Epi Biostats Flashcards
What is lead-time bias, and what can it cause?
The difference in time between diagnosis from a screening test and diagnosis from symptoms. Lead-time bias can make it seem as though patients with the condition live longer with screening, when really they just know longer.
What is length-time bias?
“A preference for finding slow-moving conditions over fast-moving with screening at regular itervrlsusually associated with cancer screenings. If the slow-moving are less fatal catching them with screening might make it seem as though you have better survivability with screening when really you are just more likely to detect less serious cancers.”
What is over diagnosis bias?
“The diagnosis of ““disease”” that would not actually have caused symptoms or death in the patient’s lifetime. (For example, why we stop screening for slow-moving cancers like cervical cancer in older patients.)”
What is referral bias?
The chance of exposed cases being admitted to the study is different to exposed controls. This happens frequently when cases are selected in a hospital whose activity is linked to the studied exposure.
What is the difference between reliability and accuracy?
Reliable tests give results at the same level accurate tests give correct information. (A test that gives a wrong answer 60% of the time is reliable but not accurate.)
How do you calculate the sensitivity of diagnostic tests?
the # who have definitive disease and test positive divided by the # with definitive disease (How likely a positive result actually is positive)
How do you calculate the specificity of diagnostic tests?
the # who do not have definitive disease and test negative divided by the # without definitive disease (How likely a negative result is actually negative)
How do you calculate the positive predictive value (PPV) of diagnostic tests?
The # with actual disease AND test positive divided by total # who test positive (many false positives make PPV small)
How do you calculate the negative predictive value (NPV) of diagnostic tests?
The # without actual disease AND test negative divided by total # who test negative (many false negatives make NPV small)
What are some criteria of an internally valid study?
Few confounding factors or biases
What are the criteria of a good screening or diagnostic test?
inexpensive, safe, valid and reliable
What is negative predictive value?
the probability of not having actual disease given a negative test.
What is positive predictive value?
The probability of having actual disease given a positive test.
