PEPconnect

PEPFAR Quality Contol and Method Validation Activity 7 Job Aid: How to Calculate TE

This PDF contains supporting material discussed in Activity 7. This information can also be found in the Participants Manual.

Job Aid: How to Calculate TE 705 Effective Date: 12/6/2016 HILS1746 I. Need 3 key numbers II. Selecting the True Value (Target Value) a. Mean a. The value published in the control's package insert for that lot 1) A fact calculated from observed data points number 1) Limitations 2) Will change due to: May not be stated in insert TO a) Recalibration b) May be based on insufficient data performed by b) Introduction of reagent or calibrator manufacturer lot number changes b. The value obtained from an interlaboratory comparison program C) Changes to the instrumentation 1) Based on your peer group performing the same method b. SD for the test 1) A fact calculated from observed data points Limitations a 2) Certain amount of imprecision is expected; Peer group may be too small to accurately reflect a true value `inherent with the analytical system b) Outlier laboratories in a small peer group will c. True Value - target value statistically affect the calculations 1) Used with the mean to determine bias An interlaboratory program may not be available by the manufacturer 2) Selected from the most appropriate source C. Cumulative mean from your laboratory's data when the or combination of sources available for the system has been stable test 1) Limitations 3) Remains constant for the life of the specific a) Difficult to correctly determine accuracy from lot number of QC material; it is the true internal sources alone value after all b) The stable system you thought you had may actually have inherent problems and constant bias 4) The value used in comparison with the Even though the data has been shown to be information gained after a change has consistent from previous results over an extended occurred to monitor the analytical system time period, you may actually be consistently wrong. Method validation or verification studies were never performed to evaluate accuracy 68

  • total error
  • TE
  • bias
  • imprecision
  • largest variation
  • true value
  • target value
  • 3 key numbers
  • percent
  • SLIPTA
  • SLIPTA checklist
  • mean
  • SD
  • standard deviation
  • total variation
  • eliminate bias
  • reduce imprecision
  • absolute bias
  • cumulative mean