In most buildings the Legionella question is close to binary: are the controls in place, or not. In a hospital it splits into something harder — which control belongs where. A staff kitchenette and a neonatal unit can sit on the same incoming main, but the people at the tap could not be more different, and a control that is proportionate for one is nowhere near enough for the other.
So the real decision is not whether to control hospital Legionella. It is how to match each method — temperature, supplementary disinfection, point-of-use filtration — to the part of the estate and the patients it serves. Choose by the patient risk you are actually carrying and the evidence each method leaves behind, not by whichever is cheapest to bolt on.
Why a hospital is not just a large office
Three things put healthcare premises in their own category.
First, the people. Inpatients are, by definition, unwell, and some are profoundly immunocompromised. The NHS notes that Legionnaires’ disease is more likely to be severe in older people and those with weakened immune systems or existing lung conditions [5] — which describes a large share of any hospital’s occupants, exposed at the point of care rather than passing through.
Second, the plumbing. Decades of ward reconfiguration leave long distribution runs, capped branches and redundant outlets unused since the last refurbishment. Stagnant warm water is the raw material Legionella needs, and hospitals generate it faster than almost any other building type.
Third, the governance. On top of the general duties in L8 and the technical detail in HSG274 [1][2], NHS premises work to Health Technical Memorandum 04-01, which expects a multidisciplinary water safety group, a written water safety plan, and control choices that reflect how vulnerable the patients at each outlet are [3]. Within that, augmented-care areas — intensive care, neonatal, transplant, haemato-oncology and burns — carry people with almost no margin. A colony count a healthy adult would shrug off can be a real exposure for them.
What to actually compare the methods on
Before reaching for a table, fix the axes. A control method is worth its place only if it scores on four things:
- The risk it addresses — growth across the whole system, or just the water leaving one tap.
- The evidence it leaves — what it lets you prove, and how cleanly that shows up in your records.
- The operational effort — the routine work, the competence, and what fails if a step is missed.
- The clinical trade-off — chiefly scald risk for patients who cannot react normally to hot water, but also water quality and flow at the outlet.
Hold those four steady and the three mainstream approaches separate out clearly.
Three control strategies, side by side
| Control strategy | What it actually controls | Where it earns its place | How you show it is working | The catch |
|---|---|---|---|---|
| Temperature control | Bacterial growth across the whole system — hot water kept hot, cold water kept genuinely cold | The baseline everywhere; the primary control under L8 and HSG274 [1][2] | Routine readings at sentinel and representative outlets, trended over time | Hard to hold at the far ends of a sprawling estate; raises scald risk, so it leans on TMVs that themselves need managing |
| Supplementary disinfection (system-wide, e.g. chlorine dioxide or copper-silver ionisation) | A treated residual carried through the pipework where temperature alone keeps slipping | Oversized or complex distribution, or a persistent problem zone temperature cannot reach | Dosing records, residual monitoring, and Legionella counts tracked over time | Adds a chemistry to run and verify; it is not a cure for bad design or stagnation, and needs competent oversight |
| Point-of-use filtration | The water leaving one specific outlet, right at the tap or shower head | Augmented-care outlets, and as a temporary barrier during an incident | The filter rating and its change schedule; the outlet is protected for the filter’s rated life | Per-outlet cost and discipline; filters have a finite life and a swap routine that cannot be allowed to lapse |
Read the table as additive, not either/or. Temperature control is the floor under the whole building. The other two are barriers you add where that floor cannot reach — a treated residual for distribution you cannot keep hot enough, and a physical filter standing directly between a vulnerable patient and the outlet.
Choosing it ward by ward
In augmented care, think in barriers. These are the outlets where a single failure has the least forgiving consequence, so they are where point-of-use filtration most often earns its cost, and where the choice stops being purely an engineering one. HTM 04-01 frames augmented-care water as a clinical risk question as much as a plumbing one [3], which is why infection-prevention colleagues belong in the decision, not just at the end of it.
