Self-medication with Antimicrobial Drugs in Europe

Picture a map of Europe and guess where people are most likely to treat themselves with antibiotics without ever seeing a doctor. If you said eastern Europe, you're right. In Lithuania, 210 out of every 1,000 people surveyed had actually done it in the past year. In the Netherlands, that number was 1. One. That is not a rounding difference — that is a two-hundred-fold gap across a single continent, and it tells you something important about how health systems either protect people from bad decisions or quietly enable them. This is the central finding of a landmark survey by Grigoryan and colleagues, who fanned out across 19 European countries. They sampled between 1,000 and 3,000 adults per country using randomized, multistage sampling and asked them directly: did you take antibiotics in the past year without a prescription? Would you do it in the future? Do you have any stored at home right now? The result was 15,548 completed questionnaires and one of the clearest geographic portraits of inappropriate antibiotic use ever assembled. The design was careful. Each country identified a region with average prescribed antibiotic consumption, then selected one city and one rural area within it. Adults were drawn from population registries by computer-generated random numbers. A standardized questionnaire, translated into each national language and back-translated for consistency, was mailed between March and July of 2003. Self-medication was defined precisely as taking any systemic antibacterial drug without a current prescription from a physician, dentist, or nurse. The team also tracked intended self-medication — respondents who answered "yes" or "maybe" when asked whether they would use antibiotics without consulting a doctor — and storage at home. The mean country response rate was 40 percent, which the authors addressed by applying a continuum-of-resistance adjustment, assuming late responders resemble nonrespondents. Where adjusted estimates fell within the observed confidence intervals, they treated results as reliable. Now here is the scale of the problem. Actual self-medication ranged from 1 per 1,000 in the Netherlands and 4 per 1,000 in Sweden, all the way up to 210 per 1,000 in Lithuania and 198 per 1,000 in Romania. But actual behavior was only part of the story. Intended self-medication — the population primed to do it next time — ranged from 73 to 449 per 1,000. Lithuania again topped the list at 449, Romania at 431, and Slovakia at 324. In every country, intended use was far more common than actual use. That gap represents a reservoir of future inappropriate antibiotic exposure. So what were people treating? Throat symptoms dominated — a dry, inflamed, or sore throat, inflamed tonsils, or tonsil pain. The most common stated reason for self-medicating, in actual and intended use alike, was a sore throat. Dental symptoms and bronchitis followed. The authors note, pointedly, that many of these conditions are self-limiting and that antibiotics would often not have been indicated at all. People were reaching for a prescription drug to treat conditions that would likely resolve on their own. Where they got the antibiotics depended on where they lived. In eastern Europe, 68 percent of self-medication courses came directly from pharmacies without a prescription. In southern Europe, leftovers from previous prescriptions were the dominant source at 51 percent. In northern and western Europe, leftovers also led at 44 percent, but pharmacy sales without a prescription were far less common at just 19 percent. Both routes are telling. The pharmacy route in eastern countries signals a straightforward enforcement failure: over-the-counter sale of antimicrobials is illegal in all 19 participating countries, yet it happened most in the east. The leftover route signals a different problem — pharmacies dispensing whole packages rather than exact doses, combined with patients stopping treatment early, creates a household stockpile that becomes its own shadow supply chain. Three courses were even obtained over the internet, all in Lithuania. Self-treatment lasted a median of five days, with some cases stretching to 100 days, and duration was significantly longer among people with chronic diseases. Which brings us to who, exactly, is doing this. The demographic findings from Grigoryan and colleagues are worth dwelling on, because they cut against the easy assumption that self-medication is mainly a behavior of the poorly informed. Three independent predictors emerged from multivariate analysis: younger age, higher educational level, and the presence of a chronic disease. Each additional year of age actually lowered the odds of self-medication — the adjusted odds ratio was 0.985 per year, a consistent protective gradient. Sex and urban versus rural location made no difference. Education moved in the opposite direction from what you might expect. High educational level was associated with higher odds of actual self-medication — adjusted odds ratio of 1.36 — and even higher odds of storing antibiotics at home. The authors are careful not to attribute this directly to education itself. They point to a plausible mediating factor: symptom interpretation. More educated respondents may apply different reasoning about when antibiotics are warranted, sometimes incorrectly. The data don't let us fully untangle this, but the pattern is real. Chronic disease had the largest association of all. Having any chronic condition raised the odds of actual self-medication by nearly 90 percent — adjusted odds ratio of 1.89 — and the odds of intended self-medication by more than double, at 2.32. This makes intuitive sense: people who interact repeatedly with the healthcare system over time accumulate more leftover medications, develop stronger views about when they need treatment, and may feel confident managing flare-ups on their own. Duration of self-treatment reinforced this — longer courses among chronic disease patients confirmed that this wasn't just opportunistic pill-taking. The geographic pattern in the regression was even starker. Compared to northern and western Europe, the adjusted odds of actual self-medication were 6.78 times higher in southern countries and 7.53 times higher in eastern countries. And critically, intended self-medication was an extraordinarily strong predictor of actual behavior. Among respondents who stored drugs at home, the odds ratio for actual self-medication was 20.9. Among those without stored drugs, it was still 17.8. Having antibiotics in the house, even without the intention to self-medicate, raised the odds 3.5 times on its own. Here is why that regional gradient matters beyond the statistics. The geographic concentration of high self-medication rates in southern and eastern Europe tracks onto the regions that have consistently reported higher rates of antibiotic resistance. Grigoryan and colleagues do not establish causation, but the concordance is not coincidental. Penicillins accounted for 54 percent of all self-medication courses, and southern and eastern countries showed significantly higher use of broad-spectrum penicillins. Using broad-spectrum drugs for self-limiting throat infections, without professional dosing guidance, applied inconsistently or stopped early — this is precisely how you select for resistant bacteria. The paper argues for targeting three groups. Prescribers, because they establish norms and dispensing practices that patients eventually internalize. Pharmacists, because they are the final barrier in countries where the pharmacy is the primary source of non-prescribed antibiotics. And the general public, because the storage problem requires people to actively decide not to hoard. Specific fixes proposed include enforcing prescription laws, limiting the number of tablets dispensed per course so leftovers accumulate less, instructing patients to discard what they don't use, and running large-scale public campaigns of the kind already launched in the United States, Belgium, Canada, and Australia. There are real limits to what this survey can tell us. Response rates in some countries were low, and in Spain, flagged explicitly, adjusted prevalence was consistently higher than observed prevalence, suggesting that actual self-medication there is likely underestimated. Self-reported recall introduces its own noise. The data are cross-sectional, a snapshot from 2003, not a measure of change over time. But the core finding survives all of that. Self-medication with antimicrobials occurred in every single participating country, from the best-resourced healthcare systems in northern Europe to the most strained in the east. The two-hundred-fold variation in prevalence across those same countries cannot be explained by biology. People in Lithuania do not get sorer throats than people in the Netherlands. The variation is a product of enforcement, access, dispensing policy, and public norms — every one of which is a human decision that can be made differently. That means this is a fixable problem. The map exists. The levers are identified. What remains is the will to pull them. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.