Coaching for quality improvement under performance-based contracting: a theory-of-change evaluation in Honduras

In many parts of the world, the challenge of healthcare is no longer just about building clinics or training doctors to show up. The harder problem is ensuring that once a patient walks through the door, the care they receive is actually good. For decades, health officials have tried to fix this by setting strict targets and paying providers only if they hit those numbers, a strategy that treats medical staff like employees who need constant monitoring to do their jobs. Another approach, known as coaching, tries something different: it treats teams as capable professionals who can solve their own problems if given the right tools, time, and encouragement. This method relies on trust and learning rather than fear and punishment. The big question for global health is whether this supportive style of management can actually work in places where the system is built on strict rules and penalties. If a health system is designed to punish mistakes, can a program designed to encourage learning survive inside it?

Researchers set out to answer this question by studying a specific health program in Honduras, a country where poverty is widespread and the healthcare system is a patchwork of different organizations working under one national government. During a three-year period, a team of international experts introduced a coaching program to local health managers and their staff. The goal was to improve how doctors and nurses handled two critical services: checking for cervical cancer and providing care for pregnant women. The coaches did not give orders or inspect work for errors. Instead, they taught managers how to use data to spot problems, how to talk to their teams without blaming them, and how to create a culture where staff felt safe to admit mistakes and try new solutions. The researchers wanted to understand not just whether the numbers improved, but how the people involved actually changed their behavior and why.

To find the answers, the research team spent months listening to the people who lived this experience every day. They interviewed nineteen key figures, including regional health directors, mid-level managers who supervised multiple clinics, and the medical leads who ran the day-to-day operations. They also watched the final training sessions and reviewed official documents. By piecing together these stories, the team reconstructed the "theory of change"—the map of how the program was supposed to work—and tested it against what actually happened on the ground. They found that the coaching did indeed spark real change, but the results were complex and depended heavily on the environment in which the teams were working.

The first major change the researchers observed was a shift in how managers saw themselves. Before the program, many managers described their role as that of a commander who gave orders and demanded obedience. They felt their training had prepared them to be bosses, not leaders. The coaching helped them realize they could be facilitators instead. One manager described the difference as moving from telling a team "do this" to saying "let's do this together." This shift was not just about being nicer; it was about building a new professional identity where the manager felt responsible for helping the team succeed rather than just policing them. This change in mindset was the foundation for everything that followed.

Once managers stopped acting like enforcers, a second powerful pattern emerged: teams began to trust the data. In the past, numbers on a chart were often just cold statistics used to punish clinics that missed targets. With the new coaching approach, managers and staff started using data to learn. They would measure a problem, try a solution, and then measure again to see if it worked. When they saw their own efforts lead to visible progress, they felt a sense of shared accomplishment. This was particularly clear in the work to screen for cervical cancer. Because the steps for this screening were clear and the results were easy to count, teams could see their success immediately. They felt capable and motivated to keep improving. This cycle of action, measurement, and success created a strong sense of collective confidence.

However, this same cycle of confidence did not work as well for prenatal care. The researchers found that the process of caring for pregnant women is much more complicated and harder to track with simple numbers. The steps are longer, and the results are less immediate. When teams tried to apply the same data-driven learning to prenatal care, they often hit a wall. They could not easily see how their specific actions led to better outcomes, so the feedback loop that had fueled their confidence for cancer screening broke down. Without that clear connection between effort and result, the teams struggled to maintain the same level of motivation. This suggested that coaching works best when the work being done produces clear, observable results that teams can claim as their own.

The third piece of the puzzle was the human connection. The coaching program created a space where staff felt safe to speak up without fear of being yelled at or punished. Managers learned to listen and support rather than blame. This psychological safety allowed teams to work together more effectively. When a problem arose, they could discuss it openly, knowing their supervisor was there to help find a solution, not to point a finger. This supportive atmosphere was the glue that held the other changes together, allowing the new identity of the manager and the new use of data to take root.

Yet, despite these positive changes, the researchers discovered a fourth, unanticipated pattern that limited the success of the program. The coaching was trying to build a supportive, learning-based environment, but it was operating inside a health system that was still run on strict, punitive rules. The national government continued to audit clinics and withhold money if targets were missed. This created a deep tension. The managers and staff felt caught between two worlds: one where they were encouraged to learn and improve, and another where they were terrified of losing their funding if they made a mistake. The researchers described this as an "enabling bubble." The coaching created a small, safe space where real improvement happened, but that space was surrounded by a harsh reality that threatened to pop it. The gains were real, but they were trapped in those local bubbles, unable to spread widely because the broader system was not aligned with the new way of working.

The study also highlighted a stark difference in how different levels of the system perceived the situation. The regional directors who oversaw the program believed that strict monitoring and clear targets were necessary to keep people working hard. They saw the audits as a legitimate way to ensure resources were used correctly. In contrast, the frontline managers and staff experienced these same audits as demoralizing and confusing. They felt the auditors would show up, list problems, and leave without ever returning to help fix them. This gap in understanding meant that the very people who needed to be supported were often the ones feeling the most pressure, while the people in charge of the system remained largely unaware of how their methods were undermining the coaching efforts.

The researchers concluded that while coaching can successfully transform how health teams work, it cannot succeed in isolation. In Honduras, the program proved that managers could change their style, that teams could learn to use data, and that trust could be rebuilt. But these changes were fragile because the wider system was still built on a different logic. The study suggests that for coaching to have a lasting impact, the entire system needs to change. The rules for funding and monitoring must be aligned with the goal of learning, rather than just punishing failure. Without this alignment, the positive changes will remain confined to small pockets of hope, unable to transform the system as a whole. The paper does not claim that coaching failed; rather, it shows that coaching is a powerful tool that needs the right environment to work, and without that environment, its potential remains limited.