Nothing Happened

Fire safety, complacency, and the drills care homes are not running

I was in a service when the fire alarm activated. Not a test. A genuine activation, panel lit, sounders going.

Do you know what happened?

Nothing. Absolutely nothing.

No staff member came to the fire panel for direction. Nobody moved towards the zone in alarm. The fire doors released and closed automatically, exactly as designed, and behind them everybody stayed precisely where they were. Residents sat. Staff carried on. The building's engineering did its job perfectly and the people inside it did not do theirs at all.

That is not a story about bad staff. Every person in that building would have told you, sincerely, that they knew the fire procedure. They had all done the training. Their certificates were current and their signatures were in the file. What had happened was subtler and far more dangerous than ignorance: the sound of the alarm had stopped meaning anything.

Why a care home is not like any other building

Fire safety in most workplaces is a question of how fast you can get everyone out. In a care home, that question is unanswerable in those terms. Your residents cannot be told to leave. Many cannot walk unaided. Some cannot process what the noise means. A person with advanced dementia may resist the very people trying to move them, and a person who is bed-bound on an airflow mattress is going nowhere without two staff and a piece of equipment.

So care homes rely on progressive horizontal evacuation: moving people through fire-resisting compartments to a place of relative safety on the same floor, buying time rather than achieving exit. That strategy is entirely dependent on two things being true at the moment it is needed. The compartmentation has to hold. And staff have to act, immediately and correctly, without instruction.

The strategy must also be self-sufficient. It cannot be built on the assumption that the fire service will arrive and rescue people. Your own team, at whatever number is actually on shift, has to move everyone who needs moving.

This is why fire safety belongs in the same conversation as clinical governance rather than in a separatseparate folder marked health and safety. It is a care issue. The Scottish Government's guidance for care homes puts it plainly: staff should understand that the concept of care includes care from fire.

The record is not reassuring

In Scotland the reference point remains Rosepark in Uddingston, where a fire that began in a cupboard on a bedroom corridor in January 2004 killed fourteen residents. The fatal accident inquiry ran for 141 days. Sheriff Principal Lockhart concluded that the management of fire safety at the home had been systematically and seriously defective, and that some or all of those deaths could have been prevented had there been a suitable and sufficient fire safety plan.

Read the list of defective systems he identified and the pattern is uncomfortably familiar to anyone who reads inspection reports: maintenance of electrical installations, staff training in fire procedure, and the management of fire safety itself. The critical failing he named was the failure to identify which residents were at risk in a fire, and the failure to think through the worst case — a fire starting at night. The home's own procedure required a staff member to locate the fire before calling 999. Nine minutes passed before the fire service was called.

In England the enforcement picture is stark. Care homes attracted 192 fire safety enforcement notices in 2024/25, around eleven per cent of all notices issued across England — the second highest of any premises type, behind only shops. Prosecutions in recent years have produced fines in the tens and hundreds of thousands of pounds, including a care home operator fined £70,000 with £22,000 in costs in early 2026 following the death of a resident, and two directors fined nearly £125,000 for offences across four premises. A Welsh operator was fined £432,944, reduced to £300,000 on appeal.

None of those organisations set out to run an unsafe building. They ran buildings where fire safety had become paperwork.

The Tuesday morning problem

Most services test the alarm at the same time on the same day every week. Ten o'clock on a Tuesday, near enough forever.

There is a reason for that, and it is a legitimate one. BS 5839-1 recommends weekly user testing at roughly the same time during normal occupancy, rotating through different manual call points, so that the system is proven and occupants become familiar with the sound. Predictability is part of the design intent. If you have moved the test around at random and told nobody, you have created a building where every activation is ambiguous, which is its own hazard.

But look at what predictability does over two or three years. Staff learn that the alarm means Tuesday. Not fire — Tuesday. The sound becomes an administrative noise, like a delivery van reversing. And when it goes off at four in the afternoon on a Thursday, the first instinct is not to move; it is to wonder whether someone got the day wrong. That hesitation is the whole problem. In a compartmented building where you are trading time for safety, hesitation is the currency you cannot afford to spend.

