Lab Tests Flashcards
reference intervals
“normal”
derived by screening many healthy individuals
encompasses central 95% of normal population
some distributions are not gaussian. examples?
things like cholesterol, LDL, tryglyceride, and HDL
how do we transform non gaussian data?
nonparametric method
.025 * (n+1)
.975 * (n+1)
troponin
biomarker for myocyte damage
anything above 99th percentile represents myocyte damage
two ways to determine reference intervals
established via parametric/non-parametric means
OR
via decision limits and treatment guidelines (A1C)
critical tests
any test that always require rapid communication to the ordering provider, whether the results are normal or not
critical result
any lab finding that requires a rapid communication of results
results represent a potentilly life threatening action
3 phases of laboratory testing
pre- deciding to test, sample collection, processing, transport etc.
pre causes most errors
analytical- receiving specimen, analysis, result
post- physician acknowledge result, implementation of treatment
preanalytic errors
order entry- similar names, duplicate order, trnascription
patient prep
needle size
tube selection
tourniquet
improper storage, delayed processing
postanalytic errors
transcription error
calculation errors
errors in result inrepretation
analytic variation
differences d/t assay imprecision
bias
spectral interference
many assays incorporate a spectrophotometer
hemolysis produces spectral interference
pseudohyponatremia
specific interference for indirect ion-selective electrodes
essentially a dilutional error produced by high concentrations of proteins and lipids (ie if serum is less than 93% water)
hook effect
found in ELISAs w/ markers w/ high levels in a pathological state
results in a false low/negative
loss of self tolerance
pts develop autoAbs to proteins, causing them to stick around and read as artifically high when there is no disease