Unit 2 Flashcards
Reflectance Photometry
- vitros analyzers use this principle
- amount of light reflected is proportional to amount of analyte
Absorption Spectroscopy
- the darkness of a colored solution quantitatively relates to the Molar concentration of the chromosphere molecules in solution
- darkness of the solution partly determines how much light that solution absorbs
Principle of absorption spectroscopy
- the number of light-absorbing molecules in solution are proportional to the amount of color of that solution (absorbance)
- absorbance is proportional to the analyte concentration (within certain limits)
6 parts of a spectrophotometer
1- light source (exciter lamp) 2- monochromator 3- primary exit slit 4- cuvette 5- photomultiplier tube (light detector) 6- readout device
Electro-magnetic radiation (EMR)
- exists in wave forms and photons
- wave form has 2 parts
——1- magnetic field (z)
——2- electric field (y)
Electromagnetic Spectrum (nm from crest-crest)
- visible spectrum from 380-725nm
- the smaller the wavelength(nm), the greater the light energy
- we do a lot of assays at 340nm because NADH absorbs the light but NAD+ does not
Light spectrum and their colors
>/= 725nm= infrared, low energy, not visible 600-725nm= oranges & reds 580-600nm= yellows 500-580nm= greens 440-500nm= blues 380-440nm= violets <380nm= ultraviolets, not visible
Light source- types of exciter lamps
- Tungsten-halide
——most common, emits wavelengths from 335-750nm - Hydrogen, Deuterium, or Mercury Arc Lamps
——limited UV use, emits wavelengths from 180-380nm
Types of monochromators
- prisms (very inefficient)
- diffraction gratings (most efficient)
- interference filters (most common)
Prisms
- rarely used
- resolving power= B(dn/d¥)
Diffraction grating
Most efficient
Interference filter
- most commonly used
- not true monochromators because they don’t break up white light into a spectrum of colors
Spectrophotometer vs. colorimeter
- spectrophotometer uses a true monochromator
- colorimeter uses an interference filter
- many times the terms are used interchangeably
Wavelength accuracy
- now that we have a light source and the wavelength is set, is it correct?
- test and document with:
—-holmium oxide: narrow bandwidth
—-didymium: broad bandwidth
Bandpass
The range of light waves that pass through the primary slit
Spectral bandwidth
- the range of the bandpass that has enough energy to potentially interact with the test solution
Also known as the Full Width Half Maximum (FWHM)
Spectral bandwidth documentation (using Holmium Oxide filter)
- find the peak at 361 no, record the absorbance (maybe 0.300A)
- move the wavelength to the left until the absorbance is half of the starting absorbance (0.150A) and record the nm reading
- repeat on other side of curve
- pretend nm reading on left was 357nm and right was 365nm
- difference between them=8nm= spectral bandwidth
Cuvettes
- made of optically perfect quartz glass or sometimes a good grade of plastic
- round cuvettes: inferior, curvature will scatter light
- square cuvettes: superior to round, lights hits 90* angle
Detection devices
Photomultiplier tube
- light energy from the sample hits the cathode and bounces back and forth, loses electrons along the way until it releases an electron to the anode at the other end of the tube
Photometric accuracy
-NBS (NIST) transmittance standards
—colored solutions with known absorbances e.g. nickel sulfate, ammonium molybdate
-NBS (NIST) SRM-930 & 931 series of neutral gray filters
—each filter has a known absorbance
—no colors to fade, just gray screens
Photometric linearity (example)
- remove cuvettes and set readout to 0.000A
- place 0.400 A filter in place of cuvettes and document absorbance reading
- remove filter, readjust air to 0.400 A, replace filter, document absorbance
- continue to 1.600 A
Stray light
- light not passing through the cuvette, but hitting the photodetector
- always causes falsely low absorbance readings
- affects high absorbances more than low
- causes of stray light: defects in cuvette, dust in bowels of instrument, degrading diffraction grating
- documented using cut-off filters at 380 & 680nm
- 680 filter will not transmit light above 680nm
—remove cuvette, dial in 700nm, put filter where cuvette goes
