2.1 Quality management Flashcards
What are the characteristics of a calibrator?
Calibration is important for reliable quantitation of an analyte
- metrological traceability ; related to CRM certified reference material, national/int standard through unbroken chain of comparisons
- commercially supplied reputable source (lot #, expiry, conc, traceable)
- matrix similar to testing sample
- stable
What is the role of the calibration curve and what are important characteristics of the curve?
CC relationship between instrument response and known conc of analyte in calibrators, to provide a result for a sample within an acceptable range
- min 6 calibrators- blank with only matrix
- independent calibrators not dilutions of a master
- range 0-150% of conce likely to be encountered
- evenly spaced over range
- run in duplicates
How are in house controls formed?
Possible sources:
-pooled samples from multiple patients
-a large specimen from one patient
(for some antibody PCS) -could have plasmapheresis samples
Once obtained -ideal to have same matrix as patient -near the clinical significant cut off level/within clinically significant measuring range -stable -sufficient volume different levels of control
Establish acceptable range for PCS
- 15-20 repeats on different runs/operators
- establish sd/cv for PCS and range
- may chart on LJ
- verify result from another lab/method
What are desirable controls/ characteristics of a good control?
Control is a specimen with known quality/quantity which allows performance assessment /assess test validity
- around the cut off
- covers analytical concentration
- clinical relevant level
- similar matrix to patient sample
- good stability/easy storage/adequate quantity
- no need for dilution
- cost effectiveness
- minimal lot to lot variation
- native rather than recombinant
- safe (virus tested)
- should be independent from manufacturer (commerical or pooled in house control)
What is a drift pattern and what can be the causes of this?
Drift- at least 5 consecutive values on one side of the mean
Instrument or control issue
Instrument: accum debris, incubation chamger temp, light filter, deterioration calibration, light source
Reagent: expiry/degradation
Action: recalibration of the instrument or change QC lot
What is a shift pattern and what can be the cause of this?
A sudden abrupt change
Reagent- new calibration, new curve, lot/change in formulation
Control- reconstitution, different lot number, diff bottle, storage
change in protocol
lab condition- temperature, water, contam
Failure in sampling system, mixing , centrifuge, washer, technician
Instrument failure- major maintenance
What is bias and what types are there?
Bias- systematic error- disagreement between the measured value of an analyte and its true value
Constant- same across measuring range
Proportional- higher error at higher concentrations, lower at lower concn (starts at 0)
Error stated as % bias
On linearity graph*
(low b+mid b+ high b)/3
What is imprecision? and what can make precision worsen?
Random error High CV (bad precision)
Can occur with
(reagent, machine, operator)
- poor technique/malfunctioning equipment
- reagent; air bubbles, sampling errors, improper mixing
- incorrect pipetting; imprecise, ill fitting tip, clogging
- power fluctuation
- contaminated water
You are out on EQA - approach
First state assay, method, whether we are in/or out, random or systematic
- unacceptable /acceptable
- Investigate that EQA
- Bring out run
- pre-analytical; correct sample, handling, storage
- analytical; calibrator change, reagent lot change, technique change, staff change/operator following SOP, instrument parameters, maintenance
- post analytical; transcription error, - Look at method group
- methodology issue
- if out within method? reagent change - pattern previous cycles
- alone or with method
- ? method change
What is the hierarchy of preferred methods for establishing performance goals?
(stockholm hierarchy)
- goals based on clinical outcome
- goals based on clinical decision making
- clinician survey
- biological inter and intra individual variability - goals based on expert opinion
- goals based on peer capability eg EQA
- goals based on state of the art
What are the causes of low lymphosum?
Accepted >95%
- Malignant/aberrant population
- Interfering substance eg. inadequate cell lysis, debris, nucleated red cell
- Reagent deterioration and binding issues
What is quality?
What are some of the markers of quality in a laboratory?
How is quality applied in the laboratory?
Quality
Objectives of quality
- support high quality health care; reduce mort/morb/economic loss
- credibility of lab; consistency (accuracy/precision), right result first time, every time
- generate confidence in lab results
Man driven:on site audit; internal/external, accreditation
Material driven; internal, external, EQA
What are some of the quality controls used in flow?
