The Cost of Health Literacy: What this Means for the Future of Digital Health

The WHO recently published a report on health literacy aiming to provide new context for understanding the economic implications of low health literacy. Data from population surveys across multiple countries, shows that between 34% and 79% of adults lack adequate health literacy. In the UK this has been estimated at 43% and this increases to 60% when numbers are involved. An estimated 7.1 million adults in the UK read at or below the level of an average 9-year-old.

Health literacy refers to people's ability to access, understand, evaluate and apply health information.

Research shows consistent demographic patterns in health literacy which follows a clear social gradient shaped by age, education and socioeconomic status, with older adults and those with lower income or education levels having lower literacy. Groups facing disadvantage, including people with language barriers, learning disabilities and dementia, are more likely to have limited literacy.

The evidence demonstrates that health outcomes and health service use are strongly correlated to literacy levels. Low literacy drives higher system costs, estimating that 3 to 5 % of annual healthcare expenditure is attributable to low health literacy. At the individual level, patients with limited health literacy incur $143 to $7,798 in additional yearly costs compared to those with adequate health literacy. In the UK it is thought that low health literacy accounts for 5% of national health spending.

Those with low health literacy generally have more unhealthy lifestyles and overall worse health. Low health literacy also drives greater use of health services, having more hospital admissions, more visits to A&E and more likely to be readmitted within 90 days of discharge. Low health literacy also correlates with poorer self-reported health, higher rates of multimorbidity and lower engagement with preventive services.

This evidence presents health system leadership with an avoidable source of system inefficiency. Health systems that assess communication effectiveness, simplify processes and train staff in clear communication can achieve better adherence, more predictable demand and lower administrative burden.

These cost estimates reflect traditional health literacy: the ability to read prescription labels, understand discharge instructions, and navigate appointment systems. However, we know that healthcare delivery has shifted substantially toward digital channels. The widespread adoption of telehealth, online portals and remote monitoring has increased the cognitive and technical demands placed on patients and professionals.

This shift introduces additional complexity. Digital health literacy refers to the ability to find, understand and apply health information in digital environments. It is distinct from general digital skills and has been described as a "super determinant of health" distinct from both traditional health literacy and general digital skills.

Limited digital health literacy contributes to inequitable access, poorer adherence and suboptimal health behaviours, particularly as digital tools become the main route to access services and information. A 2024 systematic review in the European Journal of Public Health found that low digital health literacy independently predicts poorer self-management of chronic disease and lower uptake of preventive care, even when traditional literacy is accounted for.

Population-level data identifies the same disparities seen in traditional health literacy: older adults, people with lower incomes and those with less formal education consistently record lower digital health literacy scores. Evidence also shows that members of healthcare teams often lack the digital confidence and skills required to support patients effectively, creating gaps precisely where assistance is most needed for patient to navigate digital health systems.

Artificial intelligence introduces a further layer of literacy requirements in health care because its outputs and recommendations increasingly inform both clinical decisions and consumer health information as AI tools are increasingly embedded in clinical workflows and public-facing health applications.

Their use introduces new requirements for comprehension for both professionals and patients. Patients need to evaluate AI-generated health information and understand when to question it, not just access it. Data from the Canadian Digital Health Survey found that only 7.8% of adults describe themselves as "very knowledgeable" about AI, with 42.3% reporting moderate knowledge.

Current research on AI in health highlights risks such as inaccuracies, and the amplification of misinformation when AI outputs are not properly validated, which can mislead people who lack the skills to evaluate their reliability. Low AI literacy makes it difficult for users, both patients and practitioners, to evaluate when and how to question AI-generated recommendations. As a result, gaps in AI health literacy can affect how people interpret diagnostic suggestions, assess the trustworthiness of information, and maintain informed consent in situations where AI tools play a direct role in care.

Healthcare professionals also face gaps in AI literacy: clinicians and students commonly lack structured training in how AI tools work, how to interpret algorithmic outputs, and how to discuss them with patients, which can compromise shared decision-making and safe use.

