A patient is handed a discharge sheet. It lists five medicines, two with new doses, one to stop. There is a follow-up appointment “in 2/52.” There is a phone number for “the relevant department.” The patient says she understands, because that is what people say.
Three weeks later she is readmitted. Her notes record “poor adherence.”
Health literacy is usually framed as a property of patients: some people have it, some do not, and those who do not have worse outcomes. The evidence for that last part is strong. But the framing hides the more useful insight. Health literacy is not only a skill people carry around. It is the gap between what people can manage and what our systems demand of them. And we control one side of that gap.
The evidence that it matters
In 2011, Nancy Berkman and colleagues published an updated systematic review in the Annals of Internal Medicine, pulling together 96 good- or fair-quality studies. Low health literacy was consistently associated with more hospitalisations, greater use of emergency care, lower uptake of mammography screening and influenza vaccination, poorer ability to take medications appropriately and to interpret labels and health messages, and, among older people, poorer overall health and higher mortality.
That is a long list, and it runs in one direction. The review estimated that around 80 million American adults had limited health literacy at the time.
It is not a small minority
It is tempting to imagine low health literacy as a problem of a marginal few. The data does not support that.
Europe’s largest recent effort to measure it, the HLS19 survey run by the WHO Action Network on Measuring Population and Organizational Health Literacy, surveyed adults across 17 countries between 2019 and 2021. It found considerable proportions of the adult population with limited health literacy in every participating country, with a clear social gradient. In Norway, one of the wealthiest and most educated countries in the world, the national results showed 46 percent of adults with limited general health literacy.
If nearly half the population of Norway finds health information hard to find, understand, judge or use, the problem is not only the population.
The system is the harder test
Consider what a typical health system asks people to do. Navigate referral rules. Understand risk expressed as percentages. Read consent forms written by lawyers. Work out which of several clinics to call. Use patient portals designed by engineers. Keep track of medicines with names that sound alike.
Much of this is difficult for anyone. Clinicians who have been patients often say they found it bewildering. For people who are ill, frightened, tired, older, or working in a second language, the difficulty multiplies.
This is why the HLS19 survey measured “navigational” health literacy, the ability to find your way through the system, as a separate domain. Navigation is not a personal skill in the abstract. It is a measure of how hard the system makes it.
Why “educate the patient” disappoints
The standard response to low health literacy is patient education: leaflets, classes, apps. Some of it helps. But it rests on an assumption that the patient is the thing to be fixed.
There is an uncomfortable incentive behind that assumption. Educating patients is cheap and places the burden on them. Redesigning forms, simplifying pathways, training staff and giving clinicians enough time to check understanding costs more and requires institutions to change. So institutions produce leaflets.
And when outcomes stay poor, the record says “non-adherence,” locating the failure in the patient rather than in the instruction she was given.
What a health-literate organisation does
The more useful idea is organisational health literacy: the extent to which a service makes it easy for people to understand and act on health information. That shifts the question from “what is wrong with this patient?” to “what are we asking her to do, and is it reasonable?”
In practice, that looks like:
- Universal precautions. Assume anyone may struggle, and communicate clearly with everyone, rather than trying to identify who has low literacy.
- Teach-back. Ask patients to explain in their own words what they will do. It checks the clinician’s explanation, not the patient’s intelligence.
- Plain language by default. “Come back in two weeks” rather than “2/52.” “Kidney doctor” rather than “nephrology.”
- Fewer steps. Every extra form, phone call or referral is a place where people drop out.
- Test materials with users, including people with limited literacy, before they go into circulation.
None of these is new. What is still unusual is treating them as core quality measures rather than nice-to-haves.
Measure the organisation, not just the patient
If health literacy is partly a property of systems, it should be measured partly at that level. How many steps does it take to book a follow-up? What reading level are discharge instructions written at? How often do clinicians use teach-back? Are readmissions linked to medication confusion tracked as a system failure?
Few hospitals know these numbers. Most could find out quickly.
Walk the pathway with a stopwatch and a pen
One way to see how much literacy a system demands is to walk through a common pathway and count what it requires. Take, as an illustrative example, a patient newly diagnosed with type 2 diabetes.
In the first month, she may be asked to: understand what the diagnosis means and why it matters when she feels fine; take a new medicine, possibly two, at specific times; understand the difference between blood sugar readings and the longer-term test results; book a blood test, an eye screening appointment and a foot check, each through a different system; read a diet leaflet; use a glucose meter; interpret the numbers; know which symptoms mean she should seek help urgently; and navigate a patient portal to see her results.
