Reagent Verification

1
Reagent Verification

• Compare 2 controls and at least 6 patient samples
  covering the measurable range
• Perform testing with old and new lot numbers of
  reagent on the same freeze/thaw cycle
• Use spreadsheet for making calculations
• Review comparisons from several previous lots
• Beware of bias creep (drifting in one direction
  over time)
• Look at calibration curve responses (optical
  density or clotting times versus concentration)

Lab Pointer
Contributed by Melissa Bethel,
MT(ASCP) Esoterix Coagulation Aurora, Colorado
2
Example Antithrombin Activity

• Bias is calculated as the value for the
  Difference divided by the value for the Old
  Lot x 100
• Suggested acceptable limits for the correlation
  coefficient (R2), percent bias, and slope are
  noted below however, acceptable criteria should
  be determined by each laboratory

3
Reagent Lot to Lot Comparison

• Reference
• Westgard JO. Basic Method Validation, 2nd ED,
  2003. (ISBN 1-886958-19-X)

In a Correlation Plot, values obtained from an
Old Lot number of reagent are plotted against
those obtained using a New Lot number. (Data
taken from previous slide)
The difference between lot numbers of reagent is
plotted against the actual test result. If
differences are dispersed on both sides of the
line of zero differences, results are considered
acceptable.