Lab Study Flashcards
What must be done prior to running and reporting patient results?
Quality control results must be assessed.
If not done, results are considered invalid and patient results are not reported.
What is accuracy?
Accuracy refers to the closeness to the true value. Think hitting the ‘bull’s eye’.
What type of error can be detected using the mean?
Systematic errors that affect accuracy.
What is precision?
Precision refers to the agreement of test results to be reproducible. A deviation from precision can detect a random error.
What can you use to detect imprecision?
Standard deviation and coefficient of variation.
What are possible sources of systematic errors in the chemistry lab?
- Changes in reagent lot.
- Calibration issue (change in calibration or incorrect calibrators used).
- Deteriorating reagents.
- Deteriorating instrumentation (light source, pipettor alignment, etc.)
What are some sources of random errors in the lab?
- Bubbles (reagent, reagent lines, QC, etc.)
- Instrument instability (reagent dispensing error, light fluctuations, etc.)
- Temperature variation
- Human error (pipetting, improper mixing, etc.)
What are some corrective actions or things you can do to check for systematic errors?
- Check expiration dates (reagents/calibrators). Replace as req’d.
- Preventative maintenance. Investigate instrumentation. Calls service.
- Re-calibrate the assay.
- Ensure you are using the correct lot numbers of calibrators/reagents.
- Use correct charts w/ lot numbers.
- If no charts, may need to do a crossover study.
What is a crossover study / how is it done?
Crossover study is performed before the current lot of quality control (QC) material expires, or before the lab runs out of QC materials. The new lot is run in parallel with the current lot in order to establish new control limits.
What are some corrective actions / things to check for random errors?
- Check reagents and QC material for bubbles or that it has been properly mixed then re-run QC.
- Perform regular preventative maintenance on all instrumentation to avoid errors. Check if bubbles in lines or issues with pipettor dispensing.
- Follow SOPs.
- Ensure competency prior to performing technique.
What are quality control material in a chemistry lab?
QC material are stabilized specimens with a known range of results. They are made with the same matrix as the samples and are treated and tested like patient samples. Based on results obtained from QC material, a run may be accepted or rejected.
What is the control range? What percentage of all data points should fall within it?
Control Range is the range of acceptability, it is +/- 2SD (two std. dev.)
95% of all data points should fall within this range.
What are reference ranges (also called reference intervals)? How are these ranges determined?
Reference ranges are range of results that would be considered normal. \
Samples from a minimum of 120 “healthy/normal” patients that span a variety of ages, sexes, and ethnicities. (Child/adult though are different). Must be representative of the population the lab serves.
What is ‘inclusion’ criteria?
Inclusion criteria says that the people in the sample for the reference range should represent the population the lab serves. It identifies why these individuals should be in the study. These individuals have to complete a confidential questionnaire and sign a consent form.
What is exclusion criteira?
Exclusion criteria identifies those individuals that are considered outside the “healthy” population for that particular analyte due to a disease or disorder. They are not in the study for reference range.
What else is considered in determining reference ranges?
Pre-analytical variables are considered, i.e. if fasting is required for fasting glucose, hemolysis, lipemia, etc.
What is transference?
Transference is the term where labs use a reference range from made somewhere else with a similar instrument, similar lab, manufacturer/vendor and then validate using a method of comparison study or validate with a smaller population size such as 40.
What limits if transference can be done or not by a lab to determine their reference ranges?
It can only be done if an acceptable reference range already exists for the instrument or lab method to be employed.
If transference is being used, what is the acceptance criteria? What happens if it is not met?
If using transference, no more than 10% of the values obtained can be outside of the manufacturer’s reference range. E.g. if 40 samples, no more than 4 can be outside manufacturer’s range.
If not met, than a larger population needs to be sampled, 120 people, and the reference range re -calculated. If not met again, a still larger population sample is required.
If met, then transference can occur.
What is the difference between reference range and reportable range?
Reportable range represents the ability of an instrument to provided reliable analysis within a certain range. Higher or lower values than the range report as upper limit as instrument is unable to determine true value.
Reportable ranges are much broader than reference ranges. Typically the results below and above those limits are beyond healthy and are so extreme/life threatening and so are not necessary to actually determine.
The reference range helps determine what is normal.
How would you determine the reference range for a normal gaussian distribution of data? What if the data did not follow the normal gaussian distribution?
Reference range for a normal gaussian distribution of data would be = Mean +/- 2SD.
If the data was skewed this does not apply or you determining reference range for therapeutic drugs then the reference range is based on toxicity values or minimum effective concentrations.
What is method validation and name some procedures used?
Method validation is the process of determining if an analytical procedure is consistently reliable in producing quality lab results. Includes determining accuracy, precision, medical decision limits, diagnostic sensitivity, diagnostic specificity, and reportable range.
What are medical decision limits?
Medical decision limits are values for analytes that represent specific diseases.
Qualitative medical decision limits - presence if an analyte if there or not correlates to disease state or absence of disease state.
Quantitative medical decision limits - are cutoff concentrations that if above or below correlate to a specific disease state.
Good medical decision limits will have low ________ and low ______ results.
False-positive
false-negative
What is the reportable range (or analytical range, AMR, or dynamic range)?
Reportable range is the range of specimen concentration (lowest to highest) which the method may be used without dilution.
How do you determine reportable range?
- Use elevated patient samples, spiked patient pool, or commercially available linearity materials. Must have same sample matrix as patient samples.
- Serially dilute to at least 5 to 7x’s to equally cover reportable range.
- Do testing in duplicate.
- Run from lowest to highest to avoid any potential carryover.
- Linear regression analysis is performed and compared to the manufacturer data (y = mx + b).
In the linear regression analysis for checking reportable range, what are the potential results?
- No error, lines overlap or have similar slope and y-intercept.
- Constant error, the slope is the same but the y-intercept is different (b is off).
- Proportional error, change in slope of line (difference in m).
- Combination of both of the above.
Between the manufacturer’s line and your analysis for linear regression what statistical value can be used to evaluate how well the two lines match?
Calculate Correlation of determination, R^2, value from 0 to 1. In chemistry, we aim for R^2 > 0.98.