Opening the book…
Quality assays provide evidence about a sample’s identity, concentration, purity, biological activity, or other defined characteristic. Quantitative assays report numerical values, while qualitative assays report categories under defined conditions. The test must match the sample, intended use, and decision. Concentration and potency are different: measuring the amount of protein does not necessarily show how much of it is biologically active.
A reliable result begins with suitable sampling. A representative sample reflects the characteristics relevant to the question being asked. Sampling plans should address important differences across times, locations, or containers rather than selecting only convenient material. Sample preparation also matters. Mixing, extraction, storage, and dilution can introduce losses, interference, or errors. Recovery studies can help evaluate preparation and measurement effects, but their conclusions depend on how the study is designed.
Reference materials, calibration standards, and controls serve related but different purposes. Calibration standards establish the relationship between an assigned quantity and the measured response. Controls check whether the test system performs acceptably. Positive, negative, extraction, and internal controls address different possible problems. Staff must evaluate control failures before interpreting affected samples. Passing controls do not prove that sampling, labeling, or every other step was correct.
Analytical method validation provides evidence that a procedure is suitable for its intended purpose. Relevant characteristics include accuracy, precision, specificity or selectivity, range, and detection or quantitation capability. Accuracy concerns closeness to a relevant true or accepted value. Precision concerns agreement among repeated measurements. Tightly clustered results can still be biased, and a mean near the reference does not make widely scattered results reliable. Standard deviation and coefficient of variation describe spread, but neither alone establishes overall method suitability.
Detection and quantitation limits define different capabilities. A method may detect an analyte at a level too low for reliable numerical measurement. A result below the detection limit does not prove absence. Robustness studies examine performance after small, deliberate method changes, while system suitability checks whether the measurement system can perform adequately for a run or sequence. These checks support reliable testing but do not replace validation.
Clear tables and graphs help readers evaluate results. A standard curve relates assigned concentrations to measured responses. For a suitable linear model, y = mx + b, an unknown concentration can be estimated using x = (y − b) / m. Use the model only within its supported range and follow the method’s preparation and reporting rules. A high R² alone does not establish an acceptable calibration; residuals, standards, independent controls, and relevant acceptance criteria also matter.
The calculated concentration usually describes the solution actually measured. When a valid dilution was performed, multiply that concentration by the dilution factor to estimate the original sample concentration. The dilution factor is final volume divided by sample volume. Preserve units, report appropriate precision, and avoid extending a curve beyond its supported range. Reliable assay decisions depend on the complete chain of evidence, from sampling through calculation and review.
Key Terms
|
Assay |
Analyte |
Quantitative assay |
|
Qualitative assay |
Identity test |
Purity |
|
Potency |
Stability |
Representative sample |
|
Sampling bias |
Sample matrix |
Recovery |
|
Reference material |
Calibration standard |
Positive control |
|
Negative control |
Internal control |
Analytical method validation |
|
Accuracy |
Precision |
Bias |
|
Random error |
Standard deviation |
Coefficient of variation (CV) |
|
Limit of detection (LOD) |
Limit of quantitation (LOQ) |
Specificity or selectivity |
|
Robustness |
System suitability |
Reportable range |
|
Standard curve |
Linear regression |
Residual |
|
Interpolation |
Extrapolation |
Dilution factor |
Review Questions
1. How does a quantitative assay differ from a qualitative assay?
2. How do identity, concentration, purity, and potency tests answer different questions about a product?
3. What makes a sample representative, and why might convenience sampling be unsuitable?
4. How can sample preparation affect an assay result, and what can a recovery study help evaluate?
5. How does a calibration standard differ from a control?
6. What do positive and negative controls help establish, and what should happen when a required control fails?
7. How do accuracy, precision, and bias differ?
8. What do standard deviation and coefficient of variation describe?
9. How does the LOD differ from the LOQ, and why does a result below the LOD not prove absence?
10. What are the purposes of robustness studies and system suitability tests?
11. Why is a high R² insufficient by itself to establish an acceptable standard curve?
12. How do you calculate an unknown concentration from y = mx + b and then account for a valid dilution?
Apply Your Knowledge
The following situations and numerical values are hypothetical teaching examples, not approved assay methods or product specifications.
1. A facility fills vials over five days. An analyst selects six easily reached vials from the first hour and combines equal portions into one sample for testing. Explain how this plan could miss changes across the production run or differences among vials. Propose a more suitable sampling approach and explain when separate samples would provide information that a composite sample could hide.
2. An assay plate has a smooth standard curve and a high R², but an independently prepared low-level control fails its predefined acceptance criterion. A coworker wants to report the sample results because the curve looks good. Explain why the curve alone is insufficient. Identify possible standard preparation, control preparation, matrix, and method problems to investigate before deciding which results can be interpreted.
3. A reference material has an assigned concentration of 100 units/mL. Method A gives results of 89, 90, and 91 units/mL. Method B gives 90, 100, and 110 units/mL. Calculate each mean, sample standard deviation, CV, and relative bias. Compare the precision and bias of the two sets. Explain why these three measurements per method are insufficient to establish complete method performance.
4. A method has an established LOD of 0.20 µg/mL and LOQ of 0.60 µg/mL. A sample produces a confirmed response consistent with analyte detection, but its estimated concentration is 0.35 µg/mL. Explain why reporting a precise concentration or reporting the analyte as absent would each overstate the evidence. Describe how the approved reporting rules and the purpose of testing should guide the next step.
5. A teaching assay uses the linear model y = 0.0050x + 0.020, where y is absorbance and x is concentration in µg/mL. Its supported working interval is 10–100 µg/mL. An analyst dilutes 1.00 mL of sample to a final volume of 5.00 mL and measures an absorbance of 0.320. Assuming all method criteria pass, calculate the measured solution’s concentration, the dilution factor, and the original sample concentration. Then explain why an absorbance of 0.720 would require a different response rather than simply extending the line.
This page titled 15.18: Chapter Summary was last modified on Sat, 31 Oct 2026 23:49:31 GMT and is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Emalee MacKenzie.
The book could not be loaded just now.