Each step is reasonable. Together, they form a demanding curriculum, delivered in fragments by different people, often in technical language, to someone who has just received unwelcome news.
If she misses the eye screening or takes her medicine wrongly, the record will likely describe her as non-attending or non-adherent. It will not describe the pathway as overly complex. That is the asymmetry this piece is arguing against.
The strongest counterargument
There is a serious argument that emphasising system design lets patients off too lightly.
Adults are responsible for their own health, the argument goes. Health systems cannot be redesigned around the lowest common denominator, and simplifying everything risks withholding information that more capable patients want. Patient education has a track record: people can learn to manage complex conditions, and many do. Framing health literacy as a system problem risks paternalism, treating patients as unable to cope rather than supporting them to build skills.
This holds in part. Patient skills matter, and education can work, especially when it is practical and repeated. People with chronic conditions often become highly knowledgeable about them. And patients differ: some want detailed information, others want simple instructions.
But the argument misreads what system-focused approaches propose. Plain language does not mean less information. It means information that more people can use. Teach-back is not paternalistic. It checks whether the clinician’s explanation worked. And reducing unnecessary steps in a pathway does not remove patient responsibility. It removes obstacles that have nothing to do with health. Well-designed systems support patient skills rather than replacing them.
There is also a fairness point. When the demands of a system exceed what nearly half the population can comfortably manage, as the HLS19 data suggests in some countries, the burden of that design falls hardest on people who are older, poorer, less educated or unwell. Those are the people who need health services most.
Second-order effects
Digital health can widen gaps. Patient portals, apps and online booking promise convenience. For people with limited digital or health literacy, they can become a new barrier. The HLS19 survey measured digital health literacy separately for a reason.
Misinformation spreads more easily. People who find it hard to judge the reliability of health information are more exposed to misleading claims, a problem that became acute during the COVID-19 pandemic and has not gone away.
Shared decision-making becomes uneven. Shared decision-making assumes patients can weigh risks and benefits. Without support, it may benefit mainly those with higher literacy, while others default to whatever the clinician suggests.
Staff time is consumed. Confusing forms and pathways generate phone calls, missed appointments, repeat visits and complaints. Much of that workload is created by the system itself.
What this means for different decision-makers
For health ministries, health literacy can be written into quality standards and national strategies at the organisational level, with expectations for plain-language communication, simplified pathways and accessible digital services.
For hospital and clinic leaders, practical steps include testing key documents such as discharge summaries, consent forms and appointment letters with patients before use; setting targets for readability; and tracking whether teach-back is used at discharge.
For digital health teams, the priority is testing products with users who have limited literacy and limited digital skills, and preserving non-digital routes for those who need them.
For clinicians, the most effective habit is also the simplest: use teach-back routinely, with every patient, and treat a failed teach-back as a sign that the explanation needs to change.
For educators, communication skills for people with limited health literacy should be taught as clinical skills, practised and assessed like any other.
What to watch next
Watch whether health systems begin to measure organisational health literacy, for example the readability of their materials or the complexity of their pathways, and report it alongside other quality indicators. Watch the design of digital health services and whether non-digital options are protected. And watch for the next rounds of population health literacy surveys, which will show whether gaps between social groups are narrowing or widening as health systems become more digital.
The clinician’s own literacy blind spot
There is one more uncomfortable point. Clinicians are, almost by definition, among the most health-literate people in any population. That makes it hard for them to see how their explanations land. A phrase such as “take it with food,” “your results are negative” or “we’ll keep an eye on it” seems clear to the person saying it. To a patient, “negative” can sound like bad news, and “keep an eye on it” can mean anything from a scan next week to nothing at all.
Health literacy research is sometimes read as a description of patients’ deficits. It can equally be read as a description of professionals’ blind spots. The most effective communication training often starts by showing clinicians recordings of their own consultations, or asking them to explain a plan to someone outside medicine. The gap between what was said and what was understood is usually larger than anyone expected, including the clinician.
None of this requires new technology or large budgets. It requires institutions to treat comprehension as an outcome they are responsible for, in the same way they are responsible for infection rates or waiting times. A discharge that the patient did not understand is not a completed discharge. Counting it as one hides a failure that will reappear, often expensively, a few weeks later.
The takeaway
Low health literacy is real, common and harmful, and the evidence linking it to hospitalisation and mortality is consistent. But when nearly half of adults in some of the world’s best-educated countries struggle, the problem cannot be solved one patient at a time.
The more honest question for any health service is not “how literate are our patients?” It is “how literate do we require them to be, and why?”
What is the most confusing thing your health system asks patients to do? Tell us in the comments.
