Exposure

Question 1

what is a proxy variable

Answer 1

a proxy is used for the variable of real interest, if it is difficult or infeasible to measure the other variable
e.g occupation for social class

Question 2

what is validity

Answer 2

an instrument measures what it is supposed to accurately

Question 3

what is reliability

Answer 3

consistency
degree of stability exhibited when a measurement is repeated under identical conditions

Question 4

what is sensitivity

Answer 4

proportion with positive results amongst those who TRULY have the attribute

Question 5

what is specificity

Answer 5

proportion with a negative result amongst those who truly do NOT have attribute

Question 6

what is the positive predictive value

Answer 6

proportion with true positive result amongst all those with positive result

Question 7

what is negative predictive value

Answer 7

proportion with true negative result amongst all those with a negative result

Question 8

how can we calculate how the specificity and sensitivity of the screening test change as the threshold is altered?

Answer 8

using a ROC curve - receiver operating characteristic
- its a plot of sensitivity against 1-specificity
the area under the curve compares different tests

Question 9

what are some judgment based assessments

Answer 9

face validity
content validity (all relevant components measured)
consensual validity (agreement of judges/experts)

Question 10

what are data based (empirical) assessments of validity

Answer 10

criterion validity - concurrent validity, predictive validity (against future event), responsiveness (against change)
construct validity - known-group validity, convergent and discriminant validity (different measure)

Question 11

how can reliability be measured

Answer 11

performing 2+ independant measurements and comparing the findings using an appropriate statistical test

Question 12

in what situations is reliability assessed

Answer 12

within instruments - assays, subjects and observers
between instruments

Question 13

what are approaches which ensure reliability

Answer 13

standardisation and calibration
establishment of test-retest reliability
random allocation of subjects to observers
multiple measurements

Question 14

what is the source pathway receptor model

Answer 14

hazard
exposure
body burden
target organ
physiological effect
clinical effect
death

Question 15

what is cumulative exposure

Answer 15

summed amount of exposure over study period

Question 16

what are personal exposure measurements

Answer 16

biological monitoring
e.g dosimeter to measure exposure to ionising radiation

Question 17

why should we have caution when using biological monitoring data

Answer 17

may only reflect recent exposure
subject to inter-individual variation (metabolism, excretion)
subject to measurement error
can add expense, logistical problems and ethical concerns

Question 18

what is dose reconstruction

Answer 18

collecting other info to estimate the dose receieved

Question 19

what are problems of personal level exposure assessment in epidemiology

Answer 19

may be multiple environmental media - air, water,soil,food and routes
expensive and time consuming
individual level dosimeter measurements are often not possible
may be mixture of exposures e.g water may contain metals but also microbes and pesticides

Question 20

what is a study which is an example of an exposure matrix

Answer 20

paternal occupational contact and risk of childhood leukemia
case control study based on cancer registry data
using radon detector

Question 21

what is differential misclassification

Answer 21

misclassification to exposure is DEPENDANT on individuals disease status

Question 22

what is non-differential misclassification

Answer 22

misclassification to exposure is INDEPENDANT of the individuals disease status

Question 23

what does nondifferential misclassification often bias results towards

Answer 23

the null - diminishes the magnitude of any association - but not always

Question 24

how do we minimise the impact of differential exposure misclassification

Answer 24

is it higher for those in higher or lower exposures?
considered biases associated with retrospective info collection
can we use contemporary records or measurements

Question 25

how do we minimise the impact of non differential exposure misclassification

Answer 25

can the exposures be assessed with higher validity/reliability?
assess the level of variability - can be used to adjust for relationship between exposure and disease