---
title: "How to read UK Legionnaires' disease figures without being misled"
source_url: https://legionella.io/articles/how-to-read-uk-legionnaires-disease-figures/
canonical_url: https://legionella.io/articles/how-to-read-uk-legionnaires-disease-figures/
pillar: "Legionella Basics & Science"
summary: "What 'confirmed' versus 'presumptive' means, why year-on-year swings can be testing artefacts, and the under-ascertainment caveat that makes the totals a floor."
primary_keyword: "read Legionnaires disease statistics"
date_published: 2026-06-27
date_reviewed: 2026-06-27
author: "Legionella.io editorial team (REMOTE TECH LTD)"
reviewed_against: "HSE L8 and HSG274 guidance"
region: "United Kingdom"
license: "CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote, adapt or republish with attribution to REMOTE TECH LTD and a link to source_url."
license_url: https://creativecommons.org/licenses/by/4.0/
---

# How to read UK Legionnaires' disease figures without being misled

A single year's Legionnaires' disease count tells you far less than the headline suggests. The published surveillance figures are real and carefully produced, but they answer a narrower question than most people assume, and quoting them as a clean measure of "how much Legionnaires' disease there was" is the most common mistake made with this data.

If you are about to put a number in an article, a report, a council paper or a tender, this is what that number does and does not support saying.

The figures come mainly from UKHSA's national surveillance of legionellosis in residents of England and Wales, with Scotland, Wales and Northern Ireland running their own systems alongside it, and ECDC pooling a European picture on top [1][2]. Each is a count of cases that were diagnosed, tested and reported. That chain of "ifs" is exactly where the misreadings live.

## Myth vs reality: what the numbers actually mean

| Common reading | What the figures actually support |
|---|---|
| "Cases went up this year, so risk went up" | A rise can reflect more testing, better diagnosis or a single cluster as much as a real change in underlying risk |
| "This is how many people got Legionnaires' disease" | It is how many were diagnosed and reported; under-ascertainment means the true total is higher [1] |
| "A 'presumptive' case is a probable false alarm" | Presumptive is a defined diagnostic category, not a doubtful case, both confirmed and presumptive are real reported cases [1] |
| "The figure tells me where people caught it" | Surveillance splits cases by acquisition category, but a large share are sporadic with no source ever identified |
| "I can compare the UK and another country directly" | Different countries test, define and report differently, so headline totals are not like-for-like [2] |

Read each row before you cite a number. If your sentence depends on the left column being true, rewrite it.

### Myth: a higher count means higher risk

Year-on-year movement is the figure people reach for first and misread most. A count reflects how many cases were *detected*, and detection depends on how many clinicians ordered a Legionella test, which tests were used, and whether a single outbreak inflated one year. A genuine change in community risk and a change in testing behaviour can produce the same upward line on a chart. The honest reading of a one-year rise is "more cases were detected", not "more people are at risk". Trends across several years, viewed with the caveats below, carry more weight than any single jump.

### Myth: the published total is the number of people who fell ill

It is a floor, not a ceiling. Legionnaires' disease presents as pneumonia, and not every pneumonia patient is tested for Legionella, so cases that are never tested are never counted. UKHSA is explicit that its figures are subject to under-ascertainment [1]. The practical consequence: you can correctly say "at least this many confirmed and presumptive cases were reported", but you cannot say "this is how many people developed Legionnaires' disease". The gap between those two statements is the whole point of this article.

### Myth: "presumptive" cases are dubious

Confirmed and presumptive are diagnostic categories, not a confidence ranking you should mentally discount. Broadly, a confirmed case meets a stronger laboratory standard, for example culture of the organism or a clear paired-serology result, while a presumptive case meets a defined but less definitive criterion. Both are genuine reported cases. The split also matters because the dominant test in routine use, the urinary antigen test, largely detects one species and serogroup, which shapes *which* cases get confirmed at all. Quote the category if you have it; do not treat presumptive as noise [1].

### Myth: the figure tells you where infections came from

Surveillance does record an acquisition category, community-acquired, travel-associated or healthcare-associated, and that breakdown is genuinely useful for seeing where attention is needed [1]. What it does not do is pin most cases to a source. A large proportion are sporadic, with no building, system or outbreak ever identified. So "travel-associated cases were reported" is a fair statement; "these cases were caused by hotels" is not. The category describes the patient's exposure history, not a proven source.