The standard itself acknowledges part of this: where staff work outside normal hours, additional tests should occasionally be carried out at different times so that all staff become familiar with the signal. That is worth reading carefully, because it is an admission that a single fixed test slot leaves whole cohorts of your workforce having never heard your alarm.

The resolution is not to make the weekly test unpredictable. It is to stop confusing the weekly test with a drill. The test proves the equipment. It proves nothing whatsoever about whether your people will respond, and it was never designed to. If the test is the only time your alarm ever sounds, you have no evidence at all about human response — and you have actively trained your staff to discount the sound.

The scenarios nobody runs

Here are the questions I would want answered in any service I was responsible for.

How confident are you in your night shift? Not in principle — specifically. Two or three staff, lights low, most residents asleep, a fire in a first-floor bedroom. Who goes to the panel? Who calls 999, and at what point? Who starts moving people, and which people first? Can they physically do it? Have they ever tried?

How confident are you on a Tuesday when someone has called in sick? Your evacuation strategy is almost certainly written for your establishment figure. Care homes rarely run at establishment. If your horizontal evacuation needs four staff to move eleven people and you have three, your plan is fiction on the day it matters, and the fire risk assessment that assumed four has not been reviewed.

When did you last run an unannounced scenario? Not a walkthrough. Not a tabletop. An activation nobody was warned about, observed by someone whose job is to watch and record rather than to help. How long until somebody reached the panel. Whether they read the zone correctly. Whether the person who took charge was the person the plan names. Whether anyone called 999 or everyone assumed someone else had.

Do your agency and bank staff know any of this? They are in your building overnight, and they are often the people who have had the shortest induction and heard your alarm the fewest times.

You will not enjoy the results of the first unannounced scenario. That is precisely why it is worth doing, and it is worth doing before someone else does it for you.

The fire load you have not thought about

Ask most managers about fire loading and they will talk about the boiler room and the laundry. Meanwhile the building is full of things nobody has assessed.

Wooden panelling and dados.Common in older and converted buildings, often

●      beautiful, and frequently untreated. Surface spread of flame along a corridor is exactly the mechanism that turns a contained fire into a fatal one. Panelling on an escape route needs an appropriate surface rating, achieved either through intumescent treatment or replacement. “It has always been there” is not a fire strategy.

●      Timber staircases, bannisters and handrails. The stairwell is usually your protected route. Anything combustible within it deserves specific attention, as does anything stored under it.

●      Curtains, blinds and drapes. Contract-standard flame-retardant fabric to the appropriate British Standard, with the retardancy verified after repeated industrial laundering, because it does not last forever. Domestic curtains bought in good faith to brighten a lounge are a different product entirely.

●      Furniture and furnishings. Care home furniture should meet contract standards, not the domestic standard that applies to what you buy for your own living room. Which raises the question nobody likes: what about the armchair a family brought in because Mum has always sat in it? The fabric headboard? The cushions, the throws, the crocheted blanket? These arrive one at a time, with love, and they are never risk assessed.

●      Everything we do in the name of homeliness. Memory boxes, fabric wall hangings, tactile displays, artwork covering corridor walls, seasonal decorations put up every December and assessed never. Good dementia design and low fire loading pull in opposite directions, and pretending otherwise does not resolve the tension. It just means it gets resolved by accident.

●      Paraffin-based emollients. Regulators have warned repeatedly that paraffin-containing creams transfer to clothing and bedding, where they build up over time and dramatically increase flammability. If you have residents on emollients and residents who smoke, that is a specific, documented, foreseeable risk, and it needs to be on a risk assessment with a laundering protocol behind it.

●      Pressure-relieving and airflow mattresses. Substantial volumes of polyurethane foam and air, plus a pump, in the bedroom of the person least able to leave it.

