What should a dental practice do after repeated decontamination audit failures?

Read the finding as a safety signal, not a score

The six-month decontamination audit is complete. The practice scored better overall than last time, yet the auditor found the same dirty-to-clean flow problem in two rooms and several missing records for washer-disinfector checks. A manager may be tempted to celebrate the higher percentage and file the audit for inspection. That would miss the repeated failure. An audit is useful only when it helps the practice understand whether instruments are being processed safely and whether previous action actually changed practice.

CQC's dental audit guidance says decontamination processes should be audited at least every six months, with an appropriate review depending on the findings. CQC may ask for evidence of the data collected, the action plan and subsequent action. NHS England HTM 01-05 provides the primary decontamination framework and recommends use of an audit tool. A repeated finding requires more than a new due date. It calls for a clear account of the risk, the cause and the evidence that the process now works.

This page owns the journey after a scheduled or targeted audit identifies recurring failures across the decontamination system. It does not replace the immediate response to a failed steriliser test, which is covered by the separate draft. A single device failure can create an urgent hold and trace problem. An audit may expose multiple weak controls that have persisted despite earlier action. The practice may need both routes if an audit finds an active unsafe device or instruments that may have been released incorrectly.

Do not wait for a committee meeting if a current finding may affect instruments being used today. Contain the relevant activity first, then investigate the system. The rest of the audit cycle can be deliberate, but immediate patient safety cannot be deferred until the next scheduled review.

Triage each finding by the decision it creates

An audit report often contains a long list. Sort findings by what they mean for current work. One category concerns possible patient exposure: for example, a failure to separate used and clean instruments, a missing validated cleaning step or a device check that has not been completed. Another concerns evidence: the work may have been done, but the practice cannot show it. A third concerns improvement: a written policy exists, yet staff routinely work around it because the room or timetable makes compliance difficult. These categories can overlap. A missing record must not be automatically treated as harmless, and a low audit score must not be automatically treated as proof of exposure.

Ask the infection prevention lead and clinical lead to identify whether any instrument batches, sessions or rooms need to be held pending review. If the audit discovers an active equipment fault, use the relevant device-failure process. If it discovers a cross-contamination risk, stop the affected flow and determine what can be safely processed elsewhere. If patient contact may have occurred, obtain clinical and specialist infection prevention advice on tracing and communication. Do not make a universal patient-recall rule from the audit percentage alone. The nature, timing and evidence of the breach matter.

Record interim controls in plain terms. “Use room three for clean packing until the flow is corrected” is more useful than “staff reminded.” It identifies where work can happen and how the practice will keep operating. Interim measures should have an owner, start time, expiry or review date and a clear route for staff to report that they are failing. A temporary instruction that is impossible to follow can increase the very risk it was meant to control.

The practice should preserve the original audit findings, photographs or observations and device records. Do not edit the first report to make the re-audit look cleaner. A later reviewer needs to see the initial condition and what changed. If a serious concern is found, document the clinical and regulatory advice obtained, including whether any notification or patient communication route applies to the actual circumstances. Audit failure by itself is not a universal CQC notification category.

Distinguish a missed record from a missed control

Suppose the washer-disinfector log has three blank days. The operator says the checks were performed, but the paper form was misplaced. The practice should examine device data, service records, instrument batch information and contemporaneous staff accounts. It should not simply fill in the missing entries after the fact. The investigation may conclude that a record control failed while the technical control was completed, or it may find that the checks themselves were missed. The corrective action depends on that distinction.

If the result is uncertain, make the uncertainty explicit. Ask whether potentially affected instruments can be identified and whether a competent person needs to assess them. The absence of evidence can limit what can be safely concluded. It should prompt a proportionate decision, not an unsupported assertion that all instruments were safe or unsafe. Retrospective signatures have little value if they conceal the gap.

Reconstruct the pattern across audit cycles

Bring the previous audit, action log and current findings together. Mark which issues are new, which recur, which were apparently closed and which were never tested after action. A finding may have changed wording while remaining the same underlying weakness. For example, last cycle recorded “dirty instruments placed beside packing bench,” while this cycle records “clean and used trays share trolley.” The practical failure may still be the absence of a reliable one-way flow.

Look beyond the headline score. A practice might improve many low-risk documentation items and still fail the same high-consequence control. Compare the exact observations and sample sizes. Were both audits performed on similar days and shifts? Did one include a busy late session while the other watched only the first patient of the day? Did the audit tool or standard change? A score comparison is meaningful only when the method is understood. The analysis should explain what the evidence does and does not show.

Ask workers who perform decontamination to walk through the real process, including interruptions. Where do used instruments wait when several rooms finish at once? Who is responsible when the designated operator is on leave? Does the layout force clean and dirty items through the same doorway? Are staff using equipment that works differently from the training example? CQC's IPC guidance expects an IPC lead with authority to implement change, an effective audit programme and evidence of how issues were addressed. That is a process-improvement obligation, not an invitation to blame the person observed on audit day.

