Phoenix Journal · LEV & Air Quality
An exposure incident rarely announces itself. It arrives as a diagnosis or a worrying test result, long after the harm was done. Investigating it well means working back from the symptom through the unsafe conditions to the management failure underneath - the method the HSE sets out in HSG245.
Most workplaces know how to react to an accident. Someone is hurt, the scene is obvious, and the response is immediate. An exposure incident is harder, because often there is no bang and no blood. What arrives instead is a letter from a GP, a worrying result on a lung-function test, or a cluster of people on the same line reporting the same cough. The harm has already happened, quietly, over weeks or months, and the job of the investigation is to work backwards from a symptom to the failure that caused it.
The HSE does not make accident investigation an explicit legal duty, but it plainly expects one. The Management of Health and Safety at Work Regulations require employers to have arrangements for planning, monitoring and reviewing their controls, and you cannot review a control that has just failed without finding out why. The regulator's workbook on investigating accidents and incidents, HSG245, is the standard method, and it applies just as well to a fume or dust exposure as to a fall.
There is also a hard reason to get this right. A confirmed case of occupational asthma, where the person's work involved significant or regular exposure to a known respiratory sensitiser, is reportable to the HSE under RIDDOR once a doctor has diagnosed it in writing. The report is a legal duty in its own right and should not wait for the investigation to finish. Treat the reporting and the investigating as two separate clocks that both start the day you learn of the case.
Three depths of cause
HSG245 sets out a four-step method: gather the information, analyse it, identify the risk controls that were missing or failed, and build an action plan that is actually implemented. The heart of it is the idea that every adverse event has causes at three different depths, and stopping at the first one is the classic mistake.
The immediate cause is the agent of harm itself: the fume, the dust, the mist that reached someone's lungs. The underlying causes are the unsafe acts and conditions that let it happen, such as an extraction hood switched off, a filter never changed, or a respirator that never fitted. The root cause is the management failure from which all the others grow, often remote in time from the event, such as a job that was never risk-assessed, an LEV system with no maintenance schedule, or a low priority given to health next to output. Fix only the immediate cause and the same incident returns wearing different clothes.
Good information decays fast, so gather it early. Preserve the scene and the equipment as they were, take the accounts of those involved and any witnesses while memories are fresh, and pull the paper trail together at once: the risk assessment, the COSHH assessment, the LEV examination and test reports, maintenance logs and training records. Those documents are usually where the root cause is hiding, because they show what was supposed to happen against what actually did.
The signal you nearly missed
The uncomfortable truth about exposure incidents is that the workplace rarely spots them first; a health check often does. A falling spirometry reading, a new entry on a respiratory questionnaire, or a dermatitis score creeping upward can all be the earliest warning that a control is failing, which is exactly why how to build a health surveillance programme matters as much to an investigator as to a nurse. Surveillance results, read at group level rather than one person at a time, turn a scatter of individual complaints into a clear picture of where and when people are being over-exposed. When an investigation begins with a surveillance flag, the data has effectively done the first step of HSG245 for you, pointing straight at the process and the people affected before anyone sets foot on the shop floor.
When it strikes in groups
A single case can look like bad luck; several cases on the same machine almost never are. Some of the most serious workplace lung disease appears in clusters, as the documented outbreaks of hypersensitivity pneumonitis around CNC coolant mist, the hazard operators breathe daily have shown, where a contaminated fluid and a poorly captured mist affect a whole group at once. When an investigation finds a cluster, the root cause is structural by definition: it lives in the control system, not in one susceptible person. That is the moment to test the extraction itself, because a group exposure and an LEV system that was never verified are two halves of the same finding, and the action plan has to fix the engineering rather than simply move the affected workers away.
Questions
Some are. Occupational asthma is reportable once a doctor diagnoses it in writing and the person's work involved significant or regular exposure to a known respiratory sensitiser. The report is a separate legal duty and should not wait for your investigation to finish. Keep RIDDOR records for at least three years.
The immediate cause is the agent that did the harm, such as the fume or dust that was inhaled. The root cause is the management failure behind it, such as a job never risk-assessed or an LEV system with no maintenance schedule. HSG245 asks you to reach the root cause, not stop at the immediate one.
A surveillance flag has effectively done the first step for you. Read the results at group level to see which process and which people are affected, then examine the control - usually the extraction - that was meant to keep exposure down. A cluster of cases points at the system rather than the individual.
Phoenix Duct Clean · by the numbers
Most exposure investigations end at the extraction. Phoenix examines and tests LEV to COSHH Regulation 9 and reports on real capture at source, not just airflow. UK-wide.