Page 24 - qp_17

APPENDIX 5

METHOD COMPARISON (CSLI EP-9A2)

1. Prepare sample collection plan
a. Collect minimum 40 patients samples, covering the entire reportable range of
both method (current & test method).
b. Use samples drawn on the day of analysis whenever possible
c. Make sure every sample is traceable if further investigation needed.

2. Analysis of Sample
a. Ensure the system is stable and within control*Run IQC on both analyzers
b. The analysis should be done on both test and current method within 2 hours

3. Analyse 8 samples per day
a. Perform first run according to sequence, then run
b. 2 time in reverse order
nd
• Run 1- 1.2.3.4.5.6.7.8
• Run 2- 8.7.6.5.4.3.2.1

4. Record and examine all the results in excel worksheet daily.
a. Calculate:
• Mean run1 & run2 (both method)
• Difference between run1 & run2 (both method)

b. Graph the Difference plot
• Record the all the result daily
• Check any outlier-check for reason of outlier
• Remove outlier- Analysed additional sample to replace the outlier
• Too many outlier? Stop the experiment, investigate and contact
manufacturer if required

5. Test for within method outlier
a. Calculate the mean absolute different between replicates for each samples .
• Absolute difference, Run1-Run2
• Mean absolute difference=∑ (absolute difference)/n

b. If any of the individual absolute difference exceed 4 times mean absolute
difference, do additional calculation for each method using normalized
(relative) absolute difference

c. Calculate mean normalized (relative) absolute difference
• Normalized (relative) absolute difference, (Run1-Run2)/ Mean of
replicates
• Mean normalized (relative) absolute difference= ∑ (relative absolute
difference/n)