MOOC: Validation of liquid chromatography mass spectrometry (LC-MS) methods (analytical chemistry) course

Self test 5.2A

Which are important issues in determining bias?

Homogeneous sample

Sufficient number of replicates

Reliable reference value

Concentration range match

Matrix match

Sufficiently long timeframe

Determination of bias can be hindered by large random effects

Does bias depend on the time frame?

Yes, bias determined within a single day is different from one determined on different days (and averaged)

Which effects are systematic within day and which are systematic in long term?

Effect	Systematic within day	Systematic in long term
Calibration graph of a specific day
Injection volume of autosampler is 5% higher than nominal
Delicate analyte partially decomposes at room temperature before samples are loaded into cooled autosampler
Repeatability of peak integration
Ionization suppression in the ESI source by a co-eluting compound
Baseline noise

Choose the correct answer from the drop-down menu for each case: short-term bias|not a bias or long-term bias.

All the results of a specific day are systematically influenced by the calibration graph of that day
Delicate analyte partially decomposes during sample preparation leading to lowered results
The titrant concentration determined on a particular day is slightly lower or higher than the true concentration
Because of the specifics of the used sample preparation procedure the sample is digested incompletely|leading to lowered values
During titrations we get every time slightly different result as previously

What do the parts of the figure marked with numbers mean?

Bias and related concepts

1	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect
2	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect
3	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect
4	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect
5	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect
6	Mean measured value Unselect Uncertainty range of Ci (Ci – U … Ci + U) Unselect Reference value Unselect Repeatability/Reproducibility Unselect Individual measured value Unselect Bias Unselect

What is bias?

Difference between the mean of a large number of replicate measured values and the true value

Difference between the measured value and the true value

An estimate of a systematic measurement error

Difference between the measured value and a reference value

Difference between the mean of a large number of replicate measured values and a reference value

Two analysts determined meropenem (an antibiotic) in blood plasma. Both needed to determine the recovery of the procedure. Which analyst did it more correctly?

Analyst 2 took 500 μL of the blank plasma, added 100 μL of meropenem standard solution and mixed well. She then added 500 μL of methanol for protein precipitation. She separated the precipitated proteins by centrifugation and injected the resulting supernatant into the HPLC system for analysis.

Analyst 1 took 500 μL of the blank plasma and added 400 μL of methanol for protein precipitation. He separated the precipitated proteins by centrifugation and transferred the supernatant into an HPLC vial. 100 μL of meropenem standard solution with suitable concentration was added to the supernatant and the resulting solution was injected into the HPLC system for analysis.

When we should use different bias types?

Way of expressing	Formula	When to use?
Absolute bias	bias=C_{lab_mean}-C_ref
Relative bias	bias=(C_{lab_mean}-C_ref)/C_ref
Recovery	R=C_uncorrected/C_ref
Recovery	R=(C₁-C₀)/ΔC

Course introduction
1. Validation: General
2. Selectivity and identity confirmation
3. Linearity of signal, linear range, sensitivity
4. Precision
5. Trueness
6. Precision and trueness: some additional aspects
7. Accuracy
- Self test 7
8. Stability
9. LoD and LoQ
10. Ruggedness, robustness
Acknowledgements
Frequently asked questions
References
Glossary
What our participants say?