Application of indices Cp and Cpk to improve quality control capability in clinical biochemistry laboratories

The traditional criteria for acceptability of analytic quality may not be objective in clinical laboratories. To establish quality control procedures intended to enhance Westgard multi-rules for improving the quality of clinical biochemistry tests, we applied the Cp and Cpk quality-control indices to monitor tolerance fitting and systematic variation of clinical biochemistry test results. Daily quality-control data of a large Taiwanese hospital in 2009 were analyzed. The test items were selected based on an Olympus biochemistry machine and included serum albumin, aspartate aminotransferase, cholesterol, glucose and potassium levels. Cp and Cpk values were calculated for normal and abnormal levels, respectively. The tolerance range was estimated with data from 50 laboratories using the same instruments and reagents. The results showed a monthly trend of variation for the five items under investigation. The index values of glucose were lower than those of the other items, and their values were usually <2. In contrast to the Cp value for cholesterol, Cpk of cholesterol was lower than 2, indicating a systematic error that should be further investigated. This finding suggests a degree of variation or failure to meet specifications that should be corrected. The study indicated that Cp and Cpk could be applied not only for monitoring variations in quality control, but also for revealing inter-laboratory qualitycontrol capability differences.