Standard 33: Laboratory

Why it matters. An instrument that is out of calibration gives results that look normal and are wrong. Calibration kept current on every instrument that produces a reportable result is the basis of accurate results.

Sample. Review the calibration and maintenance records for all measuring devices for the last 12 months, and the traceability certificates.

1. Each device is calibrated or verified at the set interval.

Method. Check each device on the register against its calibration date and interval.

Evidence to show. Equipment calibration and traceability (LBT-POL-001).

Pass when. Every device is within its calibration interval.

2. Calibration is traceable to a reference standard.

Method. Check the certificate for each calibration for the traceable reference cited.

Evidence to show. Calibration and maintenance register (LBT-REG-001).

Pass when. Every calibration certificate names a traceable reference.

3. A device out of calibration is withdrawn from patient testing.

Method. Check each failed or overdue calibration for the withdrawal record.

Evidence to show. Calibration and maintenance register (LBT-REG-001).

Pass when. Every failed or overdue device has a withdrawal record.

Immediate fail (any one means the criterion is not met)

• A chemistry analyser used for patient results eight months after its calibration expired.

Why it matters. A new batch of reagent can behave differently from the last. Checking each new lot against the laboratory’s own quality control before use prevents results going wrong unnoticed.

Sample. Review the reagent lot verification records for 20 new reagent lots in the last 12 months.

1. Each new lot is checked before it is used on patients.

Method. Check each lot receipt for the verification entry dated before first use.

Evidence to show. Reagent lot verification record (LBT-FRM-001).

Pass when. Every sampled lot has a verification dated before first use.

2. The check follows the manufacturer’s acceptance criteria.

Method. Check each verification for the criteria and the result recorded.

Evidence to show. Reagent lot verification record (LBT-FRM-001).

Pass when. Every verification records the criteria and the result.

3. A lot that fails is not used.

Method. Check each failed lot for its removal from use.

Evidence to show. Reagent lot verification record (LBT-FRM-001).

Pass when. No failed lot appears in patient testing.

Immediate fail (any one means the criterion is not met)

• A new reagent lot used on patients without any check against the previous lot.

Why it matters. A method used for patient results must first be shown to work for this laboratory. Validation or verification before clinical use prevents results from an unproven method.

Sample. Review the method validation records for the last 5 tests introduced or changed in the last 2 years.

1. Each method is validated for its intended use before patient testing.

Method. Check each validation report for its date against the first patient result.

Evidence to show. Method validation before clinical use (LBT-SOP-001).

Pass when. Every method was validated before its first patient result.

2. The validation covers precision, accuracy, reportable range and reference interval.

Method. Check each report for these four elements.

Evidence to show. Method validation register (LBT-REG-002).

Pass when. Every report covers all four elements.

3. A method change triggers revalidation.

Method. Check each change in the register for a revalidation entry.

Evidence to show. Method validation register (LBT-REG-002).

Pass when. Every change has a revalidation entry before use.

Immediate fail (any one means the criterion is not met)

• A new assay put into clinical use on the manufacturer’s claims alone, with no local validation.

Why it matters. Quality control shows whether a test is working on the day. A result released before the control has been run and accepted may be wrong.

Sample. Review the internal quality control log for 30 analytical runs and the rules applied.

1. Quality control is run at the set interval and with each run.

Method. Check each run in the sample for the quality control entry.

Evidence to show. Internal quality control (LBT-SOP-002).

Pass when. Every sampled run has quality control recorded.

2. Control results are reviewed against defined rules before release.

Method. Check each run for the rule applied and the review recorded.

Evidence to show. Internal quality control log (LBT-REG-003).

Pass when. Every run records the rule review before release.

3. An out-of-control run is not released and is corrected.

Method. Check each out-of-control run for the hold and the corrective action.

Evidence to show. Internal quality control log (LBT-REG-003).

Pass when. Every out-of-control run is held and corrected before results are released.

Immediate fail (any one means the criterion is not met)

• Patient results released from a run where the control was outside its limit.

Why it matters. A laboratory that does not compare its results with others may have a problem it cannot see. External quality assessment shows how its results compare with peers.

Sample. Review the external quality assessment participation and results for the last 24 months for each test performed.

1. The laboratory takes part in an external scheme for each test where one exists.

Method. Check the participation record against the list of tests performed.

Evidence to show. External quality assessment participation (LBT-POL-002).

Pass when. Every test with a scheme has participation recorded.

2. Each unsatisfactory result is investigated and acted on.

Method. Check each unsatisfactory result for the investigation and the corrective action.

Evidence to show. External quality assessment results register (LBT-REG-004).

Pass when. Every unsatisfactory result has an investigation and action.

3. Results are reviewed at a set interval by the laboratory lead.

Method. Check the review entries for the set interval.

Evidence to show. External quality assessment results register (LBT-REG-004).

Pass when. Every result is reviewed at the set interval.

