Closing the Loop on Pest Risk

Two North Wales hospitals logged 359 and 314 pest-control call-outs across four years. Both were already contracted for eight scheduled visits a year, so those call-outs sat on top of a programme that was running throughout. Whatever was missing, it was not attendance.

Freedom of Information data released by Betsi Cadwaladr University Health Board covers 2021 to 2024. Ysbyty Glan Clwyd recorded 359 pest-control call-outs across that period and Wrexham Maelor recorded 314. Every North Wales hospital is contracted for eight scheduled pest-control visits a year, and the call-out figures are additional to those visits.

That last detail does more work than the totals. Attendance was not the missing ingredient. A scheduled programme was running throughout, and several hundred reactive attendances arrived on top of it. The reporting largely stopped at the totals. The table underneath them is more interesting.

Glan Clwyd recorded 91 call-outs in 2021, 92 in 2022, 131 in 2023 and 45 in 2024. Wrexham Maelor recorded 149, then 102, then 45, then 18. Wrexham's reactive call-outs fell by roughly eighty-eight per cent across the period. Glan Clwyd spiked in 2023 and then returned its lowest figure of the four years. The health board has noted that the data runs to 2024 and that rapid remedial works followed later rat sightings at Glan Clwyd, after which no further incidents were reported.

Something at both sites changed, and the data does not say what. Call-out figures move for several reasons that have nothing to do with pest pressure. Recording practice improves. Buildings close or open. Contracts change hands and reporting conventions change with them. Nobody should read a single FOI table as proof of a control strategy. But the direction of travel is worth sitting with, because of what it can and cannot be produced by.

What a falling call-out log proves

A call-out is a request to remove a problem, and the record it creates is a record of the problem arriving. It does not, by itself, establish why the activity occurred or whether it will occur again.

A falling call-out curve is a different document. Reactive demand does not decline because the response improved. A faster technician does not reduce the number of times a drain floods, a shutter is left open, or a damaged seal admits mice. Sustained reduction in reactive attendance comes from correcting the conditions generating it, or it does not come at all. You cannot respond your way to eighteen.

A site can meet every service level for four years and arrive in year five with the same drain, the same door and the same recommendation appearing on the same report. Nothing in that record is negligent. The inspections took place, the defects were identified, the recommendations were appropriate, and the client was informed each time. The system can still fail, because the useful question was never whether anybody did their job. It was why a known condition travelled through multiple reporting periods without becoming a completed, verified corrective action.

That distance, between finding something and changing something, is one of the least examined parts of commercial pest management. It is where recurring activity is manufactured, and it is what separates a site whose call-out column falls from one whose column does not.

Stating a change is not the same as managing one

A recommendation says what should happen. Repair the damaged seal. Clean beneath the redundant line. Correct the standing water beside the waste area. Remove the obsolete stock providing harbourage. Each may be entirely correct, and none of them, by itself, is a corrective action.

A corrective action begins when the finding enters a managed process. Somebody accepts responsibility for it. The required outcome is defined rather than implied. A priority and a target date are agreed. Interim controls are considered where permanent correction cannot happen immediately. Progress is visible, completion is evidenced, and where the original finding carried meaningful pest risk, the result is checked afterwards to establish whether the correction worked.

That is a different thing from adding a line to a service report.

Pest-management systems can become very good at producing recommendations while remaining poor at resolving them. The result is familiar. The same proofing defect appears in January, February and March, each occurrence documented, each one adding a page to a file that is growing without becoming more useful. A repeated recommendation is not only evidence of persistence. At some point it becomes evidence that the process for handling recommendations does not work.

Recommendations disappear at the boundaries between departments

Most recommendations are not consciously ignored. They are lost in the handovers.

A technician finds a damaged seal at the rear of a warehouse, with daylight beneath the door and a history of mouse activity on that elevation. The finding is photographed, recorded and prioritised, and the report recommends repair. Maintenance is busy. The repair enters a works list. There is uncertainty over whether the door belongs to the landlord or the occupier. Four weeks later the gap remains, and the recommendation is raised again. A month after that there is fresh activity near the same elevation, so monitoring is added, the immediate problem is controlled, and the recommendation appears on a third report.

Every handover in that sequence is reasonable in isolation. The contractor identifies a building defect, but facilities owns the repair. Facilities agrees it, but maintenance controls the schedule. The site can authorise the work, but the landlord owns the door. Together those handovers make an efficient place for corrective actions to die.

Email is particularly good at this. A recommendation is forwarded. Somebody replies that it has gone to maintenance. A later message says a quotation has been requested. Three weeks on, everyone holds a slightly different understanding of whether the action is live. The pest report still says open. The maintenance system says awaiting parts. The person responsible for food safety believes it was completed last Tuesday. The technician standing in front of the defect knows it was not. The problem is not a shortage of information; it is the absence of a single governed state for the action.

