Sterile processing quality metrics should help a team decide what to improve. A dashboard that only counts completed trays can show activity while leaving important questions unanswered: how much work had to be repeated, which defects reached the next stage, and whether staff had the support to complete required tasks.
The goal is a small set of clearly defined measures connected to real decisions. You do not need a complicated software project to begin. You need consistent definitions, a trustworthy denominator, and a review process that leads to action.
This article offers an original starter framework for SPD leaders. The sample measures and numbers are illustrative, not national benchmarks or required regulatory indicators.
Begin with a question, then choose the measure
Start by writing the operational question in plain language. Examples include: Which tray issue is recurring? Are quality concerns found before or after release? Are staff waiting for training or equipment support? Does a handoff leave work without an owner?
A measure is useful when it helps answer one of those questions. Collecting a number simply because it is easy to export can add reporting work without improving a decision.
Assign an owner to each question. That person does not need to solve every related problem alone, but they should know who reviews the data and how a finding becomes an action.
If nobody can explain what would change when the number changes, reconsider whether it belongs on the dashboard.
Use a small family of measures
Different measures reveal different parts of performance. A throughput count describes work completed. A process measure describes whether a defined activity occurred. A quality outcome describes an observed result. A balancing measure asks whether an improvement created another problem.
For example, faster completion may be worth examining alongside rework and unresolved quality concerns. A training project may need to track both completed assessments and time waiting for supervised practice.
Do not expect one number to represent the whole department. A compact group of related measures gives you a more useful conversation than a single score that combines unlike things.
Keep the dashboard understandable to the technicians whose work produces the data. They should be able to explain what is counted and why it matters.
Six possible starting measures
Choose only the measures that match your current priorities. The following definitions are examples to adapt locally.
| Measure | Example definition | Decision it can support |
|---|---|---|
| Rework rate | Items or sets requiring defined repeat work divided by items or sets reviewed | Which recurring issue needs investigation? |
| Defects detected before release | Defined concerns found during internal review, reported separately from later findings | Where are problems being intercepted? |
| Defects reported after release | Defined reports linked to released sets, using a consistent denominator | Which problems are reaching the next team? |
| Handoff completeness | Reviewed handoffs containing all locally required information divided by handoffs sampled | Which information is repeatedly missing? |
| Assessment completion | Completed required assessments divided by assessments due for the selected period | Where is training follow-up needed? |
| Corrective-action follow-through | Actions completed and reviewed by their agreed due date divided by actions due | Are known problems actually being addressed? |
Avoid combining these into a single performance grade. Each measure needs context and may call for a different response.
Define the numerator and denominator
The numerator is what you count. The denominator is the population or opportunity you compare it with. A rate becomes difficult to interpret when either changes without explanation.
For a rework rate, decide whether the unit is an instrument, a tray, a set, or a processing event. Decide whether one set with several issues counts once or several times. Write the rule before collecting data.
For a sample-based measure, make clear that the denominator is the sample reviewed, not the entire department's production. A sample can provide useful information without pretending to cover every item.
Use the same definitions over time. If you improve the definition, document the change so an apparent trend is not mistaken for a change in performance.
A hypothetical calculation
Suppose a team reviews 200 sets and finds that 12 meet its defined rework category. The observed rate in that sample is 12 divided by 200, or 6 percent.
During the next review, 8 of 200 sets meet the same definition, producing 4 percent. That is a difference of 2 percentage points in the observed samples.
Before calling it an improvement, ask whether the samples were comparable. Were the same kinds of sets reviewed? Did the sampling method change? Did staff apply the definition consistently? Was the observation period unusual?
These numbers are invented to explain the calculation. They are not ATS results, an industry average, or a target for your department.
Separate detection from creation of a problem
A team that reports more concerns may be finding existing problems more reliably. An increase in reports does not automatically mean the team created more defects.
Keep the detection stage visible. Distinguish concerns caught during internal review from reports received after release. Define the categories clearly enough that reviewers apply them consistently.
Ask whether a new reporting process made problems easier to document. If so, explain that change alongside the chart. The first few periods may reflect a change in visibility as well as a change in the work.
Avoid rewarding silence. A dashboard that makes staff feel punished for raising concerns can reduce the information leaders need to improve the process.
Check the data before debating the result
Choose a small sample and compare the dashboard entry with its source record. Look for duplicate reports, missing identifiers, inconsistent categories, and events assigned to the wrong period.
Ask two reviewers to categorize the same fictional examples. If they disagree, clarify the definition before using the measure to compare teams.
Keep a simple data dictionary: measure name, purpose, definition, source, exclusions, owner, and review frequency. It does not need to be long, but it should be accessible.
Be particularly careful when comparing shifts. Different case mixes, assignments, staffing, and review practices can produce different numbers without establishing that one team performs better.