—if any light hits the detector, it’s stray light
Spectrophotometer QC- daily/weekly
Daily
- document wavelength calibration accuracy using Holmium Oxide or Didymium (1st)
- document photometric accuracy using SRM-930,931 neutral gray filters with a known absorbance reading
Weekly - must also document 1- photometric linearity 2- stray light 3- spectral bandwidth
Readout devices
- LED/LCD meter
- analog meter
- strip chart recorder
- printer
Beer’s Law equation
A=abc
A- absorbance at a specific wavelength (in a spectrophotometer)
a- molar absorptivity constant
b- cuvette light path in cm
c- Moles/Liter of light-absorbing product
Beer’s Law- explained
- the number of light-absorbing molecules in a solution are proportional to the amount of color of that solution (absorbance)
- absorbance is proportional to the analyte concentration (within certain limits)
- if he molar absorptivity constant is known for a given analyte, then that analyte can be measured in an unknown specimen without using a standard
- there are many applications in industry for this
- we use standards in the clinical lab
Beer’s Law proportionality calculation
- assume that the method is linear (actually follows Beer’s Law)
(Abs unknown/Abs standard)x conc standard= concentration of unknown
- you just have to include the measurement of one standard with the controls and patients
- it works because Beer’s Law says:
(Abs unknown/Abs standard)= (Conc unknown/Conc standard)
- the ratio of the two absorbances is equal to the ratio of the two concentrations
- do not use calculation if it does not follow Beer’s Law
Analytical Measurement Range (AMR)
- AKA linear, reportable, or dynamic range
- perform linearity study and plot absorbance vs concentration
- straight line is the range of analyte concentration that follows Beer’s Law with at least 90% accuracy
- must be verified every 6 months if nothing has changed
Absorbance curves
- run at least 6 standards
- plot absorbance vs concentration of each standard
- look to see if method follows Beer’s Law (linear?)
- if not linear then, this graph can be used as a calibration curve
Absorbance vs %Transmittance
%T=% of Pi hitting the detector
Absorbance=the amount of Po NOT hitting the detector
Conversions
%T to Absorbance
Abs=2.000-Log(%T)
Conversions
Absorbance to %T
Log(%T)= 2.000- Abs
- 000 Abs = ? %T
1. 000 Abs = ? %T
100%T
10%T
Absorbance curves to choose best wavelength for assay
- our reaction mixture (reagent+specimen) containing product to be measured in a cuvette
- zero spectrophotometer on water
- scan wavelength from 400-700nm
- complimentary wavelength gives the highest absorbance curve(reading)
Zeroing the Spectrophotometer (blanks)
- blank: solution used to tare or zero the absorbance reading
- water blank: used to zero the spectrophotometer when the reagent does not absorb light
- reagent blank: used to zero the spectrophotometer when the reagent does absorb light
Reagent Blanks
- reagent blanks are done by zeroing the instrument with the reagent in the cuvette
- now the instrument will subtract the absorbance of the reagent from every solution that you measure
- sometimes the reagent is too dark to zero the spectrophotometer on the reagent, so you need to do “manual” reagent blanking
Manual reagent blanking
- zero using a water blank
- place the reagent blank in the spectrophotometer and see how much light it absorbs by itself
—> if your sample uses 1mL reagent with 25uL patient sample, mimic formula for the blank but use water in place of patient sample - manually subtract the absorbance value of he reagent blank from all tube results (standards, patients) for the true, adjusted values
Strange blanks
- specimen blanks: used when a specimen has color due to hemolysis, lipemia, icterus
- this will falsely elevate test values because the color of the specimen also absorbs light
Specimen blanks- how are they done?
- if your assay uses 2.0 mL of color reagent and 50 specimen:
- blank will have 2.0mL water and 50uL specimen
- zero on a water blank then read the absorbance of the specimen blank
- subtract the specimens blanks absorbance from the patients original absorbance of the test
- you have just blanked out the absorbance from the specimen alone
When do you need specimen blanks?