### Myth: you can line the UK up against other countries

European totals exist, ECDC compiles them, and the travel network detects clusters that cross borders [2], but the European picture is a join of separate national systems, each with its own testing rates, case definitions and reporting completeness. A country that tests more, or defines cases more broadly, will appear to have more disease without necessarily having more. Cross-country comparison is a job for the epidemiologists who adjust for those differences, not something a raw league table of headline counts supports.

## What the reports genuinely let you say

Used carefully, the figures support quite a lot. You can describe the reported number of cases for a period, stated as confirmed and presumptive. You can describe how cases split by acquisition category and the broad seasonal pattern, which tends to be reported. You can describe multi-year direction with appropriate caution. And you can, and should, carry the under-ascertainment caveat so your reader knows the count is a minimum.

What you cannot do is convert a single-year change into a risk statement, present the total as the true incidence, or compare countries at face value. Legionnaires' disease is also a reportable matter in occupational contexts, and the public-health surveillance figures sit alongside, not inside, that reporting. They are not the same dataset [3].

This is general guidance on interpreting published surveillance, not a substitute for the source reports themselves or for advice from the public-health bodies that produce them. When a specific number matters, cite the original report and its stated definitions rather than a figure passed along second-hand.

## FAQ

### Can I say Legionnaires' disease "rose" or "fell" based on one year's figures?

Only as detected cases, not as risk. A single-year change reflects testing, diagnosis and any clusters as much as underlying risk. If you must use one year, say "reported cases rose compared with the previous year" and add the under-ascertainment caveat. For anything about risk, lean on multi-year trends rather than a single swing.

### What is the difference between a confirmed and a presumptive case?

Both are real reported cases; the labels describe the strength of the laboratory evidence. Confirmed cases meet a more definitive standard, presumptive cases a defined but weaker one. Treat presumptive as a genuine case in a different evidence category, not as a likely false positive, and quote the category alongside any total where you can [1].

### Why are the published numbers described as an undercount?

Because diagnosis depends on a clinician suspecting Legionnaires' disease and ordering the right test, and not every pneumonia patient is tested. Cases that are never tested are never reported, so the published total is a floor. UKHSA states the figures are subject to under-ascertainment, which is why "at least" belongs in front of any count you quote [1].

### Is it fair to compare UK figures with another country's?

Not at face value. Countries differ in how much they test, how they define a case and how completely they report, so a higher count can mean more testing rather than more disease. ECDC compiles a European view, but comparisons that adjust for those differences are specialist work, a raw table of national totals does not support a "country X is worse" claim [2].

## What to do before you cite a number

Open the original surveillance report, not a secondary summary, and find two things: the case definition it is using and any caveat it states about ascertainment. Then write your sentence so it survives both, "at least this many confirmed and presumptive cases were reported" rather than "this many people caught Legionnaires' disease". If your draft sentence falls apart once you add those qualifiers, the figure was not saying what you wanted it to say.

## Related reading

- [How UK Legionnaires' disease surveillance works](https://legionella.io/articles/how-uk-legionnaires-disease-surveillance-works-four-nation/)
- [ELDSNet explained](https://legionella.io/articles/eldsnet-european-legionnaires-surveillance-travel-network/)
- [Sporadic case, cluster or outbreak?](https://legionella.io/articles/sporadic-case-cluster-or-outbreak-legionella-classification/)
- [How Legionnaires' outbreaks are detected](https://legionella.io/articles/how-legionnaires-outbreaks-are-detected-lessons-for-duty-holders/)

## Sources

[1] UK Health Security Agency, "Legionnaires' disease: national surveillance of legionellosis in residents of England and Wales". https://www.gov.uk/government/collections/legionnaires-disease-guidance-data-and-analysis
[2] European Centre for Disease Prevention and Control, "Legionnaires' disease, Annual Epidemiological Report". https://www.ecdc.europa.eu/en/legionnaires-disease
[3] HSE, "Legionnaires' disease". https://www.hse.gov.uk/legionnaires/index.htm