Sigma Metrics and QC Design
At the monthly QC review, the chemistry director presents the following method performance from the previous 20 days: glucose TEa 10.0%, bias +1.2%, CV 1.8%; creatinine TEa 15.0%, bias +2.8%, CV 4.6%; sodium TEa 4.0%, bias +0.8%, CV 1.0%. The laboratory currently applies a single 1:3s rule with two QC levels to all three analytes. During the same period, creatinine triggered 1:2s twice in 10 days; no 1:3s occurred, and patient results were released after each event. The creatinine reagent lot changed six weeks ago, and the analyzer photometer was serviced four months ago. The director wants to know whether the current QC design is defensible and whether patient results need review.
1. Calculate sigma for each analyte and recommend a QC strategy. Is the single 1:3s rule adequate?
Read the worked answer
Sigma = (TEa − |bias|) / CV.
- Glucose = (10.0 − 1.2) / 1.8 = 8.8 / 1.8 = 4.89.
- Creatinine = (15.0 − 2.8) / 4.6 = 12.2 / 4.6 = 2.65.
- Sodium = (4.0 − 0.8) / 1.0 = 3.2.
A single 1:3s rule is not adequate for creatinine or sodium. At sigma below 3, even a 1:3s rule has poor error detection, and two control measurements give little protection. Glucose at about 4.9 is moderate and needs a multirule with more measurements (for example 1:3s, 2:2s, R:4s, and 4:1s) plus review. Creatinine needs maximum QC, three levels, and method improvement before relying on QC alone. Sodium also needs a multirule and investigation of bias and precision.
2. Interpret the two creatinine 1:2s events. What immediate patient-risk actions and investigations are warranted?
Read the worked answer
1:2s is a warning rule, not a rejection rule. Releasing results after a 1:2s without a rejection rule is defensible only if the QC chart and patient results show no systematic shift. Two events in 10 days on a method with sigma 2.65 is a signal, not a coincidence.
Immediate actions: review QC charts for trend or shift; check the new reagent lot, calibrator, photometer, water, and maintenance; repeat the most recent patient samples; review creatinine deltas; and notify the medical director if a shift is confirmed. Document the investigation and determine whether a targeted look-back is needed. Do not continue to release results on the assumption that 1:2s is always harmless.
3. Design a 90-day improvement plan for creatinine and describe how you would measure effectiveness.
Read the worked answer
90-day plan: (1) root cause-lot-to-lot bias, calibration, photometer, or sample handling; (2) reduce bias and CV by recalibrating, servicing, changing lot, or reviewing parameters; (3) recalculate sigma; (4) set QC rules according to the new sigma; (5) train staff and document.
Effectiveness measures: sigma ≥4, CV and bias within target, QC rejection rate, repeat rate, turnaround time, and no unexplained patient deltas. Review monthly with run charts. If sigma remains below 3 after the plan, change method or vendor; QC alone cannot compensate for a method that is not fit for the clinical requirement.