- hemolysis (hemoglobin Hg): hemoglobin absorbs lights from 400-600nm; below 600nm, Hg will interfere, above->no effect
- bilirubin (icteric specimen): bilirubin absorbs light from 400-530nm; below 530nm, bilirubin will interfere, above-> no effect
- turbidity: chylomicrons and VLDL (triglycerides) scatter light of any wavelength
- lipids don’t absorb light, they just scatter it; net result is that you will get falsely high absorbances
End-point Assay
- you add the specimen to the reagent, mix it well, incubate it and then read the final absorbance of the solution after the reaction stops
e. g. biuret total protein assay
Kinetic or Multiple Point Assays
- kinetic assays take 2 or more absorbance readings while the color reaction is going on
- it measures the absorbance change per unit of time (the rate of the reaction)
- the rate of the reaction is proportional to the analyte concentration of the specimen
Kinetic assays- why are they cool?
- the initial absorbance is a reagent blank and a specimen blank for each sample
- you are just measuring the rate of the reaction after that point
- you don’t have to worry about chromophore interference due to hemolysis, bilirubin or lipemia
Biochromatic Analysis (or multiple-chromatic analysis)
- reading the absorbances if each test solution at 2 or more wavelengths
- helps minimize chromophore interference effects (hemolysis, bilirubin, etc)
- primary wavelength- both the analyte and the interfering chromophore absorb light
- secondary wavelength- only the interfering chromophore absorbs light
- the difference between the readings is the absorbance of just the analyte being tested
Turbidimetric Analysis
- based upon the scattering of light by precipitate particles in the cuvette
- common method for measuring urine and CSF protein (mg/dL) range
- specimen is added to SSA and mixed, the proteins start to precipitate
- under tightly controlled conditions, the amount of precipitate formed (light scattering) is proportional to the protein concentration
- mixing and timing is critical, lots of small particles good- but as time goes on you get fewer but larger particles
- light scattering (absorbance) is read using a regular spectrophotometer at a fairly high energy wavelength (415nm)
- broad bandpass instruments do a better job with this kind of assay than do narrow bandpass instruments
Nephlometric Analysis
- also based on light scattering by precipitate particles
- about 1000x more sensitive than regular turbidimetric analysis
- uses a special spectrophotometer called a Nephlometer
- scattered light is detected at an angle to the light source
- can quantitate immune complexes- they scatter light like precipitate particles do
- immuno-nephlometric assays use an IgG antibody against an antigen plus that antigen (analyte)
Example of Immuno-nephlometric assay
- IgG anti-glucose is added to a glucose solution; the immune complexes formed scatter light
- the amount of light scattering is proportional to the glucose concentration
- used to measure hormones, transferrin, ceruloplasmin, A1A, etc.
Fluorescence
A molecule absorbs a photon of high energy light (380nm), becomes unstable and releases energy in the form of a lower energy photon of light (450nm)
Quenching
- refers to false low fluorescence
- can be cause by incorrect pH, temperature, and contaminants
Fluorometric Analysis
- under tightly controlled conditions, the amount of fluorescence is proportional to the concentration of the fluorescent molecules
- very sensitive, 1000x more than turbidimetric and spectrophotometric methods
- very prone to interferences called quenching
- contaminants that fluoresce can cause false increases in measured fluorescence, as can pH and temperature
- can either measure naturally fluorescent molecules or can utilize IgG antibodies that are tagged with a fluorescent compound (called probes)
- works like Nephlometry, except you measure fluoresced light at an angle to the exciter light source
Fluorometric Probes
- florescein (florescein isothiocyanate)
- Lucifer yellow VS
- Rhodamine B Isothiocyanate
- Phycobiliprotein
- Europium (b-Naphthyoyl-Trifluoroacetone)
- Methylumbelliferone (MUP)
- Methylumbelliferone Phosphate (MUPP)
Fluorescence Polarization (FPIA)
- a jazzed-up fluorometric, which polarizes the light that excites the stuff in the cuvette
- also has a polarized interference filter in front of the detector that will only let through light that is polarized on the correct plane
- polarization trick reduces the amount of positive quenching (false high fluorescence) due to contaminants
Polarizing fluorometer
- All molecules spin when excited by the light. The little fluorescent molecules spin very fast, the larger (antibody-bound) molecules spin slower
- only
The antibody-bound fluorescent analyte will spin slow enough so that the fluoresced light is in the proper polarized plane to be detected.