When the North Wales figures were discussed among pest-management practitioners this month, the reaction was consistent and it did not point at the technician. Several described writing clear actions into reports and then waiting weeks on stakeholders holding separate remits and separate budgets. One described building his own corrective-action tracker out of necessity, recording who each action had been assigned to, what response came back, and whether the repair had actually worked. That is a reasonable summary of where the difficulty sits. It is also an unusually clear description of a mechanism most programmes lack.

The contractor and the client own different halves

Two positions get repeated around recurring recommendations. From the contractor: we raised it on the report, and the client did not act. From the client: you are our pest-control company, so why is the problem continuing. Each contains some truth. Neither is a management system.

A pest professional should not pretend to control matters outside their authority. A contractor cannot authorise capital expenditure, alter production procedures, repair every defect or compel another department to act. Equally, identifying a material condition and restating it indefinitely is not enough. The professional obligation is to make the significance clear, keep the action traceable, reassess it as circumstances change, and escalate where an unresolved condition is contributing to continued risk.

The client's obligation is different. Once an action depends on building fabric, hygiene, operations, drainage, waste or maintenance, the site has to place it with somebody who can actually cause the change.

Commercial pest management usually leaves that handover implicit. It should not. The contractor owns the quality of the finding and the pest-risk judgement. The organisation owns the internal authority to correct conditions under its control. The system between them decides whether the matter reaches closure.

What a corrective action register has to answer

A hospital call-out log is a blunt instrument, but at least it exists and can be requested. In audited food environments the same mechanism carries a consequence a call-out log does not, because the evidence is examined by somebody with the authority to withhold approval.

On complex or audited sites, a corrective action register is often more useful than a longer service report, because it creates continuity between visits. It does not need twenty fields. It needs enough structure to answer a small set of questions reliably: what was identified, why it matters, what must change, who owns it, how urgent it is, when it is expected to close, what its current status is, what evidence will demonstrate completion, and whether the correction has been verified. Those questions read as administration until an issue persists for several months. Then they become operational.

The expectation is not only internal. SALSA, the food-safety approval scheme designed principally for smaller food and drink producers and suppliers, published Issue 7 of its Food & Drink Production Standard on 1 June 2026. From 1 September 2026 all SALSA audits are conducted against the new issues, with no transition period. The pest-management requirements carried into the standard connect three things rather than one: inspection records are expected to include the actions taken in meeting the pest-control operator's recommendations, and inspection results are expected to be assessed and analysed for trends. The useful part is not the existence of another record. It is the required link between recommendation, action and outcome.

Priority is not fixed at the point of discovery

A common weakness is that an action receives a priority when it is raised and then keeps it indefinitely. That assumes risk is static.

Consider a small proofing gap in an external warehouse elevation, recorded as a routine proofing recommendation because no activity was associated with it. Six weeks later the defect is still open, and monitoring nearby is now recording repeated mouse activity. The recommendation sits in a different risk context, and a system that still presents it as the same low-priority item has failed to absorb new evidence. Time changes risk. Recurrence changes risk. Associated activity changes risk.

That does not mean every overdue recommendation becomes urgent. A cosmetic defect does not become critical because maintenance took a fortnight longer than expected. It means priority should be reconsidered when the evidence around the action changes.

Escalation is the mechanism, and it is not a more strongly worded email. It is the point at which the system accepts that the original route to closure is no longer adequate. That may bring in a technical manager, justify an interim control, add monitoring, or produce an agreed revised completion date with the reason recorded. What matters is that an unresolved recommendation does not become administratively invisible because it has become familiar.

None of which rewards speed for its own sake. Some corrections need engineering work, shutdowns, landlords or budget approval. An action can remain open for three months and be well controlled. Another can be three days old and already lost. Age matters. Governance matters more.

An overdue action is evidence in its own right

Suppose a site holds twenty-four open pest recommendations. The number alone says little. If twenty were raised during a site review three days ago, the register may be working perfectly. If fifteen have been open for more than six months, the reading changes, and if eight concern the same building elevation it changes again.

Pest trends and corrective actions should not sit in separate documents, each reviewed on its own. Activity is one form of evidence. Open actions are another. The question worth asking is whether they correlate.

A rise in mouse activity beside three unresolved door defects means more than either fact alone. So does an increase in stored-product insects while housekeeping recommendations around obsolete raw materials remain open. Once those relationships are visible, the register stops being a compliance ledger and becomes part of the risk analysis.

Read across a longer period, the two datasets answer the question the North Wales figures raise and cannot settle. A site whose reactive attendance is falling while its oldest recommendations are closing is demonstrating control. A site whose reactive attendance is falling while the same three recommendations remain open is demonstrating something else, and it is worth establishing what.

Completed is not the same as closed

A damaged seal is replaced and the maintenance action is complete. Completion answers whether the work was done. Verification answers whether the work addressed the condition that caused concern. Those are not always the same.

A contractor can fit a new brush strip and leave twelve millimetres open at one end. A deep clean can be signed off without reaching the inaccessible void where breeding was suspected. Waste containers can be relocated while collection frequency stays inadequate. The completion evidence may be entirely genuine and the pest risk may remain.

