When establishing quality control limits, which practice is considered inappropriate?

Using exclusion of any quality control results greater than ±2 standard deviations from the mean is inappropriate when establishing quality control limits because it can lead to a biased representation of the assay's performance. Quality control (QC) results that fall outside this range can contain valuable information regarding the assay's reliability and capability. Every QC data point plays a role in understanding the variability and potential issues in the testing process.

Establishing quality control limits ideally should consider all relevant data, as eliminating results arbitrarily may mask systematic errors or biases in the results. Instead, QC limits should be set based on an analysis of all data, including points that are outside the expected range, to ensure a thorough assessment of the method's performance and stability over time. This comprehensive approach allows for more accurate determination of the assay's characteristics and helps to maintain the integrity of laboratory testing.

In contrast, using last month's QC data to determine current target limits can provide insight into recent performance, using control results from all shifts gives a broader view of variability, and employing limits determined by reference laboratories ensures consistency with accepted standards. All these practices contribute to building a more robust quality control framework than simply excluding outliers.