FPIA
- the more analyte the patient has, more of the antibody will bind to it and less is bound to the analyte-probe
- fluorescence in the correct light plane is inversely related to the samples analyte level
- high fluorescence=low patient levels
- low fluorescence=high patient levels
Chemoluminescence
- some chemicals give off a “flash of light” (scintillation) when they become oxidized
- this is detected in a special instrument similar to a nephlometer, where the only light involved is that given off by the reaction
- super sensitive (2000x spectrophotometry)
Biochemiluminescence
Very similar, but a special form of chemiluminescence wherein the reaction is mediated by an enzyme
Acceptable range for control
- test the control pool many times (n=30, 60 is better)
- exclude any outliers (>3SD)
- when all remaining values fit within +/-3SD then,
- set the acceptable range at +/-2SD around the mean
- this range is called the 95.5% confidence interval
Standard
- a solution with an exact known concentration of an analyte
- used as a reference point, with which you can assign analyte concentrations to other specimens such as unknowns and controls
- also known as a “calibrator” and may or may not be biological
Control
- A biological solution which has been previously assayed many times and the target value and acceptable range is known
- used to assess the accuracy and validity of the test
– Space biological source with a similar matrix as the patient samples such as serum CSF urine etc. - if you can assay 2 levels of controls and get acceptable values for both of them, then we can assume that the calibration curve is accurate and patient values are reliable
- if we can’t get acceptable values for the controls, then something is wrong
-do not ever report patient values when you are out of control, this is a CLIA violation
Control is out, what do you do?
- disregard all patient values
- plot all controls
- make sure you ran the correct controls, they are not expired, no floaties, or signs of degradation
- check all instrument settings, volumes used
- repeat entire run
Causes of a control being out
- bad calibration, reagents, operator, pipettor, instrument, wavelength or control
- wrong reagent (lot # or expired) or temperature
- fluke
Random Analytical Error- probability of a control out?
- 1 out of 20 times, or 4.5% probability
- although we know this will happen, 95.5% of the time we expect random analytical error to allow the controls to be within their acceptable ranges
- remember, if your control is out, the greatest probability is that technical error has caused it
Random, technical and systemic error
– Random analytical error: pure chance your data will be outside too but within 3SD
Random technical error: scattering error, due to imprecision
- systemic error: bias error that causes and accuracy, E.G.shifts and trends
Controls out again, what do you do?
-controls moving in the same direction? Calibration issue
– controls moving in opposite directions? Sloppy tech work
- one control near the main, other out of range?
– probably just that one control- contaminated, concentrated by evaporation, or diluted
When do I run controls?
- batch assay
- 2 controls, each and every run, unless specified otherwise
- random analysis analyzer
- 2 levels of controls once per day, or more as specified by the instrument manufacturer
Bad standard? Bad assay?
– your controls will tell you this, they will move away from their means and be out of acceptable limits
– if the standard (calibration) is bad, both controls move in the same direction
Bad standard? Standard diluted or low
- diluted, or you used a lower standard than you were supposed to by mistake
- both controls are high (low standard=high controls)
Bad standard? standard concentrated or high
– Concentrated or you used a higher standard than needed
- both controls are low (high standards=low values)
- both controls will be out of range in the same direction
- can not assume anything about patient values based on this because we don’t know what they should be
Shift
- 5 or more values that deviate from the mean in about the same degree
- it represents a new mean in the controls
- represented by 5 or more consecutive values to be on the same side of the mean
- cause by things that change suddenly such as: new lot # of reagents, bad standard, new lamp, new pipette, new tech, etc
Trend
- 5 or more values moving in the same direction
- caused by something changing very gradually or slowly: deteriorating reagents/standards, light source wearing out, pipette wearing out, temperature drifting
Why do we use QC charts?
- to detect shifts and trends
Only work if every point is chronologically documented/plotted
2-2s rule
- can be split into 3 different sub-rules
- if a control has 2 consecutive 1-2s flags——REJECT!
