Quality Control (QA/QC): Automated Control Charts and Analytical Trend Detection

1. Why quality control isn’t optional in a water testing lab

When a laboratory reports that a drinking water sample contains 0.08 µg/L of PFAS — just under the 0.10 µg/L parametric value set by the EU Drinking Water Directive 2020/2184 or the US EPA’s PFAS limits under the Safe Drinking Water Act — the line between compliance and non-compliance depends entirely on whether the analytical process is reliable. An undetected 25% error would push that same result to 0.10 µg/L, crossing the regulatory limit.

Validating the reliability of results is the mechanism that lets a laboratory demonstrate its analytical processes are under statistical control, and that reported results are trustworthy within their declared uncertainty. Without it, every result issued is an act of faith, not a scientific data point.

2. The pillars of QA/QC under ISO/IEC 17025

2.1 Control charts: the Shewhart chart

A Shewhart chart is a time-series plot of a control parameter — typically the result of analyzing a reference material — on the vertical axis, against time on the horizontal axis. It includes three zones: the center line (mean of the control values), warning limits (usually ±2 standard deviations from the mean), and action limits (±3 standard deviations).

As long as points scatter randomly around the center line within the warning limits, the process is in control. When a point crosses the action limit, or when a non-random pattern appears (for example, 7 consecutive points above the mean), the system flags a loss of control that requires immediate investigation.

2.2 Westgard rules: the language of QC

The Westgard rules are the de facto standard for interpreting control charts in clinical and environmental laboratories. Each rule catches a specific type of error:

1-3s (reject): A control result exceeds ±3 standard deviations. Signals a serious error, random or systematic. The batch must be rejected and repeated.

2-2s (warning): Two consecutive control results exceed ±2 standard deviations in the same direction. Suggests an emerging systematic error (bias).

R-4s: The difference between two controls in the same batch exceeds 4 standard deviations. Indicates increased method imprecision.

10x: Ten consecutive control results fall on the same side of the mean. Even if none crosses a limit, the pattern signals a systematic drift that needs investigating.

2.3 Proficiency testing

Proficiency testing programs are interlaboratory comparisons in which an accredited provider sends identical samples to multiple laboratories. Results are evaluated statistically to determine whether each laboratory’s values are comparable to the consensus. Accreditation bodies operating under the ILAC MRA — including A2LA and NELAP in the US, and UKAS in the UK — require regular participation in these programs as a condition of accreditation.

An unsatisfactory proficiency test result doesn’t automatically mean losing accreditation, but it does trigger a root-cause investigation, documented corrective actions, and often a repeat round with a new reference material.

2.4 Trend analysis

Trend analysis is the tool that closes the QA/QC loop. It involves evaluating how control data evolves over time to catch progressive drift before it crosses action limits — the difference between reacting to a problem and getting ahead of it.

3. How Zendo LIMS automates QA/QC

3.1 Automatic generation of control charts

Zendo LIMS generates Shewhart charts automatically from the results of analyzed reference materials. The system calculates the mean, standard deviation, and warning and action limits, and plots the data in charts that update in real time.

This eliminates the common practice in labs without a LIMS of building control charts in spreadsheets at month- or quarter-end, by which point the ability to react to a deviation has already been lost.

3.2 Configurable validation rules

The system lets labs configure Westgard rules alongside other lab-specific validation rules. When a control result breaks a rule, Zendo LIMS automatically alerts the analyst and the quality manager, halting the workflow until the deviation is investigated and resolved.

3.3 Trend alerts

Beyond the Westgard rules, Zendo LIMS includes drift-detection algorithms that spot drift patterns even before a control rule is broken — turning QA/QC from a reactive system (catching the problem) into a proactive one (anticipating it).

3.4 Proficiency testing records

Proficiency testing results are logged in Zendo LIMS and linked to the relevant method and parameter. This gives the quality manager a consolidated view of the lab’s performance across interlaboratory comparisons, making it easier to identify the methods with the most variability and prioritize improvement actions.

4. QA/QC and the new generation of contaminants

Regulations such as the EU Drinking Water Directive 2020/2184 and the US EPA’s PFAS regulations have introduced parameters that demand extremely high-sensitivity analytical techniques, such as quantifying PFAS by LC-MS/MS at ng/L levels. At these concentrations, analytical variability is proportionally larger, and quality control becomes critical.

Laboratories analyzing PFAS face specific QA/QC challenges: the risk of cross-contamination during sampling (PFAS are present in gloves, containers, caps, and reagents), the need for extremely low detection limits, and the requirement for isotopic standards in quantification. A robust QA/QC system, automated through the LIMS, is the only reliable way to manage this complexity.

5. The cost of skipping QA/QC (or running it in Excel)

Undetected invalid results: Without real-time control charts, a laboratory can issue results for weeks without realizing its method has drifted out of calibration. The cost: report retractions, repeat analyses, and lost client trust.

Accreditation audit findings: Missing or outdated control charts, or control data that was never evaluated, are among the most common nonconformities cited in accreditation assessments.

False alarms and excess repeats: Without correctly configured Westgard rules, labs tend to either rerun analyses unnecessarily — a direct cost in reagents and time — or, worse, ignore genuine warning signals.

6. Conclusion: QA/QC as the lab’s operational intelligence

Quality control isn’t a bureaucratic add-on to analytical work — it’s the lab’s nervous system. Control charts, Westgard rules, proficiency testing, and trend analysis are the tools that let a laboratory know, at any given moment, whether its results can be trusted.

A LIMS like Zendo LIMS automates these tools, turning scattered data into operational intelligence available in real time. The question is no longer whether your laboratory does QA/QC. The question is whether it does so in time to act before an out-of-control result reaches the client report.