Work-order data, a blocked fire exit and a disputed finding
During the stage 2 audit of a 900-bed hospital's FM system, the FM team presents its CMMS dashboard: planned preventive maintenance completion is 98.4% for the year against a target of 95%. While sampling you notice that 412 work orders across the year were closed on the last working day of each month, all with the identical free-text note 'done', and all by three supervisors. A technician tells you, off the record, that jobs are often closed in batches so the monthly figure 'stays green'. On the building tour you see a fire exit on a patient corridor obstructed by two linen trolleys and an unlabelled five-litre container in a plant room. Your trainee auditor wants to announce a major nonconformity immediately to the operations director, who is escorting the tour. Later the FM head says: 'The client only looks at the percentage - that is what they want.'
1. How do you handle the obstructed fire exit and the unlabelled container, both immediately and in the audit?
Read the worked answer
The immediate hazard comes before audit protocol. I would stop, tell the escort plainly what I have seen and where, ask for the trolleys to be moved and the container to be identified and labelled or removed, and stay until something is done or escalated. I would record the time, location and what action was taken, and note who was informed. If the escort were unable to act and patients were at risk, I would escalate to the most senior available person, because duty of care does not pause while an audit is in progress. Photographs, if permitted, support the record.
Then I would separate the immediate correction from the audit judgement. A single obstructed exit does not automatically prove a system failure; it may be an isolated lapse. What matters is whether the FM system would have detected it: were routine fire safety inspections scheduled and recorded, and do the records claim the exit was clear on the day? Was there a hazard reporting process, and is there evidence staff use it? Are chemical containers covered by a labelling and storage control, with trained staff and stock checks? That line of enquiry tests clauses 8.1 and 9.1.
If the checks were recorded as satisfactory when the condition existed, the monitoring process is producing false assurance, and a nonconformity is well founded. If everything else works and this was a genuine one-off outside the inspection cycle, an observation and an agreed correction may be proportionate. I would not turn the tour into a full health and safety inspection, but I would not walk past it either.
2. How do you build the finding on the closed work orders, and what determines whether it is minor or major?
Read the worked answer
The dashboard figure alone proves nothing; the question is whether the closures represent work actually done. I would triangulate. Take a sample of 25 of the 412, and for each compare the closure date with the supervisor's shift pattern, the technician's timesheet, building access logs, spare parts issued, and any subcontractor invoice or attendance record. Physically inspect a few assets claimed to have been serviced and look for evidence of recent intervention. Interview two of the three supervisors and a technician openly, not relying on the off-the-record remark, and ask what happens at month end when jobs are outstanding.
Then test the control: does anyone validate completion before closure; is there a second signature or photo evidence; is there a backlog report; does management review the backlog? The 'client only looks at the percentage' remark points to clause 5.1 leadership and to the design of the KPI itself - a measure that rewards closure rather than completion invites exactly this behaviour.
Grading depends on scale and consequence. If 412 closures are a small fraction of the annual workload, the work was largely done and only the recording was batched, a minor nonconformity against 9.1 may be right. If a material proportion of PPM was recorded as complete without evidence of work, and the figure drives client reporting and management decisions, the performance data is unreliable and the finding is major against 8.1 and 9.1. The proportion matters, and I would calculate it before grading.
3. Your trainee wants to announce a major nonconformity on the spot, and at the closing meeting the operations director disputes the finding. How do you manage both?
Read the worked answer
The trainee's instinct is understandable and wrong. Findings are presented at the closing meeting after verification, and announcing one mid-audit from a single observation risks being wrong in public and converting an audit into a confrontation. I would explain that the batching evidence needs triangulation first, that the tour observation needs its control tested, and that premature disclosure can also compromise the audit by prompting people to change records or rehearse answers. If the fire exit hazard needed immediate action, that is a safety escalation, which is different from declaring a nonconformity.
The closing meeting has a structure that protects everyone: purpose and scope, summary of what was covered, findings presented with evidence and the requirement quoted, the auditee's response recorded, agreement on correction and corrective action timescales, and explanation of the appeal route and confidentiality. Findings are stated against processes and requirements, never against named individuals.
If the operations director disputes a finding, I listen for new evidence. If it is genuinely new and checkable, I verify it and amend the finding if the evidence warrants it. If it is disagreement with the interpretation, I restate the evidence and requirement once more, record the disagreement in the report, and point to the appeal process. What I would not do is withdraw a well-evidenced finding to keep the meeting calm, or trade it against another finding. The report goes out as the evidence shows, with the auditee's position recorded alongside it.