Glucose Quality Control Failure Before Releasing Patient Results
On a Monday morning in a primary health centre laboratory, the glucose analyser has just been calibrated with a new reagent lot. Two levels of quality control material are run: Level 1 is 96 mg/dL against a mean of 100 mg/dL and SD of 5; Level 2 is 278 mg/dL against a mean of 250 mg/dL and SD of 10. Yesterday, before the lot change, Level 2 was 272 mg/dL. The laboratory information system flags no rule violation because only one control level is outside ±2 SD. Fifteen fasting patient samples are waiting, including a pregnant woman with a glucose of 180 mg/dL whose clinician has called twice. The senior technician says, 'It is only 28 mg/dL high; release the results and recalibrate tomorrow.' The analyser has no back-up, the calibration was accepted without a repeat control, and the QC chart has not been updated. The laboratory has a written QC policy requiring investigation of any control outside ±2 SD before patient reporting.
1. Interpret the QC results using the available numbers and decide whether the patient results may be released. Show your reasoning.
Read the worked answer
Level 1: 96 mg/dL against mean 100 and SD 5 is (96 - 100)/5 = -0.8 SD, which is acceptable.
Level 2: 278 mg/dL against mean 250 and SD 10 is (278 - 250)/10 = +2.8 SD. That is outside ±2 SD. Yesterday's Level 2 value was 272 mg/dL, or +2.2 SD. Two consecutive control results on the same level exceed +2 SD, so the run violates the 2_2s rule and indicates a systematic shift, not a random warning. The fact that the software did not flag it does not make it acceptable; the written policy requires investigation.
The run is therefore out of control. Patient results from this run should not be released because a positive bias of about 28 mg/dL at the upper level could change clinical decisions, especially in a pregnant woman. The appropriate action is to stop reporting, investigate the new reagent lot and calibration, recalibrate, repeat both control levels and then rerun the patient samples. If the clinician cannot wait, the laboratory should explain the delay and, with laboratory director approval, arrange testing on a validated alternative method or referral laboratory rather than release unverified results.
2. What investigation and corrective action should be completed before rerunning the patient samples?
Read the worked answer
Check the pre-analytical and analytical chain systematically:
- Controls: reconstitution date, storage, expiry, evaporation and lot number; repeat with a fresh vial.
- Reagent: new lot number, expiry, storage temperature, preparation and lot-to-lot validation record; compare the calibration curve with the previous lot.
- Analyser: calibration acceptance, sample and reagent probes, tubing, lamp or detector, temperature and wash steps; run a maintenance check.
- Calibration: was it performed correctly, and was a repeat control run immediately? If not, recalibrate and repeat both levels.
- Environment and operator: water quality, power stability and whether the same operator prepared both control and reagent.
After recalibration, both control levels must be within acceptable limits before patient samples are rerun. All patient samples from the affected run should be repeated, not just the urgent one. If controls remain outside limits, the analyser should be taken out of service and the laboratory director informed. Record the investigation, corrective action and verification results in the QC log. The expired or missing repeat control after calibration is itself a process failure requiring a corrective-action note.
3. The clinician demands the pregnant woman's glucose immediately. How should you respond professionally, and what ethical and documentation issues arise?
Read the worked answer
Acknowledge the clinical urgency and explain clearly: 'The glucose analyser failed quality control this morning; results from that run may be falsely high. I cannot release them until the analyser is recalibrated and controls pass. I am repeating the sample now and will call you as soon as it is verified.' Offer an alternative if available, such as a validated method or referral, and give an expected time.
Do not alter, hide or override the QC result, and do not report a patient result with a false 'normal' comment. If the laboratory director authorises release under a documented clinical-urgency policy, the report must state the QC limitation and the clinician must be informed and the decision recorded.
Document the clinician's calls, time, advice given, staff member spoken to and the final result. Preserve the original QC data. Ethically, the duties are patient welfare, professional competence, integrity of results, transparency and accountability. Discussing the patient's result in a corridor or with the relative would breach confidentiality. A QC incident report and root-cause review should follow, because 'release and recalibrate tomorrow' is not an acceptable quality decision.