Best way to perform duplicate or triplicate analysis for every registered sample?

Technical Questions and Answers
1

Hi,

For every sample that I register, I need to perform duplicate or triplicate analysis (for example, SiO2 for precision check).

I am not sure if partitions are the correct approach, because partitions seem more suitable for splitting samples across different units or physical sub-samples.

In my case, I simply want to:

  • Perform the same analysis multiple times on the same sample

  • Record independent results for each replicate

  • Maintain proper audit trail for ISO/IEC 17025 compliance

Questions:

  1. What is the recommended way in SENAITE to handle routine duplicate or triplicate analyses?

  2. Should this be configured in Analysis Profiles?

  3. Is there a way to automatically create multiple instances of the same Analysis Service per sample?

  4. Is partitioning considered best practice for precision checks, or is there a better approach?

Thank you.

2

Hi @khatib-JMG

does this meet your use case: Secondary Testing of Samples ?

3

No. That is not exactly the same as my use case.

In our laboratory we sometimes perform duplicate or triplicate measurements on the same sample within the same analytical batch. This is mainly for precision and repeatability checks, not for re-analysis at a later time.

For example, in a LOI analysis, the same sample may be analysed three times:

LOI-1
LOI-2
LOI-3

Then we calculate the average value and report that as the final result.

So the workflow is:

One registered sample
→ multiple replicate measurements (duplicate or triplicate)
→ average result calculated

The Secondary Testing of Samples described in the documentation seems to address re-analysis of a stored sample at a later time, which creates a new sample record with a suffix (e.g. -S01, -S02).

What I am looking for is a way to handle replicate analyses within the same worksheet or analysis request, rather than creating secondary samples.

Could you advise on the recommended approach in SENAITE for handling duplicate or triplicate measurements and averaging the results?

4

This would be a great improvement! In our laboratory, we also make replicates of analysis, besides other QC’s; nonetheless, replicates only work for measuring and controlling the absolute deviation from the first obtained result, which works as a control, but it could be so much more dynamic.

This model for several replicates would need to:

  1. Calculate an Average result of all replicates, but also allowing to rule out values outside of tendency
  2. Calculate the RPD when comparing only two replicates; or
  3. Calculate the RSD when comparing more than two replicates

I believe this to be somewhat complex for the dynamic values have to be stored in objects and then recalculated according to the different models.