Critical results under ISO 15189: procedure and records
A critical result is a laboratory result that deviates so far that the doctor must act immediately. Under ISO 15189, your laboratory must have a fixed procedure to get those results quickly and reliably to the right person. The procedure has three parts: what counts as critical, how you pass it on and how you record it. An assessor from the RvA can check all three parts. Make sure you can show all three at any time.
The subject falls under the post-analysis phase in the requirements of ISO 15189. If you want to see which phase that is exactly, first read what ISO 15189 is. The procedure also belongs in the quality manual.
Short answer: a critical result must reach the doctor or nurse, and the laboratory must be able to show who reported it, to whom, when and how. A result that is only in the system has not been passed on.
What counts as critical?
Each laboratory decides for itself which values are critical, in consultation with the specialists. These are usually values that are immediately dangerous to the patient, such as a very high or very low potassium level. The list belongs in your procedure, with the limits per test and per age group where needed.
For the list, you can use the guidelines of your professional group, such as the minimum standards in the NVKC (Dutch Society for Clinical Chemistry and Laboratory Medicine) sector guidelines. The RvA expects you to justify that choice and to update the list when new insights arise.
The procedure step by step
The procedure describes what happens from the moment the analyst sees the critical value until the doctor confirms the notification.
The analyst recognises the critical value
The system flags the result automatically, or the analyst checks the list. The analyst starts the notification straight away.
The notification goes to the right person
Phone the requesting doctor or the duty nurse. Ask for the name of the recipient and check the patient number. Choose a route where you are sure the recipient will see the result. An email that sits unread in an inbox does not count.
The recipient confirms
The recipient reads the result back. The analyst notes the name and the time of the confirmation. If the notification fails, start a second attempt straight away through a fixed route.
The notification is recorded
Release the result according to your own procedure. The file shows the critical value, the reporter, the recipient and the times.
Follow-up if no contact is made
If there is no confirmation after a set time, an escalation follows. Your procedure states who does this and within what time limit. Also record the escalation.
The steps are an example. Your procedure may differ, as long as you can demonstrate that the result reached the right person.
Example of a record
An assessor wants to be able to trace the trail of one critical result. Below is a fictional example of a record.
| Field | Example |
|---|---|
| Patient | Patient number 100245 |
| Test and value | Potassium, 6.8 mmol/l (fictional) |
| Reporter | Analyst M. Jansen, shift 14:00 to 22:00 |
| Recipient | Dr A. Peters, GP, by phone |
| Time of notification | 15:42 |
| Time of confirmation | 15:44, result read back |
| Recorded in the file | 15:45 |
| Escalation | Not needed |
With this one line, you can reconstruct the whole chain. Pay attention to the times: the gap between notification and confirmation tells you something about the speed of your process.
What does the assessor ask?
An assessor often picks a critical result from the past year and asks for the file. That is why you must be able to find the records for each result directly.
They also want to know whether staff know the procedure. Ask an analyst who was on shift to explain the steps, without the document in front of them. If that works, the procedure is embedded.
A third point is evaluation. Did you miss a critical result or pass it on too late last year? Then a non-conformity belongs with it, with a cause analysis. The root cause analysis describes that approach, and a structural problem calls for a CAPA process.
Common mistakes
Only a system message. A critical result that appears in the laboratory system but was not phoned through has not been passed on. Record the telephone or in-person notification.
No second route. If the doctor cannot be reached, the procedure must say what you do then. A procedure without escalation leaves a gap in the process.
A list that is never updated. Check the critical values regularly and record when you last reviewed them.
Records filled in afterwards. An assessor looks at the timestamps. Complete the form at the moment itself, not later from memory.
What should the procedure contain?
A good procedure answers six questions without the reader having to guess. What counts as critical? Who reports? To whom is it reported? How is it reported? How is confirmation obtained? What happens if nobody can be reached?
Fill in each answer concretely. Write “the duty analyst phones the requesting doctor or the duty doctor” instead of “the right person is informed”. The second version sounds better, but an assessor cannot check anything with it.
Do not forget the time limit either. How quickly must the notification follow the finding, and when is a result too late? That limit belongs in the procedure and in the system.
Who is responsible for the list?
The list of critical values is the responsibility of the laboratory, in consultation with the specialists. Record who drafts the list, who approves it and how often it is reviewed. Every change to a method or device means the list must be checked again.
This is a shared responsibility. The laboratory sets the limit, but the specialists help determine its clinical meaning. Show that in the minutes, so the assessor can see the link.
How do you test the procedure?
Practise the process. Once a quarter, run through a simulated critical result and measure how long the notification takes. Write the time down. That shows whether the procedure also works in practice, not only on paper.
Use the outcome of that exercise as input for the management review. The discussion of the process belongs there.