Session 4 Chemical Pathology Flashcards
what can cause physiological variation in measurements in clinical chemistry?
age e.g. serum alkaline phosphatase- high in rapidly growing adolescents and increases gradually from age 40 to 65, also increases in pregnancy
circadian rhythm e.g. cortisol- must measure at same
time of day for comparisons, GH higher on a night
menstrual cycle e.g. progesterone- high- days 14-28
food intake e.g. glucose tolerance test- must know if patient has fasted before glucose measurement- if fasted, >7mmol/L, if not, >11.1mmol/L =diabetes
give an example of pathological variation?
with time after injury e.g. enzyme changes after MI e.g. creatine kinase
how could the clinical value of a test be improved?
combinations of test e.g. free T4 and TSH to assess thyroid status, liver function tests- bilirubin, ALP and ALT
sequential tests to show trend e.g. beta hCG doubling time for ectopic pregnancy- doubling every 2 days is normal, EP= increase is usually less
dynamic function tests- use stressor to uncover differences not apparent in basal samples e.g. synacthen- analogue of active portion of ACTH, want to stimulate cortisol prod as low, should increase cortisol prod by >200nmol/L- normal response usually excludes Addisons- primary adrenal insufficiency
use of venous sampling to localise a hot spot e.g. PTH venous sampling- determine which side prod. more PTH, so can target correct side
what does the reference range usually approximate to?
the mean value +/- 2SD
what sort of diagram is produced from the distribution of values of an analyte in a healthy population, and what does this mean?
skewed ‘‘normal’’ distribution
so 95% of normal (healthy) biochemical analyte values will fall withing reference range, but 5% will fall outside, 2.5% above and below, so these patients would have abnormal results, so may need repeat investigations and follow up
what does a high sensitivity test mean for false -ves?
low number of false -ves: test good at giving a +ve result when the patient does have the disease, so low number of -ve results if patient did have the disease (false -ves)
why might a value within the reference range still mean that a patient was unhealthy?
the reference range is for the population as a whole, not the individual, and the distribution of normal values within an individual is very small, so movement within the reference range may be abnormal for that individual
Define sensitivity in terms of application to a clinical value or efficiency of a test in distinguishing health from disease
the percentage of +ve results in patients who DO have the disease in question
Define specificity in terms of application to a clinical value or efficiency of a test in distinguishing health from disease
the percentage of -ve results in patients who do NOT have the disease in question.
high specificity= low false +ve rate
what is sensitivity in terms of operational performance of the analysis?
lowest concentration of analyte the test can reliably measure
what is specificity in terms of operational performance of the analysis?
ability of test not to falsely cross-react with substances (often closely related chemically) other than the 1 it claims to assay e.g. prednisolone and cortisol
what is accurate in measuring a clinical chemistry analyte?
measured value close to true concentration- possibly obtained by suitable gold standard
what is precise in measuring a clinical chemistry analyte?
repeated analyses of same sample give very similar results
if good precision, it’s easy to re-calibrate and then get good accuracy, but harder to get good precision if poor and accuracy good
what would under-calibration of an assay cause?
poor accuracy
what ideally should methods be?
precise, accurate, robust, inexpensive, quick, sensitive and specific
