Performance Verification of CardioChek PA Lipid Point-of-care Devices

  Objective  To validate the accuracy of CardioChek PA lipid point-of-care devices in determining total cholesterol (TC), triglyceride (TG), and high-density lipoprotein cholesterol (HDL-C) in whole blood and the comparability with results determined by full-automatic biochemical analyzer.

  Methods  We determined the low, medium, and high levels of TG, TC, and HDL-C in whole blood with single reagent lot number 20 times using 3 different CardioChek PA devices to evaluate inter-run and devices' coefficient of variations(CVs), and used single CardioChek PA device with 3 reagent lot number to determine the whole blood 10 times for evaluating inter-run and total CVs. Fifty-four volunteers whose lipid profiles covered up high, medium, and low levels were recruited. Lipids in fasting periphery whole blood and venous serum were collected and determined using 3 different CardioChek PA devices and 4 kinds of full-automatic biochemical analyzers, respectively. Bland-Altman plot was made to analyze the comparability of results from CardioChek PA and the full-automatic biochemical analyzers. Linear regression was analyzed using results of CardioChek PA and full-automatic biochemical analyzers. Bias and percentage bias were determined between CardioChek PA and different automatic biochemical analyzers; meanwhile, whether they satisfied the requirements of medical decision levels was determined.

  Results  The total CVs of TC in low, medium, and high levels in 3 different devices were 2.69%, 4.88%, and 3.51%, respectively; for TG, they were 5.51%, 5.27% and 4.96%; and for HDL-C, they were 7.27%, 6.84% and 6.79%. The total CVs of TC, TG, and HDL-C determined with the same device but different reagent lot number were 4.70%, 7.66%, and 8.61%, respectively. Comparison of the results from CardioChek PA devices and the 4 kinds of full-automatic biochemical analyzers showed lowest deviation for TC with -2.21%-2.56%, and for HDL-C with -1.12%-5.57%; the deviation of TG results of BeckmanDxC800 was relatively high with 25.85%, but in other 3 systems were -4.55%-13.34%. Deviation of TC, TG, and HDL-C in different medical decision levels were -3.27%-1.96%, -11.05%-13.06%, and -5.86%-11.56%, respectively, all of which could satisfy the requirements of the National Cholesterol Education Program(NCEP). Compared with reference methods, the biases of TC in medical decision levels were 1.96% and 0.77%; for HDL-C, they were 2.34% and 4.87%.

  Conclusion  The accuracy of CardioChek PA lipid point-of-care device can satisfy the clinical requirements, and the device can be used in the screening and monitoring of dyslipidemia.