Proficiency testing (PT)
- mandated by CLIA-88 for all moderate/high complexity tests
- a validation of your lab that verifies your lab, instruments and methods can produce quality results
- 3 times a year (challenge cycles)
- 5 unknowns or challenges per each analyte
- minimum passing score=80%
- your lab results are compared to the other labs who ran the same samples
- if you do not get >/=80% for each analyte
- you have to sen PT service (&CMS) documentation as to why you failed and how to correct it
- you have to pass the next 2 cycles or lose your CLIA license for that analyte or group of analytes
Standard Deviation Index (SDI)
- developed and used when all testing variables cannot be matched
- e.g. my lab uses a homemade reagent or lot number of reagents are different
- passing SDI values= +/- 2 deviations
(Your value-Peer value)/standard deviation of peer group= SDI
- CAP require that you document what went wrong for every Allstate which gave an SDI of 2.0 or greater
- CAP forwards this information on to CMS/CLIA
Competency testing and documentation
- documents that each individual can perform the test accurately
- one on one instruction and demonstration, reading SOP, passing a written quiz, correctly assaying several unknowns under observation
- must be done annually (CAP requirement)
Quantitative tests
- designed to accurately measure specific fluids
- examples: blood glucose assays, electrolyte assays for Na+ and K+
- used when there is a “normal range” of analyte and are measured and reported as mg/dL or mMol/L
Qualitative tests
- designed to detect but not quantitate the presence of analyte in body fluids
- usually not detectable in healthy people, but is when disease is present
- typically reported as positive/negative for the analyte
- could be an antibody, foreign antigen, or hormone
- examples: mono-spot test(mononucleosis), RPR syphilis rest, and urine pregnancy test
Normal ranges
- the range of values typical of healthy, non-diseased people
- alternate names: expected ranges/values, reference intervals
Therapeutic range
- similar to normal range but applies when a patient is taking prescription drugs or antibiotics
- denotes the blood level range of the drug which will offer the desired therapeutic effect
Critical value
- another name is critical action value
- compounds or analytes in the blood that will cause catastrophic effects to patients if the blood levels deviate drastically from the normal range
Ex: blood glucose and blood potassium
Accuracy
- ability to obtain the correct value
Precision
- ability to reproduce the same value over and over again
Sensitivity (of an assay)
- refers to how small or how great an analyte concentration the test is able to detect or accurately measure
- typically we are concerned with the minimum amount of analyte capable of being accurately measured
Specificity (of an assay)
- concerned with the ability of the method to detect one and only one analyte
Predictive value
- refers to the tests ability to correctly classify diseased and non-diseased patients
True positives (TP) And True negatives (TN)
TP: patients that had abnormal results and were really diseased
TN: patients that had normal results and were non-diseased
False positive (FP) And False negative (FN)
FP: patients that had abnormal results and were really non-diseased
FN: patients that had normal results and were diseased
% sensitivity
- ability to detect disease from a population of diseased patients
%sensitivity= [TP/ (TP+FN)]x 100%
sensitivity CANNOT be increase without a loss in specificity
% specificity
- ability to yield normal results from a population of non-diseased patients
%specificity=[TN/(TN + FP)] x 100%
*sensitivity CANNOT be increase without a loss in specificity
Gaussian distribution attributes
1- most of the values are clustered around the mean
2- the distribution is symmetrical. For every value that is 5% higher than the mean, there is a value that is 5% lower than the mean
3- there is a very predictable manner in how the values will be found within +/- 3 SD FROM THE MEAN, 95.5% of the values will be found within +/- 2 SD from the mean, 68.3% of the values will be found within +/- 1 SD from the mean
2 types of QC charts
1- Shewhart
2- Levey-Jennings
Quality assurance
(The overall goal of quality control)
- to assure that all test results produced by the lab provide accurate and timely information to the physician
Statistical quality control (SQC)
- the practice of assaying QC specimens when performing each test
- it is very important but can only monitor variables or sources of error that are directly related to the performance of the particular test
Laboratory quality assurance program (QA)
- monitors all variables or sources of error, beginning with the ordering of the test by the physician through the reporting of the laboratory info.
- SQC is one part of the QA program
A good QA program includes procedures which monitor and standardize things such as:
1- ordering of the test by the physician 2- preparation of the patient by the nurses 3- specimen collection and transport 4- specimen receiving and processing 5- statistical QC procedures 6- reporting of test results 7- preventative maintenance of all lab equipment 8- proficiency testing 9- validation of tests used by the lab 10- hiring qualified personnel 11- personnel safety 12- in-service training and continuing education 13- providing job satisfaction
