Topic 1: Scope of Clinical Biochemistry; Quality Management

What are the reasons for requesting a test?

• establish/confirm a diagnosis

• reassure a patient

• avoid litigation

• monitor disease state

• monitor therapy progress

• population screening

What type of clinical specimens can be taken?

Bodily fluids/excreta

• saliva

• urine

• blood

• CSF

• faeces

tissue samples

• biopsy

When do most diagnostic errors occur?

80% in pre-analytical phase

What are some ways to reduce sample variability?

• patient ID checks and quality management

• use standards to monitor performance

• have appropriate protocols in place to handle labelling and data

• accreditation ensures accuracy across labs

What are some errors in blood sampling?

1. technique: may lead to haemolysis

2. stasis during venipucture: false increase of plasma content level

3. insufficient speciment taken

4. errors in timing: fasting vs unfasting

5. inappropriate container

6. inappropriate sampling site

7. incorrect speciment storage or transport

8. patient related factors: smoking, medications, stress

Describe the types of testing errors? Random vs Systematic errors. Which type reduces precision and which reduces accuracy?

random errors

• unknown and unpredictable source

• normal distribution

• reduce precision: measure of how closely measurements cluster together

systematic errors

• opperator error or from measuring instruments

• can show trend: alteration from standard w pattern of divergence

• can show shift: abrupt change in measurements

• reduce accuracy: measure of how closely measurements are to the true value of quantity being assayed

• constant error: specific diff between test and reference methods

• proportional error: variable diff between test and reference

Explain the interpretation of a normal distribution. This distribution is what makes the ______ of a diagnostic marker?

95% of normal population have values within ± 2SD of mean

5% have values outside this range

• may be abnormal or normal

reference range

How would you effectively distinguish between a healthy and diseased population on a distribution curve

determine the diagnostic cut off value: the uncertainty range

• draw a line where the two populations intersect

• values in this range can be FP or FN

What do the terms: Imprecision, Within assay imprecision, Day-to-day imprecision, and coefficient of variation (CoV) mean?

Imprecision = random changes that reduce agreement between replicate measurements

Within-assay Imprecision = diff between results for same speciment assayed repeatedly at the same time

Day-to-day imprecision = diff between results for same speciment assayed repeatedly at different times

Coeffecient of variation (CV) = expresses precision and should be close to zero as possible

• CV% = (SD of assay) x 100% / mean of assay results

How do you determine the total number of negatives and positives /diseased and not diseased in a population

	Number (diseased)	Number (not diseased)	Total Number
Positive Result	TP	FP	Total (positive)
Negative Result	FN	TN	Total (negative)
	= total diseased	= total not diseased	Total

What is sensitivity a measure of? How would you calculate it?

a measure of positive results

• measure of freq of a positive test result when a particular disease is present

• represents the % of TP results out of TP and FN results

sensitivity% = TP / TP + FN x 100%

What is specificity a measure of? How would you calculate it?

a measure of negative results

• measure of freq of negative test results when a particular disease is absent

• represents the % of TN results out of TN and FP results

specificity% = TN / TN + FP x 100%

What is the receiver operating characteristic? What information does it give?

• usefulness of tests expressed visually as ROC curves

• the greater the area under = the more useful the test

Briefly describe spectrophotometry? What does it measure?

• measurement of light by a specific molecule

• measures concentration or purity of a molecule

Briefly describe atomic absorption spectrometry? What does it measure?

• measures the absorption of light by atoms in a gaseous state

  • flame dissociates metals from chemical bonds

• measures metal ions via flame

Briefly describe nephelometry? What does it measure?

• measures the precipitation/scattering of light from an antibody:antigen complex in solution

• measures proteins

Briefly describe ion selective electrodes? What does it measure?

• measures the potential difference across a selective membrane - electrode only allows specific thing across

• measures electrolytes and gases

Briefly describe electrophoresis? What does it measure?

• molecules move in an electric field based on charge, size and shape

• migration depends on:

  • electric charge of molecule

  • size and shape of molecule

  • electric field strength

  • properties of support

  • temp

• used to separate proteins and DNA for ID and quantification

Briefly describe immunoassays? What does it measure?

• uses antibodies to bind specific molecules/target antigens which is then detected and measured (signal intensity)

• measures hormones, drugs, proteins etc

Briefly describe high performance liquid chromatography? What does it measure?

• separation technique that separates components in sample based on their affinity for a stationary phase and a mobile phase

• measures drugs, hormones, etc

What is the purpose of internal quality controls and what does it involve?

• needed to minimise errors and ensure test result accuracy

• rigorous, systematic checks on instrumentation, methodologies and processes used in daily operations

• lab equipment calibration

What is the purpose of quality control? but what does it not guarantee

• ensures each analytical run is consistent with results from the previous internal runs

• lab sets the QC and confirms consistency of lab results

• but does not guarantee the correct result (i.e. if the QC set was not accurate)

What are some types of analytical quality errors usually encounted?

