A statement that a vaccine is effective needs an object: effective against what, for whom and over what period? Preventing infection, preventing symptoms and preventing hospital admission are different outcomes. Two percentages can both be correct while answering different questions. Reading the label on the number matters as much as reading the number itself.
Efficacy usually describes performance in a controlled clinical trial. Effectiveness describes performance in real-world use. Neither word means perfect protection, and neither is a permanent score attached to every person receiving the vaccine. Both describe comparisons made under particular conditions.
Start with the outcome being counted
Imagine two hypothetical studies. One counts every laboratory-confirmed infection, including infections without symptoms. Another counts hospital admissions attributed to the disease. The second study could report a different result without contradicting the first, because the outcomes are different. The populations and follow-up periods might differ as well.
When reading a headline, find the study's actual endpoint. Ask whether testing was routine or only performed when someone became ill. Ask how severe illness was defined. If a summary does not provide those details, avoid treating its percentage as a prediction of your personal chance of any infection. A clear explanation should make the measured outcome visible.
Use rates, not just counts
Suppose a fictional trial follows 1,000 vaccinated and 1,000 unvaccinated participants for the same period. If 10 vaccinated participants and 50 unvaccinated participants develop the specified outcome, the risks are 1% and 5%. The relative reduction is 1 minus 1 divided by 5, which equals 80%. These numbers are invented only to explain the calculation.
The absolute difference in this example is four percentage points, or 40 fewer cases per 1,000 participants during that period. The relative and absolute measures answer different questions and are both useful. The result does not mean 80% of vaccinated people are completely protected while the other 20% receive no benefit. It describes a comparison between group risks.
Why trial and real-world estimates differ
A controlled trial can use random allocation to make groups comparable, along with defined enrolment criteria and follow-up. Real-world studies observe vaccination under ordinary conditions, where recipients may differ from non-recipients in age, health, exposure and healthcare use. Researchers must consider those differences when estimating effectiveness.
For example, people at greater risk of severe illness may be more likely to be vaccinated. Simply comparing their raw admission counts with those of healthier unvaccinated people would be misleading. Statistical adjustment and careful study design can help, but assumptions and limitations remain. A good report explains the design rather than presenting the final percentage as if it emerged without methodological choices.
Exposure and follow-up change the picture
Two groups can have different opportunities to encounter an infection. Healthcare work, household contacts, travel and local circulation can all affect exposure. If those differences are ignored, the apparent vaccination effect can be distorted. Researchers also need to consider when protection is assessed and how much time has passed since vaccination.
A short observation period during low circulation may produce few events and an imprecise estimate. A longer period may include changes in circulating strains or waning protection. These are reasons to examine the study carefully, not reasons to abandon measurement. The useful question is whether the design answers the question being asked with acceptable uncertainty.
Read uncertainty alongside the point estimate
An estimate is usually accompanied by a confidence interval or another measure of uncertainty. A wide interval can indicate that the data are compatible with a broad range of effects. Quoting only the central value may make a small study sound more precise than it is. Precision also differs from freedom from bias: a very large study can still have design problems.
Do not rank two vaccines or programs by headline percentages from unrelated studies. A comparison needs compatible outcomes, populations, timing and methods, or an analysis that accounts for the differences. If those conditions are not met, saying one percentage is larger is arithmetic, but saying one option is better may not be a justified clinical conclusion.
Why vaccinated cases can outnumber unvaccinated cases
When most people in a population are vaccinated, there may be more vaccinated people among the cases simply because that group is much larger. The relevant comparison is the risk or rate within each group, with appropriate adjustment, not the proportion of all cases who were vaccinated. This is a denominator problem.
Consider an illustration in which two groups differ tenfold in size. Even a much lower individual risk in the larger group can still produce more total cases there. Without group sizes, raw counts cannot tell you how well vaccination worked. Age and other risk differences may further complicate the comparison, so a simple count is rarely the full answer.
Use evidence to inform an individual decision
Population evidence helps clinicians make recommendations, but your decision also depends on your risk, vaccination history, timing and any precautions. Ask which outcome the recommendation aims to reduce and what is known about people in your situation. It is reasonable to ask for both expected benefits and recognised limitations.
Keep other preventive measures in view where relevant. A vaccine can be useful without eliminating every infection or removing every need for clinical care. The most informative discussion translates the evidence into an appropriate plan, rather than asking you to choose on the basis of one impressive percentage. Clear denominators, outcomes and uncertainty make that translation possible.
Sources & further reading
General information only. This article is educational and is not a substitute for personal medical advice. Your immuniser will confirm eligibility and contraindications on the day.
Ask your pharmacist or GP about suitability, benefits, risks and alternatives before making a treatment decision.