Immediate fail (any one means the criterion is not met)

• Three unsatisfactory external results for one test with no investigation.

Why it matters. A clinician who orders a test without knowing what it can and cannot tell them will misread the result. Clear written information on each test’s purpose and limits is part of a safe result.

Sample. Review the test catalogue and request forms for 15 tests, and 10 request forms completed in the last 3 months.

1. Each test states its indication and limitations.

Method. Check each catalogue entry for the indication and limitation text.

Evidence to show. Test request information sheet (LBT-FRM-002).

Pass when. Every catalogue entry states indication and limitation.

2. Patient preparation is stated for each test that needs it.

Method. Check each entry that needs preparation for the instructions given.

Evidence to show. Test request information sheet (LBT-FRM-002).

Pass when. Every test that needs preparation states it.

3. Clinicians receive the information at the point of request.

Method. Check each completed request form for the information attached or referenced.

Evidence to show. Test information for users (LBT-SOP-010).

Pass when. Every request form carries or refers to the test information.

Immediate fail (any one means the criterion is not met)

• A fasting test requested with no instruction given to the patient.

Why it matters. A specimen that is haemolysed, clotted or too small can give a wrong result. Written rejection criteria, applied the same way by everyone, stop it being tested.

Sample. Review the specimen receipt and rejection log for 30 rejected specimens and 30 accepted specimens in the last 3 months.

1. Each specimen is checked against the acceptance criteria on receipt.

Method. Check each receipt entry for the criteria checked.

Evidence to show. Specimen acceptance and rejection (LBT-SOP-003).

Pass when. Every receipt records the criteria checked.

2. Each rejection records the reason and the notice given to the requester.

Method. Check each rejection for the reason and the notice date.

Evidence to show. Specimen receipt and rejection log (LBT-REG-005).

Pass when. Every rejection records reason and notice.

3. Accepted specimens are logged and identified.

Method. Check each accepted specimen for the log entry and identifiers.

Evidence to show. Specimen receipt and rejection log (LBT-REG-005).

Pass when. Every accepted specimen is logged with its identifiers.

Immediate fail (any one means the criterion is not met)

• A specimen with no patient identifier accepted and tested.

Why it matters. Some specimens change in transit if they are too warm, too cold, or left too long. The transport conditions must keep the specimen as it was taken.

Sample. Observe specimen transport from collection to the laboratory for 10 deliveries and review the transport records.

1. Specimens are packed and labelled to preserve integrity.

Method. Observe the packaging and labelling of sampled deliveries.

Evidence to show. Specimen transport (LBT-SOP-004).

Pass when. Every sampled delivery is packed and labelled correctly.

2. Temperature and time limits are kept and recorded.

Method. Check each transport record for temperature and time against the limit.

Evidence to show. Specimen transport record (LBT-FRM-003).

Pass when. Every record is within its temperature and time limit.

3. Each deviation is recorded and the specimen is assessed.

Method. Check each deviation for the record and the decision on the specimen.

Evidence to show. Specimen transport record (LBT-FRM-003).

Pass when. Every deviation has a record and a decision.

Immediate fail (any one means the criterion is not met)

• A specimen left in a hot vehicle for three hours and tested without a deviation record.

Why it matters. An urgent (STAT) specimen that waits in the queue can delay treatment for a critical patient. A written procedure to flag, handle and track urgent specimens gets them to the front.

Sample. Review the STAT specimen log and the turnaround time records for the last 3 months.

1. STAT specimens are identified and prioritised on receipt.

Method. Check each STAT specimen for its identification and its priority in the workflow.

Evidence to show. STAT specimens and turnaround (LBT-SOP-005).

Pass when. Every STAT specimen is identified and prioritised.

2. Turnaround is measured against the defined targets.

Method. Check the turnaround log for each target against the actual time.

Evidence to show. Turnaround time monitoring log (LBT-REG-006).

Pass when. Turnaround is measured for every test against its target.

3. Targets missed are reviewed and acted on.

Method. Check each missed target for the review and action.

Evidence to show. Turnaround time monitoring log (LBT-REG-006).

Pass when. Every missed target has a review and action.

Immediate fail (any one means the criterion is not met)

• A STAT potassium result left in the queue for four hours with no alert.

Why it matters. A specimen used as legal evidence must show, at each step, who held it and what happened to it. Without that custody record, the evidence can be challenged.

Sample. Review the chain of custody forms for 10 forensic or legal specimens in the last 12 months.

1. Each specimen has a chain of custody from collection.

Method. Check each form for the first entry at collection.

Evidence to show. Chain of custody (LBT-SOP-006).

Pass when. Every sampled specimen has a form from collection.

2. Each transfer is signed and dated.

Method. Check each transfer on the form for a signature and date.

Evidence to show. Chain of custody form (LBT-FRM-004).

Pass when. Every transfer is signed and dated.

