Conceptualising undervaccinated populations in high-income settings: why consensus and clarity matter

Routine immunisation coverage across high income countries is challenged by ongoing decline/attrition and persistent disparities.¹ Terminology to describe groups with suboptimal vaccination rates has evolved from “hard-to-reach,” “marginalised,” and “disadvantaged” to “underserved.” Despite differing implicit assumptions, these terms are often used interchangeably, obscuring their distinct biases. Clarity in terminology is important for shaping programme delivery and outreach methods in ways that avoid stigmatising communities. The term hard-to-reach has been used since the 1980s to describe marginalised communities (by ethnicity, region, or socioeconomic status) seen as less responsive to public health campaigns.² It implies community resistance, though many experience services as inaccessible or perceive agencies as “hardly listening,” reinforcing existing marginalisation.³ In contrast, underserved shifts responsibility to the system for failing to meet population needs, highlighting that one-size-fits-all delivery does not ensure equitable access.

The language shift from hard-to-reach to underserved communities was intended to move the onus from the individuals onto the health system.⁴ However, the attempt to eliminate wording that attaches blame to certain groups, and the subsequent focus on equity in service delivery and system limitations,⁵ has arguably gone beyond removing the responsibility from communities. While underserved usefully avoids blaming groups and highlights the need for equitable delivery, it can also diminish community agency in shared health responsibility. By focussing on service delivery, the term may overlook the role of individual decision-making in vaccination outcomes.⁶

Reducing vaccination disparities to either individual traits or service flaws overlooks the complex, multifactorial nature of vaccination decisions and hinders solutions that address system-, population-, and individual-level barriers. Attributing undervaccination solely to health system failures can further discourage community-led vaccine initiatives and overlook individual agency, homogenising so-called underserved groups.¹ This terminology creates a one–way relationship between the health system and those it serves, reinforcing rigid links between vaccination status and identity markers such as ethnicity, socioeconomic status, or residence. In other words, when public health professionals attribute non-vaccination to group membership alone, they overlook that individuals make vaccine decisions as complex, multidimensional beings shaped by more than their interactions with the health system.⁷

Similarly to the terms that came before it, underserved is a value-laden term, which assumes that all undervaccinated individuals are not vaccinated as a result of a shortcoming of the health system. It positions the shortcomings of health services as the primary, if not sole, contributor to disparities in vaccination. As a result, crucial elements such as vaccine confidence, proliferation of misinformation, health literacy, self-efficacy, and institutional trust are not captured. While it is essential to continuously evaluate public health programmes to ensure they are optimally designed to deliver vaccines to the populations it is responsible for, it is equally important to recognise the agency of individuals and populations and allow for non-binary individual vaccination behaviour.

We suggest abandoning terminology that highlights only one of the numerous drivers of vaccination disparities.^8,9 Instead, we call for adoption of more outcome-oriented phrasing. During the UNIVACC 2025 conference at Karolinska Institutet in Sweden, the conceptualisation of underserved groups was debated from many perspectives. We note that terms such as underprotected and underimmunised have been put forward following the same rationale. These terms however implicate processes dependent on physiology such as individual-level immunogenicity, waning, or vaccine failures, that cannot be routinely measured or controlled by vaccination programmes. We propose the term undervaccinated to describe individuals or groups in high-income settings who are un- or partly vaccinated (see Fig. 1). The term focuses on vaccination status without assuming the relative influence of individual or systemic factors. Using consistent terminology such as undervaccinated helps public health agencies and healthcare services align language, improving clarity, comparability, and community engagement. As a value-free term, it makes no assumptions about the causes of non-vaccination and is thus applicable across contexts where factors, from access barriers to confidence, vary. Care and consistency in language can nurture more effective community partnerships, and hold broader relevance as evidence shows declining vaccine confidence across low-, middle-, and high-income countries.

Fig. 1.

The term undervaccinated refers to those who have not received all doses of a particular vaccine they are eligible for according to local recommendations, without attributing this outcome to a particular cause. As such, it encompasses, but is not restricted to, the sentiments embedded in more value-laden terminology, while recognising that the possible causes of under-vaccination are neither exhaustive nor mutually exclusive.

Contributors

Conceptualisation: All authors (Sibylle Herzig van Wees, Ben Kasstan-Dabush, Michael J. Deml, Ifrah Mohamed Ibrahim, Sandra Mounier-Jack, and Michael Edelstein) jointly conceptualised the idea during a conference discussion. Writing Original Draft: Sibylle Herzig van Wees, Ben Kasstan-Dabush, and Michael J. Deml led the drafting of the manuscript. Writing Review & Editing: All authors contributed to reviewing, refining, and approving the final manuscript. Verification of Data: As this is a Comment piece without original data collection, all authors verified the accuracy of the arguments and sources cited. Final Approval: All authors reviewed and approved the final submitted version.

Declaration of interests

The authors report no conflicts of interest related to this work.