Policy vs Reality: Why tracking antibiotic resistance is so hard

By Emily O’Neill

Antimicrobial resistance (AMR) has become one of the defining global health challenges of our century. Governments, international agencies, and research funders have invested in tackling AMR through national policies, stewardship programmes, surveillance systems, and drug development incentives. As we mark World Antimicrobial Awareness Week, the fundamental question remains: Do these actions work?

Why measuring AMR progress is so difficult

Evaluating efforts to mitigate AMR is challenging for several reasons. Like many public health interventions, outcomes often lag behind action, but in the case of AMR, this delay is particularly pronounced. Changes in bacterial resistance can take years, even decades, to manifest, and resistance rates can be further influenced by a number of factors, including differences in reporting and surveillance, trade and travel, and even climate. As a result, drawing a straight line between a policy intervention and a measurable change in resistance rates is rarely possible. Many times, policymakers are left to evaluate a moving target, trying to measure progress in a system where biological, behavioural, and social factors interact in complex, time-dependent ways.

The challenge of measuring AMR outcomes is further compounded by the difficulty of defining the "right" outcomes. Should reduced antibiotic consumption be seen as success, or might higher use in particular contexts reflect improved access to essential, life-saving treatments? Should we celebrate lower observed bacterial resistance, or could higher rates indicate improved surveillance? Do fewer documented infections signal better prevention and control, or simply more transparent reporting? Each metric can tell a different story, and what looks like progress in reports may not translate to actual impact.

The metrics dilemma and what counts as success?

Current assessments often focus narrowly on inputs such as laws passed, surveillance systems established, and guidelines issued, rather than the outcomes of those actions. Even when antibiotic use is tracked, metrics vary widely, and only a small number of evaluations are embedded within One Health frameworks that capture the interconnected dynamics of human, animal, and environmental health. Surveillance networks within a country may be mandated by policy, but operate at only a handful of sample sites, producing patchy or non-comparable data. In some cases, policies within countries are launched without the laboratory capacity, regulatory enforcement, or information systems needed to track progress. So, on paper, an achievement was made, yet in real life, the problem was never addressed.

Generating objective evidence to guide AMR policy may seem daunting, but in a recent BMJ Global Health study, myself and colleagues Jinru Wei, Sara Machado, Omar Galarraga and Irene Papanicolas illustrate that more rigorous quantitative analysis is possible. In this study we linked the quality of 68 countries’ AMR national action plans to antibiotic sales data over time. We found that higher-quality plans were associated with measurable improvements in antibiotic use, particularly in using first-line antibiotics rather than potentially harmful broader-spectrum antibiotics. Complementing this, a Lancet Global Health analysis of pharmaceutical sales and resistance data from over 70 countries explored the link between antibiotic use and regional patterns of bacterial resistance. The authors found that regions using more antibiotics tended to have higher levels of drug-resistant bacteria. Together, these studies illustrate that quantitative evaluations can begin to map the real-world patterns in policy, antibiotic use, and resistance, helping to identify effective strategies and guide future investment and action.

From reporting to learning: building an evidence-based response

While these studies do not solve the challenge of defining and measuring outcomes, they represent an essential first step toward evidence-based analysis and learning in the global AMR response.

If AMR is to be managed, governments, global health institutions, researchers and the like must treat evaluation as a tool for learning, not just reporting. Success should not be measured by the number of policies written or meetings convened, but by how clearly, we define positive health outcomes and by how we measure the impact of interventions against them. Achieving this is no easy task, but concrete steps can and must be taken.

The urgent need for stronger evidence-based oversight has never been more apparent. As World Antimicrobial Awareness Week reminds us, AMR is a slow-moving crisis – but policy inertia can accelerate it. The first few glimpses of evidence-based research indicate that well-designed policies can work, but only if they are implemented effectively, monitored rigorously, and continuously refined. The global community must commit to learning from evidence, adapting strategies, and holding itself accountable to ensure evidence finally drives global policy rather than merely being discussed.