WHO's report highlights economic analysis suggesting health literacy interventions generate financial returns. Economic modelling projects that reducing unnecessary emergency department visits by patients with low health literacy to the level of those with adequate health literacy could save $47 billion annually in the United States. The Better Beginnings, Better Futures program in Ontario demonstrated $2.30 saved for every dollar spent on school-based health literacy programming.

The following ten considerations outline practical steps to embed greater health, digital and AI literacy within strategy, service design and workforce development.

1. Prioritise literacy as a strategic capability

The WHO Regional Digital Health Action Plan for Europe 2023–2030 identifies digital health literacy as a core enabler of equitable digital transformation. Yet only 17 of 53 European countries have formal strategies, education plans or policies addressing it. Health systems should treat literacy, traditional, digital and AI, as system capability, not optional support.

2. Embed literacy indicators in performance frameworks

Measurement of literacy should sit alongside other quality and equity indicators. Validated population tools such as HLS19 and HLS19-DIGI enable benchmarking across different geographies and over time. Routine data collection allows systems to identify gaps, implement solutions and monitor progress.

3. Expand validated assessment tools

Current assessments focus largely on traditional literacy. Instruments such as the Digital Health Readiness Questionnaire (DHRQ) provide structured ways to assess digital health literacy and readiness. Integration of such measures in clinical and community settings can make literacy visible within transformation programmes.

4. Integrate literacy into digital design and governance

Digital and AI systems should be evaluated for usability, comprehension and equity outcomes, not only for technical accuracy. Literacy must be considered at project inception, embedded in procurement, service design and testing, rather than added after deployment.

5. Build trustworthy AI by design

AI introduces unique literacy challenges. Design processes should explicitly consider users with differing levels of digital and AI competence. The FUTURE-AI framework for trustworthy and deployable AI in healthcare outlines 30 recommendations, including collaboration among policymakers, providers and developers, and continuous bias and risk assessment.

6. Strengthen workforce capability

Clinicians, navigators and non-clinical staff require training to recognise and support diverse literacy needs. This includes communication skills, awareness of digital barriers and confidence in explaining algorithmic tools. Workforce development should address both use of digital tools and support for patients using them.

7. Invest in infrastructure and equitable access

Improving digital health literacy depends on more than individual skills. Reliable connectivity, access to devices and culturally appropriate content are prerequisites for meaningful participation.

8. Maintain assisted and non-digital pathways

Rapid digitalisation risks deepening inequalities for people with disabilities, language barriers or limited confidence. Health systems should maintain blended access routes, digital, telephone and face-to-face, rather than assuming universal adoption.

9. Address the AI literacy gap

AI implementation requires targeted education for both patients and professionals. Studies show that clear explanation of how AI systems work, transparency in recommendations and maintained human oversight significantly increase trust and safe use. Every AI deployment should be accompanied by a literacy and communication plan.

10. Evaluate digital health equity, not just adoption

Success metrics must extend beyond uptake rates. Regular assessment of digital health equity outcomes, including access, comprehension and benefit across demographic groups, ensures that technology improves, rather than widens, health outcomes.

Low health literacy has substantial economic and clinical consequences. Digital transformation has increased the complexity of navigating care, and AI is adding new layers of comprehension that most systems and populations are not yet equipped to meet.

Digital literacy and access are now recognised as determinants of health in their own right Digital and AI technologies can improve access, quality and efficiency, but only if people can understand and use them safely. There is a human cost of low literacy as it drives preventable illness and avoidable barriers to care. The economic argument for investing in literacy is now well established and interventions to improve literacy have been shown to have a strong return on investment.

We need to design and scale interventions that address the full spectrum of literacy: traditional, digital and AI, within routine service delivery. Realising those benefits at population scale depends on embedding literacy as a design principle, not an afterthought.

Health systems therefore face a strategic choice. Continuing to expand digital and AI tools without addressing literacy will deepen inequity. Treating literacy as a determinant of health and building it into policy, procurement and workforce planning offers a path toward more equitable, trusted and sustainable healthcare.