• data input errors: standards, calibration values

• instrument errors: accuracy, precision

• observer errors: reading errors, errors in equipment selection, analysis and interpretation

• environmental errors: external factors

How is uncertainty in measurement defined? What is the definitions of these terms? Repeatability, Reproducibility, Within-lab reproducibility, Between-lab reproducibility, Stability, Bias, Drift, Resolution, Standard uncertainty?

Repeatability = variation among measurements made on same object using same conditions

Reproducibility = variation among measurements made on same object using diff conditions

Within-lab reproducibility = precision within single lab over time

Between-lab reproducibility = precision between diff labs

Stability = variation among measurements made on same object using same conditions at diff time

Bias = diff between measurements average for same object and its true value

Drift = continuous and incremental change over time

Resolution = smallest change in quantity that causes detectable change in output/indication

Standard uncertainty = uncertainty expressed as SD

What are some sources of uncertainty?

• weighing

• volume uncertainties of pipette

• fluctuations of density

• temp effects

• metrological traceability

• homogeneity of sampling material

• matrix of sampling material

• complex steps in preparation of sample

• stability of measuring signal

• calibration of device

What constitutes a suitable quality control?

• shelf-life/open vial stability

• liquid vs lyophilised

• medically relevant levels

• challenge range of instrumentation

• within-run/vial-to-vial variability

• matrix preservatives

QC material is often specially prepared plasma or serum and should be used in every batch analysis? T or F
If there are a large number of specimens, how often should a QC be used?

true

1 every 20 specimens

duplicate specimen testing assesses what? and what type of errors are detected vs what is not detected?

• assesses precision

• detects random errors not gradual drift

What are the limitations of IQC? and what is required to check IQC?

• only detects change in performance between current and stable/OG operation

  • if OG mean is systematically wrong → cannot be detected

  • original method must establish accuracy

• needs EQA to ensure systematic errors do not slowly increase and are undetected

What do anomalies such as Outlier, Shift and Drift mean?

Outlier: QC material contamination

Shift: new QC batch component

Drift: new calibration

What is the purpose of EQA and how is it organised? What is the objective?

• external comparison of labs performance against other labs involved in similar activities

• evaluation by outside agency of performance by a number of labs on material which is supplied specifically for the purpose

• organised on a nation or regional basis

• analysis is retrospective

• objective = achieve comparability, not necessarily accuracy

What are the 3 functions of EQA schemes and the outcomes?

primary = measurement

• QA material, data collection and analysis

secondary = assessment

• analytical standards: allowable limits

tertiary = corrective action

• protocols for dealing with poor performance

outcomes:

• compare lab results with peer group for:

  • accuracy, precision, linearity

• compare methods and instruments

Compare and contrast IQC vs EQA

IQC

• monitor and minimise errors and ensure result accuracy within a lab

• QC values set by the lab and confirms consistency of results

• but does not guarantee correct result

EQA

• ensures results from diff labs are comparable using same test, methodology and instrument

• external proficiency organisation provides samples

• independent external evaluation of labs performance

• checks IQC and compentency of staff

What are the major activities of QAP

1. preventative

   • activities performed before examining specimen that establish systems to promote accuracy in testing

   • equipment maintenance

   • instrument calibration

   • media testing

   • staff training

2. assessment

   • activities performed during testing to determine if systems working correctly

   • assaying standards + controls

   • maintaining control charts

3. correction

   • activities performed to correct system after error detected

   • equipment troubleshooting

   • recalibration of instruments

What are some agencies that address quality

• Royal College of Pathologists of Australia (RCPA)

• National Pathology Accreditation Advisory Council (NPAAC)

• National Association of Testing Authorities (NATA)

What are the functions and components of a Levey-Jennings control chart?

monitor and assess stability and correctness of lab processes over time

X axis = time or run

Y axis = measured value from QC sample

mean = horizontal lines

SD lines = control limits

What is the westgard rule 1_2s? Do we reject or accept the run?

• warning rule

• violation if value lies > ± 2SD

• inspection of QC data

  • investigate and repeat test

What is the westgard rule 1_3s? What kind of error does it detect? Do we reject or accept the run?

• detects random error but may be indicative of large systematic error

• violation if value lies > ± 3SD

• run rejection

What is the westgard rule 2_2s? What kind of error does it detect? Do we reject or accept the run?

• detects systematic error

• violation if two consecutive control values from separate runs lie between 2SD and 3SD from mean in same direction

• run rejection

What is the westgard rule R_4s? What kind of error does it detect? Do we reject or accept the specimen?

• detects random error

• violation if there is a 4SD difference between two control values within a single run

• specimen rejection

What is the westgard rule 4_1s? What kind of error does it detect? Do we reject or accept the run?

• indicates bias

• violation if there is 4 consecutive values for a single level of control greater than 1SD on the same side of mean

• run rejection

What is the westgard rule 10_x? What kind of error does it detect? Do we reject or accept the run?

• indicates bias

• violation if there are 10 consecutive values for same level of control on the same side of mean

• run rejection

What are some good problem solving habits?

• inspect control charts, ID rules broken and error type

• relate type of error to possible causes

• consider factors in common on multi-test systems

• relate causes to recent changes

• verify solution and document remedy