3. The specimen is disposed of only as recorded.

Method. Check the disposal entry against the form.

Evidence to show. Chain of custody form (LBT-FRM-004).

Pass when. Every disposal matches the recorded custody.

Immediate fail (any one means the criterion is not met)

• A legal specimen with a gap in custody between two staff members.

Why it matters. A critical result, such as a dangerously high potassium, must reach a responsible clinician at once. A result left in a report or on a screen can cost a life.

Sample. Review the critical value log for 30 critical results in the last 3 months.

1. Each critical value is communicated to a responsible clinician.

Method. Check each log entry for the recipient named.

Evidence to show. Critical value communication (LBT-SOP-007).

Pass when. Every critical value names a responsible recipient.

2. Communication happens within the defined time.

Method. Check each entry for the time from result to communication.

Evidence to show. Critical value communication log (LBT-REG-007).

Pass when. Every communication is within the defined time.

3. Receipt is verified and recorded.

Method. Check each entry for the read-back and receipt confirmation.

Evidence to show. Critical value communication log (LBT-REG-007).

Pass when. Every entry records receipt confirmed.

Immediate fail (any one means the criterion is not met)

• A critical potassium result left on a voicemail and not acted on.

Why it matters. A result released without being checked by qualified staff can carry an error to the clinician. A review and authorisation before release is the last check.

Sample. Review the authorisation records for 30 released results and the competence records of the authorising staff.

1. Each result is authorised before release by a qualified person.

Method. Check each result for the authoriser and the time before release.

Evidence to show. Result authorisation and release (LBT-POL-003).

Pass when. Every released result has an authoriser before release.

2. The authoriser is competent for the test.

Method. Check the authoriser against the competence register for the test.

Evidence to show. Result authorisation record (LBT-FRM-005).

Pass when. Every authoriser is competent for the test.

3. The authoriser is recorded with the result.

Method. Check each result record for the name recorded.

Evidence to show. Result authorisation record (LBT-FRM-005).

Pass when. Every released result shows the authoriser.

Immediate fail (any one means the criterion is not met)

• A result released by a trainee with no authorisation.

Why it matters. A report without the patient’s identity, the specimen type or the test date cannot be used safely. Every report must carry a defined minimum of information.

Sample. Review 20 laboratory reports and the delivery records for the last 3 months.

1. Each report contains the defined minimum information.

Method. Check each report against the minimum content list.

Evidence to show. Laboratory report content and delivery (LBT-SOP-008).

Pass when. Every report contains all the minimum content.

2. The report reaches the correct recipient.

Method. Check each delivery against the requester recorded.

Evidence to show. Laboratory report content and delivery (LBT-SOP-008).

Pass when. Every report reaches the requester named.

3. Reports carry the authoriser and reference interval.

Method. Check each report for the authoriser and the reference interval.

Evidence to show. Laboratory report content and delivery (LBT-SOP-008).

Pass when. Every report carries both.

Immediate fail (any one means the criterion is not met)

• A report sent to the wrong ward, with another patient’s result on it.

Why it matters. A result that is changed after release must show what it was and what it became, and why. Without that, the change cannot be trusted or reviewed.

Sample. Review the result amendment log for all amendments in the last 12 months.

1. Each amendment is recorded as an amendment to the original result.

Method. Check each amendment for the link to the original result.

Evidence to show. Result amendments (LBT-SOP-009).

Pass when. Every amendment is linked to the original.

2. Each amendment states its reason, date and person.

Method. Check each amendment entry for the reason, date and person.

Evidence to show. Result amendment log (LBT-REG-008).

Pass when. Every amendment states reason, date and person.

3. The original result remains visible in the record.

Method. Check the record for the original value kept.

Evidence to show. Result amendment log (LBT-REG-008).

Pass when. The original value is kept visible in every sampled case.

Immediate fail (any one means the criterion is not met)

• A released result overwritten with no record of the original value.

Why it matters. Reagents, consumables and outside services that are poor quality cause results to be wrong. Buying only from qualified suppliers, and monitoring them, keeps quality up.

Sample. Review the supplier qualification register and the purchase records for all reagent and service suppliers in use.

1. Each supplier is qualified before first use.

Method. Check each supplier on the register for the qualification date before the first purchase.

Evidence to show. Supplier and reagent qualification (LBT-POL-004).

Pass when. Every supplier was qualified before first use.

2. Qualification is reviewed at the set interval.

Method. Check each supplier for the review date against the interval.

Evidence to show. Supplier qualification register (LBT-REG-009).

Pass when. Every supplier is reviewed at the set interval.

3. Unqualified suppliers are not used.

Method. Check the purchase records against the register.

Evidence to show. Supplier qualification register (LBT-REG-009).

Pass when. No purchase is made from an unqualified supplier.

Immediate fail (any one means the criterion is not met)

• Reagents bought from an unlisted supplier with no quality documents on file.