On general inpatient wards, temperature control does most of the work, backed by monitoring you can stand behind. If you are fighting the same far-end outlets month after month and the readings will not hold, that is the signal to consider a system-wide supplementary approach rather than chasing flushes forever. Getting the hot side right is the foundation — on hot water temperature guidelines covers the numbers and how to actually hit them at the outlet.
Outpatient clinics, offices and staff areas behave more like ordinary commercial buildings, but the hospital habit of long, lightly used pipe runs follows them. A clinic room used twice a week feeds a tap that sits warm and still in between — the same low-use trap that catches intermittently occupied buildings, and the same flushing discipline answers it, as in on managing intermittently used properties.
Decommissioned and reconfigured zones need a deliberate decision, not neglect. A capped-but-not-removed branch is a dead leg feeding the live system beside it. Pulling it out once beats flushing it forever, and it tidies the picture an inspector will eventually want to follow — worth knowing what they look for, in on HSE Legionella inspections.
Where the comparison goes wrong
The table is a way to structure a conversation, not a ruling. None of it replaces your own water safety group’s judgement, and HTM 04-01 expects that group to bring clinical, estates and infection-prevention input to bear on the choices it describes [3]. Two cautions in particular.
The scald trade-off is real and local. Pushing outlet temperatures up to suppress growth raises the danger for confused, sedated, very young or elderly patients who cannot pull away from hot water, so the balance is struck outlet by outlet through correctly maintained thermostatic mixing — which is why healthcare control is not simply “run everything hotter”.
And no method here is a substitute for the risk assessment that sits underneath them. Temperatures, dosing levels, filter change intervals and monitoring frequencies are set by your competent assessment of this site and these patients, not by a figure copied from another building [2]. A negative sample tells you about one outlet at one moment; it does not retire the control.
A practical next step
Take your outlet schedule and add one column: the vulnerability of the people who use each outlet. Map augmented-care and high-dependency areas first, then everywhere temperature control is hardest to hold. That single overlay — risk against where your evidence is weakest — tells you where a barrier method is worth its cost and where disciplined temperature control is already enough. Bring that map to your next water safety group meeting rather than a generic action list, and the spending decisions tend to make themselves.
FAQ
Do hospitals have to follow HTM 04-01 as well as ACoP L8?
The L8 duties apply to healthcare premises like any other workplace. HTM 04-01 sits alongside them as the healthcare-specific guidance, adding the expectation of a water safety group, a water safety plan and control choices weighted to patient vulnerability [1][3]. Read L8 and HSG274 as the legal and technical baseline, and HTM 04-01 as how that baseline is applied in a clinical setting.
Are point-of-use filters a replacement for temperature control?
No. A filter protects the water leaving the outlet it is fitted to, for the life of that filter, which makes it a strong barrier at high-risk taps and during incidents. It does nothing for the rest of the system, and the change routine has to be kept up or the protection lapses silently — a targeted addition to system-wide control, not a way to stop doing it [3].
Why do healthcare premises sometimes run hot water hotter than other buildings?
Where temperature is the primary control, the margin against bacterial growth matters more when the people exposed are vulnerable, so healthcare guidance generally expects hot water to reach outlets hot enough to suppress growth [3][4]. The exact figures live in HTM 04-01 and your risk assessment, always weighed against scald risk — which is why thermostatic mixing at the point of use is part of the same conversation.
Sources
[1] HSE, “Legionnaires’ disease. The control of legionella bacteria in water systems - Approved Code of Practice and guidance (L8)”. https://www.hse.gov.uk/pubns/books/l8.htm [2] HSE, “Legionnaires’ disease: Technical guidance (HSG274)”. https://www.hse.gov.uk/pubns/books/hsg274.htm [3] NHS England, “Health Technical Memorandum 04-01: Safe water in healthcare premises”. https://www.england.nhs.uk/publication/safe-water-in-healthcare-premises-htm-04-01/ [4] HSE, “Hot and cold water systems”. https://www.hse.gov.uk/legionnaires/hot-and-cold.htm [5] NHS, “Legionnaires’ disease”. https://www.nhs.uk/conditions/legionnaires-disease/