What keeps a good idea alive in a public health system?

Many public health practitioners will recognise this pattern. A pilot succeeds. The numbers improve, the district team is proud, the report is well received, and a presentation travels from one meeting to the next. Then, a year or two later, the work has quietly stopped. Nothing went wrong, exactly. The project ended, the staff moved on, and the routine of the system closed over the space where the innovation had been.

A successful pilot is a milestone, not proof. It shows that an idea can work under particular conditions, often with extra people, extra attention, extra money and a team that cares. What it does not show is whether the idea can keep delivering its benefits once those conditions are gone, even if its design changes along the way. That is what I mean by sustainability here, and it is a different question, answered by different things.

In my experience, several organisational and institutional factors can strongly influence whether an innovation becomes part of routine practice.

The first is ownership, whether someone inside the system, an officer, a department or a programme, regards the work as theirs rather than as a partner’s project they are hosting. The second is money in the right place, whether the cost has found its way into a government budget line, however small, rather than sitting entirely in a project budget. The third is fit with routine work, whether the innovation has become part of how a frontline worker or a district manager already does their job, or remains an extra task alongside it.

The fourth is how the idea sits with the rules. An innovation that fits existing guidelines, staff roles and procurement arrangements may be easier to build into routine work. One that needs a new task for a cadre whose role is defined elsewhere, a payment no scheme provides for, or a device no standard yet covers, may need further approvals or policy changes. The permissions that let a pilot run do not always provide a lasting basis for routine use.

The fifth is boundaries, between departments, between levels of government, and between people. Many good ideas in public health need more than one department to work together, such as health and women and child development, health and labour, or a state and a private estate. Convergence sounds simple in a meeting and is harder in practice, because each department has its own budget, targets, reporting lines and timelines. An innovation that depends on continued cooperation between departments is more vulnerable when that cooperation is not built into how both of them work.

Running through all of it is bureaucracy, in the neutral sense of the word, meaning files, approvals, transfers and the annual planning cycle. An officer who championed a pilot may be transferred before it ends. A proposal that misses the window for the annual programme plan may wait a year. These processes serve important purposes, including accountability and continuity, although they can also create delays. An innovation that is not designed to work through them may struggle to last.

None of these factors is enough on its own. Ownership matters, but it needs to come with adequate resources, the capacity to implement, continuing relevance and a willingness to adapt. And these factors are not always captured in a pilot’s headline results.

In 2025, the disruption to US foreign assistance brought questions of continuity into sharper focus for many people working in global health. In January, a United States executive order titled Reevaluating and Realigning United States Foreign Aid paused foreign development assistance for review, and in the months that followed, many colleagues saw the programmes they worked on come to an end. In the work I observed, activities that were already part of government protocols, systems and routine responsibilities seemed better placed to continue. That is an observation, not a systematic comparison, and continuity depended on many factors beyond ownership alone.

The project I was part of, USAID’s Saksham, was among the projects that continued, recognised as life-saving work. I cannot attribute that to any single factor. But it reinforced my sense that work woven into government processes, rather than sitting beside them, is better placed to continue when funding arrangements change.

The lesson I take from it is not about any one funder. Funding of every kind, from donors, philanthropies or government schemes, changes over time. An innovation that depends on one source of support is more fragile, however good its results. So a question worth asking at the start of a pilot is not only “will this work?” but also “whose will this be when we leave?”

A pilot that answers the first question earns the right to be taken seriously. The second question helps decide whether its benefits can last.

Views are my own.