Connect a measure to an observation
A chart can tell you where to look, but it rarely explains the cause by itself. If a category rises, observe the relevant workflow and speak with the people doing the work.
In a June 2026 Infection Control Today practice perspective on multidisciplinary rounding, Eduardo Velazquez describes bringing different operational partners into the review of department conditions. The useful idea is that a problem may require expertise beyond a single supervisor.
For your own review, invite the people who can clarify the suspected cause. A document issue may need the educator; an equipment issue may need service support; a supply issue may need purchasing.
Do not invite a large group without a defined question. A focused observation should produce a finding, an owner, and a next step.
Avoid claiming an infection outcome from a process metric
A complete record, an acceptable inspection result, and a lower rework rate can be useful findings. They do not by themselves establish a reduction in patient infections.
Describe what the data actually measure. If you reviewed documentation completeness, report documentation completeness. If you observed a workflow change, identify the scope and period of the observation.
Patient-outcome questions require the appropriate infection prevention and clinical surveillance expertise. Do not use a small local dashboard to imply a causal relationship that has not been evaluated.
This precision strengthens a resource request. Leaders can trust a modest claim supported by good evidence more readily than a dramatic claim with a missing link in the reasoning.
Treat targets as decisions that need a basis
A target can focus attention, but an arbitrary number can distort behavior. Before setting one, ask what requirement, baseline, capacity, or improvement aim supports it.
Some tasks may have explicit requirements established by policy or applicable instructions. Other measures describe a local improvement project. Keep those categories distinct.
If you choose a local target, label it as such. Explain why it was selected and when it will be reviewed. Do not present it as an industry standard unless you have a current source establishing that claim.
Also identify what the team should do when the measure falls outside expectations. A red cell without a response plan is a decoration, not a management tool.
A practical monthly review agenda
Use a brief meeting to connect the data with decisions. The following agenda is an original suggestion for local use.
- Confirm whether definitions or data sources changed.
- Review the current result and the recent pattern.
- Identify one finding that needs investigation or action.
- Check the status of the previous action.
- Assign an owner and the next review point.
- Decide what feedback staff should receive.
Keep a record of the decision and its reason. If no change is warranted, say why. A stable measure may need continued monitoring rather than a new project every month.
Avoid using the meeting primarily to explain away results. Ask what the evidence supports and what information is still needed.
Hypothetical example: a dashboard points to the wrong fix
A department sees an increase in reports categorized as missing instruments. The first proposal is to retrain assembly staff.
A review of the source records shows that several reports describe a mismatch between the current approved set contents and an outdated reference used by the receiving team. Other reports are genuine missing-item concerns.
The combined category hid two different problems. The team clarifies the reporting definition, coordinates the reference update, and investigates the remaining missing-item reports separately.
The lesson is not that training is unnecessary. It is that the measure should lead to a more precise question before the department selects a response.
This scenario is illustrative and does not describe ATS performance or a partner facility.
Include staff in choosing what to improve
Technicians often know where data entry is confusing or where a category fails to describe the work. Ask them to review the definitions and explain what the dashboard misses.
Show how their observations influenced a change. If a metric creates unnecessary duplicate documentation, examine whether an existing approved source can supply the information.
Keep individual personnel matters within the appropriate management process. A public dashboard should not become a ranking of employees based on unadjusted counts.
The goal is a shared understanding of work that needs attention. For the leadership habits behind that approach, see building a quality culture in sterile processing.
Start with one measure you can maintain
Choose one operational question, write a clear definition, and test the collection process on a small sample. Check whether the data are understandable and whether someone can act on the result.
Only add another measure when it answers a different useful question. More columns do not automatically produce more insight.
You can use a simple approved spreadsheet or an existing quality system. Protect access appropriately, avoid unnecessary patient information, and follow the organization's record requirements.
After a few review periods, ask whether the measure has changed a decision. If it has only created reporting work, revise it or retire it through the responsible process.
Frequently asked questions
What is a good sterile processing defect rate?
There is no universal benchmark established in this article. Definitions, units, case mix, sampling, and detection practices matter. Begin with a trustworthy local baseline and applicable requirements.
Is a lower incident-report count always better?
No. Reporting access and staff willingness to raise concerns can affect the count. Review the events and the reporting process before interpreting the change.
Should throughput be removed from the dashboard?
No. Work volume can help explain demand and capacity. Pair it with measures that address quality and the consequences of changing the pace of work.
Can a small department do this without new software?
Yes. Clear definitions and consistent review matter more than a complex display. Use an authorized tool the team can maintain reliably.
Turn the next number into a decision
Choose one measure you currently report and ask: what action would we take if this changed? If the answer is unclear, start by refining the question.
For help discussing department education and operational priorities, contact ATS about consulting. Bring a defined problem and the data you already have so the conversation can focus on a practical improvement.