●      Lithium batteries. Mobility scooters, powered wheelchairs, e-cigarettes, chargers of unknown provenance brought in by families and plugged in behind bedside cabinets. Where are they charged, on what surface, and who checks them?

●      Alcohol hand sanitiser. Necessary, flammable, and often mounted in quantity along the escape routes.

●      Wedged fire doors. Still the most common, and the most consequential, finding of all.

Where this actually goes wrong

The failures I see are rarely failures of equipment. Detection systems work. Doors close. Extinguishers are serviced and tagged.

What fails is the management system around them. The weekly test is logged but nobody analyses the log. The drill happens and the debrief does not. Someone notices that only two staff attended the last drill and records the number without asking why, or what changed as a result. The fire risk assessment is reviewed annually as an administrative event rather than after the refurbishment, the new admission with a hoist requirement, or the change to night staffing that materially altered the risk.

This is the same pattern regulators are describing across the whole of care governance at the moment, and it is the “so what” that keeps going missing. You found something. What did you do with it? What changed for the people living there? Who did the work, and who is accountable for making sure it does not quietly slip back?

For fire safety, that pattern has a consequence that other governance failures do not. Most of the time, poor oversight produces poor outcomes gradually. Fire produces them in about four minutes.

Five things worth doing this month

1.    Run one unannounced scenario, observed and recorded. Night shift first, because that is where your exposure is greatest and your evidence thinnest. Write down what actually happened, including how long nothing happened for.

Test your plan against your real rotas, not your establishment.Take the minimum staffing you have actually run in the last three months and work out whether the 1.    evacuation strategy holds. If it does not, that is a fire risk assessment finding, not a staffing grumble.

2.    Walk the building looking specifically for fire load nobody assessed. Family-brought furniture, soft furnishings, decorations, panelling, charging points, emollient use. Log what you find and act on it.

3.    Separate your test from your drill, in writing. Keep the weekly test doing its job. Then create a genuinely separate programme of varied, unannounced scenarios, and make sure night, weekend and agency staff are inside it.

4.    Debrief every activation, including the false alarms. Especially the false alarms. They are free data on how your people behave when they think it is nothing, which is exactly how they will behave the first time it is something.

What has changed recently, and one thing to be careful about

Both jurisdictions have moved. In Scotland, the Scottish Government's practical fire safety guidance for existing care homes was revised in February 2022 and supersedes the 2014 version — worth checking which edition your fire risk assessment was actually written against. BS 5839-1 was revised in 2025, and notably now treats remote alarm monitoring in residential care homes as a default requirement rather than an optional extra. BS 9991 was extended in its 2024 revision to cover residential care homes specifically. In England, MHCLG updated its enforcement and sanctions guidance under the Fire Safety Order in July 2026.

One caution. There is guidance circulating that states the Fire Safety (Residential Evacuation Plans) (England) Regulations 2025, in force from 6 April 2026, apply to care homes. Read the regulations. They apply to buildings containing two or more sets of domestic premises that are 18 metres or seven storeys and above, or over 11 metres with a simultaneous evacuation strategy. A conventional care home is a single premises, not multiple domestic dwellings, and generally falls outside that scope.

That is not a reason to relax. Personal emergency evacuation planning has been expected practice in care homes for years, driven by the fire risk assessment and by the Regulatory Reform (Fire Safety) Order itself rather than by the 2025 Regulations — and in Scotland by person-centred fire risk assessment under the Fire (Scotland) Act framework. If you do not have an individual, current, tested evacuation plan for every person who cannot self-evacuate, you have a problem regardless of which set of regulations does or does not bite. But get the legal basis right, because citing the wrong instrument in your documentation undermines everything else in it.

Fourteen people died at Rosepark in a building where the fire safety paperwork existed. The inquiry did not find an absence of documents. It found an absence of management.

The alarm in that service I visited was working perfectly. Everything worked. Nothing happened.