Check whether the previous action was actually delivered. “Training completed” might mean a slide deck was emailed but no one observed the revised task. “New trolley ordered” might mean it was purchased but stored in a cupboard. “Policy updated” might mean the document changed while the room layout and shift coverage stayed the same. For each prior action, identify the completion evidence and ask whether it could reasonably have prevented the repeated finding.

Find the cause at the point of work

Root-cause work should be specific enough to change a control. A finding such as “staff did not follow policy” is usually a description of the event, not an explanation. Why was the policy not followed? Was it unavailable at the workbench, inconsistent with equipment instructions, taught differently by different supervisors or impossible during peak instrument demand? Was the clean area too small, or did temporary storage migrate into it? Were staffing and equipment capacity adequate for the treatment schedule?

Use observations, records, equipment data and staff accounts together. A single interview can miss intermittent pressure. A single photograph can misrepresent normal practice. Repeat a walk-through at the time when the problem occurs. If the practice is concerned about a device, check service and validation records and seek competent technical advice. If the concern is packaging or storage, trace an instrument set from use through cleaning, inspection, packaging, sterilisation and release. The aim is to see where the intended barrier fails.

Consider whether the audit itself contributed to weak learning. If the form allows a whole decontamination area to be marked “compliant” after one brief observation, it may not detect a late-session problem. If the same person both performs and checks every step, independent challenge may be limited. If actions are closed by document upload alone, no one may check what happens during care. Improve the audit method where necessary, while keeping the applicable HTM and CQC standards visible.

Do not make the investigation punitive by default. Staff need to speak honestly about workarounds and near misses. A fair review can still identify individual competence or conduct issues when supported by evidence, but the first question should be whether the system gives people a workable safe method. A recurrence across several operators, rooms or shifts strongly suggests looking at design, resources and supervision rather than issuing the same reminder again.

Choose actions that alter the process

Each action should target an identified cause. If clean and dirty flows cross because of room layout, a new poster alone is unlikely to solve it. The practice may need a revised route, physical separation, different equipment location, timetable change or a separate decontamination area. If washer-disinfector records are missed because no one owns the first-use check on a split shift, assign the task and handover. If staff use an incompatible product because the approved one is often out of stock, fix purchasing and substitution controls as well as training.

State the action in observable terms. “Improve decontamination compliance” cannot be verified. “Move the clean packing bench and observe ten instrument transfers on two different shifts without dirty-to-clean crossing” gives the team a testable target. The sample numbers in such a local action are a management choice, not a CQC mandated universal sample. Set a measure that suits the risk and practice size, and record why it is enough to detect the failure that recurred.

Separate immediate containment from durable change. A temporary hold on one room may protect patients today. A redesigned work area and new handover may prevent the problem next month. Both need owners and dates. If a capital change is required, set interim controls and review them until the work is complete. Do not close the audit action because a business case was submitted; the risk remains until the safe method is operating and checked.

Staff should receive a concise explanation of the changed method and the reason for it. Show the actual route for instruments, not just a policy paragraph. Assess competence at the task where necessary and allow questions. Update local instructions, diagrams and training material so they agree with the new practice. If a change affects cleaning contractors or maintenance staff entering the area, include them in the handover. The revised control must survive normal staffing variation.

Allocate a decision owner and a verifier

The action owner coordinates delivery. The verifier tests whether it works. In a small practice the same senior person may take both roles, but a second view is valuable for recurring findings. The record should show what evidence the verifier examined, what they observed and what remained open. A manager should be able to ask why a finding was closed and receive more than “the policy was updated.”

If several findings share a cause, one substantial action may address them together. Keep each finding linked to the shared action so none disappears. Conversely, one finding may need multiple actions: an equipment repair, staff instruction and re-audit. Do not force every line into one superficial tick. CQC expects evidence of data collection, action planning and subsequent action, so a coherent trail matters more than a large number of tasks.

Re-audit the control that failed

CQC describes re-audit as repeating the cycle to measure what changed after action. Design the follow-up around the failure, not the easiest metric. If the problem was dirty-to-clean transfer during high-volume sessions, observe those sessions. If documentation was unreliable, sample records over enough days to include different operators. If a device was repaired, compare post-repair checks with the previous defect. The re-audit should be able to show whether the original risk is controlled in normal work.

Set a review point that allows the new control to operate but does not leave an unresolved risk open for months. The interval depends on the seriousness, frequency and nature of the change. A practice should not present the minimum six-month decontamination audit cycle as permission to postpone a targeted check after a significant finding. CQC explicitly says review should be appropriate to audit outcomes. Where immediate containment remains in place, monitor it until the durable control has been verified.

Keep the same standard where comparison is intended, or explain why it changed. If an audit tool is updated, map the old and new criteria. A lower score may reflect a more demanding method rather than deterioration; a higher score may reflect a narrower sample rather than improvement. Record the sample, observations, exclusions and limitations. The aim is a defensible conclusion about actual practice, not a favourable number.