Pre-analysis
- CS& T beads; polystyrene beads 3 diff intensities
- check fluidic system, laser(s), optics, and electronics; LJ chart PMT voltage
- rainbow beads: fluoresce in all channels 8 colours 3 lasers, check application settings; median MFI for each fluorochrome LJ chart; check voltage settings to optimise visual
- CS& T beads fail then
- check correct beads run expiry/storage
- -beads warm otherwise clump
- -rinse fluidics bleach/clean
- de gas flow cell air bubbles
- laser warm up
- optics/detectors instrument maintenance
- corect values entered
– engineer
Other controls;
- lymphocyte subsets CD chex LJ chart
- healthy patient control
- other post analytical lymphosum, T cell sum, delta check
- EQAP
Pitfalls of flow ?
preanalytics - correct patient sample, cell viability, handling, storage,
analytical
- instrument set up/tracking (fluidics, lasers, optics, electronics),
- QC specific to run
- processing; lysis, antibody (expiry, non sp binding poly vs mono), incubation, washing inadequate
- interfering antibodies
- instrumentation maintenance, contamination
- operator within SOP
Gating/compensation
Voltages
Raw plots
post analytics
- QC lymphosum
- interpretation
- transcription
- rounding issue
What are critical results? What is the procedure in your lab
Critical- results that will affect or change a patient’s management and should be communicated urgently
e.g vasculitis, GBM, dsDNA/ENA SSA preg, neuronal, tryptase, HIV, DHR,
NATA requirement- documented procedure
Procedure
- scietnsit tells reg/smo of result
- Result checked
- Interim
- phoned
- FU w verify/document/
- document
Changing lot number of QC material, how would you go about changing over to the new lot number batch?
- Run in parallel old and new QC
- Run 20 values sequential runs or/and within runs
- Compare mean and SD value with previous QC to see if acceptable performance limits for new QC material
- Avoid changing lots of reagents whilst change in new QC
- Document in multi QC program
- Periodic monthly (ACT) QC checks to refine performance characteristics
What are some of the ways to improve TAT?
- Test selection/order entry- easy access/standardisation
- Specimen collection/delivery- access to info handling/phlebotomy notice/barcodes/delivery systems
- Testing- automation, minimal downtime, automated dilutions, verifications, incomplete lists
- Reporting- interface, preliminary results
Evaluate flow to maximise efficiency
What are the causes of a gradual drift/trend?
Trend of at least 5 consecutive values- gradual change
Reagents/controls- incorrect storage/unstable/refridgerator problem/degradation
Analyser issues
Lab temperature
Fix: recalibrate instrument or use fresh control
What can cause between assay imprecision?
- reagent lot fault
- contro problem
- variation in coated beads/well/tube
- other analytics; pipette/washing/centrifugin
What can cause within plate/ poor reproducibility
- excessive time taken to add samples/controls
- multichannel pipette
- inconsistent washing
- poor distribution of ab in sample(mixed if thawed)
- inconsistent incubation
- reagent lots diff
- inconsistent washing
What can cause high background signal?
- poor quality water for washing plate
- substrate solution deteriorated
- insufficient washing
- cross reac/non specific binding
- conjugate concentrate too high
- washer microbial contam
- reader malfunctioning not blanked properly - repeat zero reading
What can cause low optical density?
- lab temp too low
- too many wash cycles
- incubation period too short
- plate/reagents too cold
- if no colour then conjugate not added/wrong bottle used
- assay plate read at wrong wavelength/reader malfunctioning
What is the difference between a calibrator and control?
Calibrator- point of reference accurate known value to calibrate a method and standardise results, used to establish relationship between amount of signal produced and the analyte concentration
curve, single or multipoint, usually from kit
Control- determines validity of run and monitor performance of assay by assessing precision
known value, can be 1st /3rd/ PCS, pos/neg charted on LJ chart
How do you establish value of a new QC material?
Run at least 20 times (30) over different batches and runs
Calculate mean, SD and CV of data
mean +/-2 SD of the mean
Needs satisfactory CV
Plot on LJ chart
