QA/QC In Case Questions

Levey-Jennings And The Westgard Multirules

Following quality assurance protocols is explicitly tested, and the centerpiece is statistical quality control. A Levey-Jennings chart plots each control result against the established mean, with horizontal lines at +/-1, 2, and 3 standard deviations (SD). A run is judged by how the control points fall relative to those limits. The Westgard multirules, published by Dr. James Westgard in 1981, formalize this judgment. Know each rule, what it detects, and whether it warns or rejects:

The 1-2s rule is a warning only; it does not by itself reject a run but prompts you to check the dedicated rejection rules. A common trap option is to reject a run on a single 2 SD point, which violates the warning-rule concept. The remaining rules are rejection rules: when violated, you stop, do not report patient results, troubleshoot, and rerun the controls before releasing data.

The statistical reasoning behind the chart is itself testable. Each control material carries an assigned mean and SD established from at least 20 runs over multiple days, and roughly 68% of in-control values fall within +/-1 SD, 95% within +/-2 SD, and 99.7% within +/-3 SD of the mean. That is why a single point beyond 2 SD is only a 1-in-20 expected event (the basis of the 1-2s warning), but a point beyond 3 SD is rare enough (about 3 in 1000) to justify outright rejection by the 1-3s rule.

Multilevel QC strengthens this: running two controls (for example a normal and an abnormal level) on each run lets rules like 2-2s and R-4s compare across levels, catching bias and imprecision that a single control would miss. Knowing the probability logic helps you reason about borderline patterns rather than memorizing rule names blindly.

Random Versus Systematic Error And Troubleshooting

The most testable distinction is random versus systematic error. Random error is imprecision, scatter that worsens unpredictably, flagged by 1-3s and R-4s. Systematic error is bias, a consistent offset detected as a shift or trend, flagged by 2-2s, 4-1s, and 10-x. Recognize the chart patterns:

• A shift is an abrupt move of consecutive points to one side of the mean (often after a reagent lot change, recalibration, or a new control vial).
• A trend is a gradual drift in one direction (often deteriorating reagent, an aging light source or electrode, or a slow temperature change).

Tie pattern to cause. Six consecutive points jumping to the +1 SD side immediately after opening a new reagent lot is a shift suggesting a calibration or lot problem; the action is to recalibrate and verify with fresh controls. A slow climb of control values over two weeks is a trend suggesting reagent degradation or instrument drift; the action is to investigate the reagent and instrument, not simply repeat the control once.

QA/QC also extends to the broader quality framework, the small Laboratory Operations content area (roughly 5-10% of the guideline). Distinguish the layers: accuracy is closeness to the true value (assessed against a reference or proficiency target), while precision is reproducibility (assessed by repeated measurements and the coefficient of variation, CV = SD/mean x 100). Proficiency testing (PT) under CLIA evaluates a lab against peers using blind samples; PT failures and unresolved QC are reportable quality events.

Calibration verification confirms the analytical measurement range, and delta checks (covered earlier) guard individual patient results.

A worked QA/QC case: the level 1 control reads +3.5 SD while level 2 reads -0.5 SD on the same run. The range between them exceeds 4 SD, violating the R-4s rule and indicating random error. The correct procedural answer is to reject the run, withhold patient results, investigate (check pipetting, bubbles, a clot, or a degraded control), and rerun. Selecting "accept and report because one control passed" is wrong because a single rejection-rule violation invalidates the entire analytical run, regardless of how the other control behaves.

QA/QC items also test the lifecycle around the run itself. When a run is rejected, the corrective sequence is to identify whether the error is random or systematic, fix the assignable cause, document the corrective action, and verify acceptable controls before releasing any held patient results. If patient results were already reported before the failure was caught, the protocol requires evaluating affected patients and issuing corrected reports as needed.

Beyond daily QC, recognize the broader quality vocabulary the exam draws from: preventive maintenance schedules, temperature logs for refrigerators and water baths, proficiency testing cycles under CLIA, and lot-to-lot verification of new reagents against the existing lot. A stem describing a new reagent lot installed without verification, followed by a systematic shift, is testing whether you connect the lot change to the QC failure and prescribe verification as the fix rather than blaming the instrument.