Quality Assurance Basics

Quality Assurance

• Quality assurance aims to get the correct answer.
• Answers should be accurate and precise enough to support later decision-making.
• Data quality standards involve obtaining the right data, getting the data right, and keeping the data right.
• The goal is to ensure results meet customer needs, so keep in mind:
  • Raw data: measurements.
  • Treated data: concentrations from raw data using calibration methods.
  • Results: quantities reported after statistical analysis of treated data.
  • Specifications: standards state how good the numbers need to be and what precautions analytical procedures need to have.
• False positive: concentration exceeds the legal limit when it is actually below the limit.
• False negative: a concentration is below the legal limit when it is actually above the limit.

Choosing a Method

• Selectivity is the ability to distinguish an analyte from other species in a sample.
• Sensitivity is the ability to respond reliably and measurably to changes in analyte concentration.
• Blanks are samples not intended to contain the analyte, types include:
  • Method blank: all components except analyte taken throughout all steps of the procedure are included here.
  • Reagent blank: accounts for trace amounts of analyte found in reagents.
  • Field blank: indicates if the analyte is picked up by exposure to field conditions.

Spike Recovery

• Response to an analyte can be decreased or increased, the following should be considered:
  • Matrix: everything in the unknown, other than the analyte.
  • Spike: a known quantity of analyte added to a sample.
• Spike recovery percentage can be given as:
  • % recovery = (C spiked sample - C unspiked sample) / C added * 100).

Dealing With Large Numbers of Samples

• Periodic calibration checks are required.
• Method drift can be due to changes such as room temperature, spoilage of reagents/standards.
• Calibrations checks should be different from the ones used to prepare the original calibration curve.
• Performance test samples help eliminate bias introduced by the analyst who knows the concentration of calibration check samples.
• Standard operating procedures outline the steps taken for a procedure and serve as control experiments to detect problems in the lab.

Assessment

• Assessment involves collecting data to show analytical procedures are operating within specific limits.
• Assessment is the process of verifying that final results meet use objectives.
• Control charts are used to monitor performance and standard protocols outline what must be documented.

Method Validation

• The process of proving an analytical method is acceptable for its intended purpose.

• Demonstrated requirements met for method selectivity, linearity, accuracy, precision, range, limit of detection & quantitation, and robustness.

• Selectivity is the extent to which an analytical method can distinguish the analyte from everything else in the sample.

• Linearity measures how well a calibration curve follows a straight line.

• The response is proportional to the quantity of analyte.

• Residual plots emphasize the difference between the calibration data and the least-squares line.

• Accuracy can be achieved by analysing certified reference, comparing results, and analysing spiked blank samples/standard additions of analyte.

Detection Limit (DL)

• The detection limit (LOD) is taken as 3x the standard deviation of the blank, expressed as LOD = 3s/m.
• The lower limit of quantitation (LOQ) is 10x the standard deviation of the blank, expressed as LOQ = 10s/m
• Reporting limit: the concentration below which regulations say an analyte is reported as not detected, even when it is observed.
• Robustness is the ability of an analytical method to remain unaffected by small changes in operating parameters.
• Range is the concentrations where linearity, accuracy, and precision are acceptable.

Standard Addition

• In standard addition, known quantities of analyte are added to the unknown.
• A increase in signal indicates how much analyte was in original unknown.
• The method requires a linear response to analyte concentration.
• There is higher precision when standards are added by mass instead of volume.

Internal standards

• Standard addition: known amount of a compound - same substance as analyte added to the unknown.
• External standards: solutions with known concentrations of analyte used to prepare a calibration curve.
• Internal standards: known amount of compound, different from the analyte, and is added to the unknown.
  • Especially useful for analysis when the quantity of sample analysed or instrument response varies from run-to-run.
  • Desirable when sample loss occurs during sample preparation steps prior to analysis.
  • Helpful for internal standards by making them chemically similar to the analyte so the effects offset each other.
  • Can be expressed as:

One-Point Calibration (Internal Standard)

• Prepare a known mixture of a standard and analyte.
• Measure the relative response of the detector to the 2 species.
• Detectors generally have a different response to each component.
• The response factor (F) is given as: Signal from analyte/Concentration of analyte = F * signal from standard/concentration of standard*.

Multipoint Calibration Curve (Internal Standard)

• Multipoint calibration curve preferred as it averages out experimental variability and verifies the linearity of response.
• To do this, construct a graph of the signal ratio vs concentration ratio.
• Graph should be linear with zero intercept.
• The slope of the graph is the response factor (F).