urgent one.

Why immunization is almost a perfect example

Of all the areas where this gap shows up, immunization might be the clearest illustration you could ask for. Think about it. The product is proven beyond serious doubt. Routine vaccination is credited with preventing an enormous number of deaths globally. By one widely cited estimate, more than 150 million over the past five decades. The cost-effectiveness is extraordinary. The political will, at least on paper, broadly exists.

And still, coverage stalls. Globally, the share of infants getting the third dose of the diphtheria-tetanus-pertussis vaccine, the number everyone watches as a proxy for a functioning system, has hovered around 85% without moving much, while the number of completely unvaccinated "zero-dose" children remains higher than it was back in 2019. We are, in places, going backwards.

When a product this good still can't reach everyone, you've ruled out the product. What's left is everything around it: the planning, the logistics, the data, the people, the trust, the hundred small operational decisions that determine whether a vaccination session happens on a given Tuesday in a given village. That "everything around it" is exactly the territory implementation science was built to map.

The "zero-dose child," a whole strategy in two words

If you spend any time in global immunization conversations lately, you'll hear the phrase "zero-dose children" constantly. It's worth understanding, because it's a beautiful little example of implementation thinking baked right into the language.

A zero-dose child is, in the operational definition used globally, a child who hasn't received even the first dose of the DTP-containing vaccine. Why measure it that way? Because that first dose is usually a child's first real contact with the routine immunization system. If a child never even gets that, it's a strong signal that the system isn't reaching them at all. Not that a single vaccine was missed, but that an entire family is likely missing from health services more broadly.

So "zero-dose" isn't really about one vaccine. It's a flag for the most excluded, hardest-to-reach families, and a way of forcing programmes to ask the implementation question directly: who are we systematically failing to reach, and why? Reducing the number of zero-dose children is now a central equity goal of the global Immunization Agenda 2030, with the aim of cutting their number by half from 2019 levels by the end of the decade.

That's implementation science wearing a name tag. The whole concept exists to point programmes at their delivery failures rather than their product.

What this looks like when it actually works

Let me give you a real, established example: the "Reaching Every District" approach, usually shortened to RED. It was developed by WHO, UNICEF and partners back in the early 2000s precisely to tackle this kind of problem, and it's a lovely case of implementation thinking turned into practical routine.

RED doesn't introduce a new vaccine. It reorganizes how the existing ones get delivered, around five fairly humble-sounding components:

• Re-establishing regular outreach. Making sure services actually reach communities beyond the clinic walls, on a predictable schedule.
• Supportive supervision. Supervisors who coach and problem-solve with health workers, rather than just inspecting and scolding.
• Community links with services. Building genuine, ongoing relationships between health workers and the communities they serve.
• Monitoring and using data for action. And the operative phrase there is for action, not data for filing in a drawer.
• Planning and managing resources. Including the unglamorous heroics of microplanning at district and facility level.

Notice what's on that list. Outreach logistics. Supervision style. Community trust. Using your own data. Planning. Not one of those is a medical breakthrough. Every one of them is about how the work gets done. And in many African settings, getting these basics right has been credited with meaningful, sustained gains in coverage over the years. That's the quiet power of the implementation lens. It takes the boring operational stuff seriously, because the boring operational stuff is usually where coverage lives or dies.

Why this matters for you, specifically

Here's the part I most want to land. You do not need to be a researcher to think like an implementation scientist. In fact, the people best placed to use this lens are usually the ones closest to the work: the district officer, the programme manager, the supervisor, the health worker who already knows in their gut which ward is being left behind and probably has a few theories about why.

Implementation science gives those instincts a structure. It offers a way to ask sharper questions. Where exactly is the breakdown: access, or trust, or supply, or scheduling? What's actually driving it in this context, which may be completely different from the district next door? What small, locally feasible change might shift it? And, crucially, did the change we tried actually work, or did we just feel busy?

That last question matters more than it sounds. A great deal of effort in immunization goes into activities that feel productive but never get checked against results. The implementation mindset is relentless about that loop. Try something, watch what happens, adapt. It treats the district itself as a place where useful knowledge gets made, not just a place where instructions from above get carried out.

Where to go from here

If any of this resonated, if you've ever sat in a review meeting staring at a coverage figure that refuses to budge and thought we're doing everything we're supposed to, so what's going on?, then you've already had the first implementation science thought. The rest is just learning the tools to act on it more systematically.

That's exactly what Achieve IS is built for. Our courses, games, and simulations take these ideas out of the textbook and put them into scenarios that feel like the work you actually do: the broken motorbike, the market-day clinic, the three-month-old data. Because reading about implementation is useful, but practising it is where it starts to change how you work.

The gap between knowing and doing is real. But it's not a mystery, and it's certainly not permanent. It's a set of problems, specific, local, and far more solvable than they feel at 4pm on a hard day. Implementation science is simply the toolkit for solving them, one district at a time.