If the re-audit still fails, do not merely reset the due date. Reopen the cause analysis. Was the action delivered fully? Was it the wrong intervention? Did it work only when the IPC lead was present? Did an equipment, staffing or schedule change introduce a new problem? A second failure is useful evidence that the first explanation was incomplete. Escalate the unresolved risk and revise the control. Keep clinical activity restricted where the safety basis remains inadequate.

Build an evidence trail that can be followed

A useful audit file has the standard used, scope and method, raw findings, immediate risk decisions, cause analysis, action plan, delivery evidence, re-audit and final management review. Those items can be linked in a compact record. They do not need to be duplicated in several folders. The trail should let a new manager understand why the practice considered an action effective and what still needs checking.

Protect patient and staff information. Audit observations may describe a clinical session or individual worker, but the general action log does not need unnecessary identifiable details. Keep any patient-specific clinical review in the approved clinical record. Store staff performance material in the appropriate restricted process. A compliance platform may track tasks and documents; it should not become an uncontrolled copy of patient records or disciplinary notes.

Make evidence trustworthy. Date observations when they occur. Preserve failed results and superseded instructions. Link to service reports rather than writing “engineer fixed it” without the report. Record who approved a restart and what they relied on. If an action is deferred, show the interim control and the reason for the new date. A transparent audit can contain open problems; an apparently perfect audit with no trace of known failures is less credible and less useful.

Use the management review to test whether the improvement lasts. At a later point, check incident reports, equipment faults, instrument demand and staff feedback for signs that the workaround has returned. Compare recurring themes across rooms. A practice with repeated findings may need to reconsider capacity or layout, not only retrain individuals. CQC's well-led guidance treats audit and learning as part of governance.

Work through three practical failure patterns

The same flow problem returns. An earlier audit found used trays placed near the clean packing bench. Staff completed refresher training, but the next audit finds the same problem during a busy afternoon. Observe the actual tray movement and capacity. The cause may be a bottleneck, a poorly placed trolley or lack of protected clean space. Hold unsafe practice, redesign the route and test it during the busy period. A staff signature on a training sheet is not enough to close the issue.

The records are incomplete again. The practice cannot find several first-use equipment checks. It should establish whether checks were missed or only records were lost, using device data and contemporaneous evidence. Any uncertainty about affected instrument batches needs a competent clinical and technical decision. The durable action may be a clear shift owner, a point-of-use log and a daily supervisory check. Re-audit across different shifts. Do not backdate entries or treat a new template as proof of use.

The audit score rises but one high-risk finding persists. Many low-risk policy and storage items improve, but the washer-disinfector validation concern remains. The manager should keep that item separately open, obtain technical advice and control instrument processing until there is an adequate basis for use. An aggregate score can be reported with context, but it must not override the unresolved safety decision. The re-audit should focus on the validation and release control, not simply repeat every easy item.

These patterns show why an audit is a management tool. It tells the practice where to observe, decide and change work. It does not award an automatic safe status to every instrument or room.

A practical closure test for recurring findings

• Has someone assessed current patient and staff risk, including whether any work must stop?
• Is the original observation preserved and linked to the previous audit finding?
• Does the cause analysis explain the repeated failure at the point of work?
• Is each interim control assigned and still functioning while durable work is incomplete?
• Does the action change a physical, technical, staffing or decision control where needed?
• Do staff and contractors who perform the task understand the new method?
• Has the re-audit sampled the circumstance in which the failure originally occurred?
• Does the evidence show the control working across relevant rooms, shifts and operators?
• Have remaining uncertainties been escalated rather than closed with a generic statement?
• Is there a later review trigger if equipment, staffing, layout or guidance changes?

A “complete” action is one that has been delivered and verified against the problem. If the practice cannot answer these questions, keep the finding open with a named owner and current interim control. A due date alone does not protect the next patient.

What Complys may support, subject to product review

The live Complys dental page presents general compliance evidence, audit and task oversight. A practice could potentially use a reviewed workflow to assign corrective actions, attach a service report and track re-audit dates. Public copy does not verify that Complys runs the Infection Prevention Society audit tool, reads device logs or judges whether instruments are safe for use. Those functions must not be promised in this page without a product demonstration.

Keep the clinical and technical decisions with the practice's competent people. Complys should not be described as replacing HTM 01-05, the IPC lead, device manufacturer or patient record. The valuable question for a product review is whether the current system can link a finding to an owner, evidence and verification outcome in a way staff can actually maintain.

Suggested CTA: Ask Complys to demonstrate current audit-action and evidence-tracking features for dental practice governance. Have the IPC lead and relevant technical specialists approve the risk decisions and closure evidence.

Related tool opportunity: A decontamination audit action register could connect each finding to immediate containment, cause, corrective work, re-audit sample and verifier. It should remain a decision aid, not an automatic instrument-release certificate.