Session 1 Introduction Flashcards
When would a test have perfect sensitivity and specificity?
if there was complete separation of the normal and diseased populations by the test
what is the identity line?
the line where y=x, slope of 1
what is accuracy?
the degree to which measurements have a constant bias in 1 direction away from true value
what is precision?
random, but unbiased component of measurement error.
if little scatter of values, so values lying along line of identity, then high precision- so limited variation
what would be considered in deciding whether to carry out a physiological measurement?
acceptability of test for patient cost effect on patient management effect on treatment outcomes whether it will add info to current diagnosis probabilities sensitivity and specificity
what would a low specificity of a test imply?
the test is not good at giving -ve results when they are actually -ve, so there would be lots of false +ves, and hence over diagnosis, so may be making lots of patients have a tment with horrible SEs when they don’t even have the disease
main applications for physiological measurement?
diagnosis patient monitoring screening research sports medicine
what may cause a systematic bias?
miscalibration of equipment
standardisation and training can improve this
Clinical applications of a pressure measurement?
arterial blood e.g. heart failure venous blood bladder oesophageal tissue intracranial intrauterine intraoccular- e.g. glaucoma
physiological measurement used in cardiology clinical setting?
heart rate stroke volume aBP blood gases temperature electrical activity of heart
problems of physiological measurement?
access
safety and comfort
environment
interference
what is the hierachy for acceptance of a physiological measurement?
social impact- difference at pop level- QOL, life expectency, disability improvements?
patient outcome- better outcomes? clinical trials, condition dependent, tment dependent
therapeutic impact- does it lead to new tmnets
diagnostic impact- does it lead to new diagnoses, redundancy?
diagnostic accuracy- detect disease?- sensitivity, specificity- accuracy, ROC curve, reliability, predictive value- prevalence of condition in test pop.
technical demands- does it work?- safety testing, accuracy and precision, pilot studies, environmental testing
components of reproducibility?
measurement error- sytematic error- bias- accuracy, random- noise- precision
physiological variability- systematic- physiological interference, system design fluctuations, random- noise and fluctuations
Biological variability
causes of systematic measurement error?
equipment
measurement technique
operator dependent- inter-observer variability
causes of random measurement error?
equipment- weak signals
environment- temp
subject movement