Verified closure therefore needs a final piece of professional judgement. On a simple action that may be no more than checking the repair on the next visit. On a significant one the relevant evidence is what the monitoring shows afterwards. If the same problem reappears immediately, the work may have been completed without correcting the cause. Closure should describe the state of the risk, not the state of the work order.

Most pest corrective actions are not pest actions

Pest activity is the event that attracts attention. The action usually belongs somewhere else. A cluster of fruit flies leads to drainage or hygiene. Repeated external rat pressure leads to waste management. Stored-product insects lead into stock rotation, incoming goods or cleaning. Bird activity exposes building fabric.

Removing the pest deals with the immediate biological problem. The corrective action deals with the condition that made recurrence likely, which is why the phrase itself is slightly misleading. Many of the most important corrective actions in a pest-management programme are maintenance, hygiene, structural or operational actions discovered through pest management. The technician is often the person best placed to identify the condition and rarely the person best placed to correct it.

Closure should leave the programme better informed

Return to the warehouse door. The repair is completed, the technician verifies it, and monitoring over the following period shows no further internal activity. That should not simply end the record.

The evidence that this elevation was a genuine ingress route is now stronger, which may justify inspecting neighbouring doors, altering the site risk assessment, or moving where monitoring is concentrated. If similar defects have appeared across an estate, it may point to a maintenance standard worth addressing more broadly.

A corrective-action process therefore produces two outputs: closure of the individual problem, and information that improves the next period of control. Without the second, organisations become efficient at solving the same class of problem repeatedly. The action closes. The lesson does not travel. Six months later the same weakness appears elsewhere and is treated as a new event. That is technically corrective. It is not particularly intelligent.

This is the part SEPOS governs. SEPOS, the Stratton Environmental Professional Operating Standard, is Stratton's own operating standard for commercial pest risk. It is an internal operating standard rather than an external certification, and it does not replace a client's audit scheme.

Its structure treats the recommendation as the middle of the process rather than the end of it. Six stages run in the same order on every site. Inspect the site against its risk profile, monitoring plan and current operating conditions. Record findings, activity and material observations in a dated service record. Recommend what needs to change, why it matters and at what priority. Assign the required action to a responsible party with an agreed timeframe. Verify completed work and retain the evidence supporting closure. Review the accumulated record to adjust monitoring, priorities and the next period of control.

Stages four and five are the ones absent from most reports. Assignment is what stops a finding dissolving at a departmental boundary. Verification is what separates a closed work order from a corrected condition.

At programme level the same logic runs as a longer cycle: capture what the site shows, assess what it means, close it to a defensible position, then learn from what the record now holds before the loop begins again. Stratton Pulse is where that record sits, so findings, corrective actions and closure evidence are held in one place rather than across three systems returning three different answers. It is being brought onto sites progressively rather than all at once.

On many conventional pest-control reports, the recommendation is where the process stops.

A test worth running on your own programme

None of this requires redesigning the programme to find out whether the mechanism works. Three checks will do it.

Start with the shape of your own call-out data across four years rather than the total for the last one. It is worth knowing which curve your site is producing before somebody else calculates it.

Then take the oldest meaningful open pest recommendation on site, not the easiest one, and establish what condition caused it to be raised, who owns it now, whether that person knows, whether there is an agreed completion date, whether the reason for any delay is visible, whether the risk has changed since it was raised, and who will verify that the correction worked. If those answers are readily available, the action is being managed even though the permanent work is outstanding. If nobody can establish the position without searching old emails and asking three departments, that is useful information too.

Finally, take one recently closed action and follow it forward from the original finding: the recommendation, the assigned owner, the completed work, the evidence, the verification, and where relevant the effect on the activity that prompted it. That chain should be unremarkable. Strong control systems usually are.

Pest management is never finished when the report is sent

Good inspection matters. Technical judgement matters. Accurate reporting matters. But on commercial sites a significant proportion of pest risk sits in conditions the technician cannot correct alone: building fabric, doors, drainage, waste, hygiene, storage and operational behaviour. The programme is therefore judged on what happens after those conditions are identified.

A recommendation that reaches the right person, is completed properly and is subsequently verified is part of a control system. The same recommendation copied onto four consecutive reports is a record that the site knew. At some point another visit and another identical line stop offering reassurance and start asking a harder question: if the system already knew what was wrong, why did nothing change?

Which returns to the two columns of figures. Nobody outside those health boards can say from an FOI response what produced 149, 102, 45 and then 18. What can be said is that no amount of responsiveness produces that shape on its own. Reactive attendance falls when the reasons for it are removed, and the reasons are removed by corrective actions that somebody owned, completed and checked.

For most commercial sites, closing that distance is not an administrative detail around pest management. It is the part that decides whether the pest management works at all.

Corrective actions should not end at recommendation.

See how Stratton governs findings from inspection through assignment, verification and review.

Explore SEPOS → https://www.strattonem.com/sepos

Discuss a commercial site → https://www.strattonem.com/sectors

This article provides general operational information and does not constitute legal advice. Organisations should confirm the requirements applicable to their activities, products and jurisdiction.