Appraising Diagnostic Research Flashcards
(25 cards)
Reference Interval
range of scores that captures individuals without a movement problem.
Cohort Designs
most common design for development of diagnostic test.
Prospective cohort study
group of participants who are likely to develop a condition or outcome are followed into the future.
retrospective design
researchers look back in medical records for information about their participants that relates to their research question
prospective studies of diagnostic tests plan to collect what?
collect data on both the clinical test under the study and a gold standard comparison test
prospective design can be either
cross sectional or longitudinal
cross sectional study
participants are sampled and measured at one point in time (prospective, cross-sectional) data obtained once and within a short time frame
longitudinal design
participants are followed over a longer time frame to determine is target condition is confirmed.
Quality assessment of diagnostic accuracy studies (QUADS)
includes 14 questions to guide the appraisal of the applicability and quality of diagnostic study.
A dichotmous variable has how many categories ?
2 - an example of this (gender: female and male)
continuous variables used for diagnostic studies -
has a specific cutoff point is determined to define a positive and negative result
false postitive
person is negative for the condition according to the gold standard but tests positive on the reference test
Is a type I error - alpha
false negative
person test positive on the gold standard but negative on the reference test
Type II error - beta
sensitivity (snout)
ability to rule out a condition.
Does not have a lot of false negatives
specificity (spin)
Ability to rule in a condition
receiver operating characteristics curve
is a graphic representation of the sensitivity and specificity values that are generated from a series of cutoff points.
positive predictive value
are computed from the rows in 2x2 table
positive predictive value express what?
participants with a positive result on the reference test who also have a positive result on the gold-standard.
negative predictive vale
participants with a negative result on the reference test who also have negative results on the gold-standard test.
likelihood ratios
Change of probability
positive likelihood ratio
expresses the numeric value of a positive test of the movement problem is present versus absent
negative likelihood ratio
expresses the numeric value of a negative test if the movement problem is present
pretest probability
can be estimated from the prevalence of the problem in the population
posttest probability
the problem exists, thereby warranting performance of the test