High and low quality control tests should be run at least once every day of patient testing, and again any time something changes in the testing process: a new reagent lot, a calibration, major maintenance, or an unexpected patient result. The exact frequency depends on the test’s complexity, the potential harm from an error, and your laboratory’s workload.
Daily QC as the Baseline Standard
Running both a high and low level control at least once per day of testing is the most widely accepted minimum. This confirms your instrument is producing accurate results across the reportable range before patient samples are analyzed. Many laboratories run controls at the start of each shift or at the beginning of the day’s testing, then repeat them at regular intervals if testing continues around the clock.
That said, daily QC is a floor, not a ceiling. The right frequency varies by analyte. High-sensitivity troponin, for example, carries catastrophic consequences if results are wrong because it drives decisions about heart attacks. Professional organizations recommend running three levels of QC material at least once daily for that test. A routine chemistry like creatinine, where a small error is less immediately dangerous, may not need the same intensity. The principle is straightforward: the greater the potential harm from a wrong result, the more often you should run QC.
Events That Trigger Immediate QC Testing
Beyond your scheduled daily runs, certain events require you to run high and low controls before reporting any more patient results. These triggers exist because each one introduces a new variable that could shift your results:
- New reagent lot: Every time you open a shipment or switch to a different lot number, QC must pass before you use those reagents on patients.
- Calibration or recalibration: After calibrating an instrument, controls verify that the calibration actually produced accurate results.
- Major maintenance or repair: Any significant service work, such as replacing a lamp, cleaning an optical system, or fixing a mechanical issue, requires QC verification afterward.
- Unexpected patient results: When a result doesn’t match the clinical picture, running controls helps determine whether the instrument is the problem or the result is genuinely unusual.
- Equipment malfunction: Even after a brief error message or power interruption, controls confirm the analyzer has returned to normal function.
Think of these as “prove it still works” moments. Any time the testing system has been disrupted or altered, you need fresh evidence that it’s performing correctly.
Why You Run Both High and Low Levels
A single control level only tells you the instrument is accurate at one concentration. Patient samples span a wide range, and an analyzer can drift differently at different concentrations. A low control checks performance near decision points for ruling out a condition, while a high control checks performance in the abnormal or critical range. Running both levels together catches problems that a single level would miss entirely.
Some tests call for a third, mid-range control as well. This is common for analytes where clinical decisions happen across multiple thresholds. The key point is that every level of control must pass before you release patient results.
How to Tell If QC Has Passed or Failed
Most laboratories use a set of interpretation rules, often called Westgard rules, to evaluate control results. These rules look at how far your control value falls from the expected mean, measured in standard deviations (SD). A few of the most common:
- 1-3s rule: A single control result more than 3 SD from the mean fails the run. This catches large, sudden errors.
- 2-2s rule: Two consecutive results both exceeding 2 SD in the same direction indicate a systematic shift. The run fails.
- R-4s rule: One control is more than 2 SD high and the next is more than 2 SD low (or vice versa), with a total spread of 4 SD. This pattern signals random error.
- 10x rule: Ten consecutive results all fall on the same side of the mean. Even if each individual result looks acceptable, this trend indicates a gradual drift that needs correction.
When any rule triggers a failure, you stop reporting patient results, investigate the cause, correct the problem, and rerun controls until they pass. Any patient samples tested since the last acceptable QC run may need to be retested.
Point-of-Care Devices Follow Different Schedules
Point-of-care testing (POCT) devices, such as glucose meters and handheld blood gas analyzers used at the bedside, follow the same core principles but with adjusted schedules. Central lab analyzers typically run multi-level QC throughout the day combined with monthly external quality assessments and regular maintenance. POCT devices may not run continuously, so their QC timing is tied more to events than to a clock.
At minimum, QC on a point-of-care device should be performed when a new delivery of reagents arrives, when the reagent lot number changes, when a patient result seems inconsistent with the clinical situation, and after any major maintenance or repair. Some programs structure this into a regular cycle. Australia’s national POCT quality program, for instance, requires operators to run two levels of QC in the first two weeks of each month and two external quality samples in the second two weeks.
Regulatory Requirements in the U.S.
Under the Clinical Laboratory Improvement Amendments (CLIA), laboratories performing moderate or high-complexity testing must follow specific quality control requirements. CLIA also requires enrollment in proficiency testing programs, with at least three testing events spaced at roughly equal intervals per year for most analytes. For tests not covered by formal proficiency testing, laboratories must verify accuracy at least twice annually through alternative methods.
CLIA offers some flexibility through a program called the Individualized Quality Control Plan (IQCP). This allows a laboratory to customize its QC frequency based on a risk assessment that considers the test method, the testing environment, and personnel competency. An IQCP doesn’t eliminate QC. It lets you tailor the plan so your resources focus where the risk is greatest, while still providing equivalent quality assurance. The process requires documented risk assessment and must be approved through your laboratory’s accreditation pathway.
International Standards and ISO 15189
Outside the U.S., many laboratories follow ISO 15189 for accreditation. This standard requires quality indicators across all phases of testing, from sample collection through result reporting, but it does not prescribe a specific QC frequency. Instead, it requires each laboratory to establish its own quality indicators, define measurement goals and action plans, set limits for acceptable performance, and periodically review the system to confirm it remains appropriate.
An international comparison program run by the International Federation of Clinical Chemistry and Laboratory Medicine gives participating laboratories a way to benchmark their error rates and quality indicators against peers worldwide. This external comparison functions similarly to proficiency testing, helping laboratories identify drift or systematic problems that internal QC alone might not reveal.
Building a Practical QC Schedule
Your QC schedule should combine time-based and event-based testing. At minimum, run high and low controls once every 24 hours of patient testing. Add QC runs after every reagent change, calibration, maintenance event, or instrument malfunction. If your laboratory processes high volumes or runs critical tests like troponin or blood gases, consider running controls at the start of each shift or more frequently throughout the day.
Track your QC data over time using Levey-Jennings charts or your analyzer’s built-in QC module. Patterns of gradual drift are just as important as outright failures. A control that consistently trends upward over several days, even while technically still “in range,” signals a developing problem worth investigating before it triggers a failure and delays patient results.