B5-073 EBM: Colon Cancer Screening Flashcards
number one risk factor for colorectal cancer
age
risk factors for colon cancer
4
- age
- history of CRC or polyps
- Lynch syndrome
- hx of inflammatory bowel disease
decreasing incidence and mortality due to colon cancer is due to
- detection of early stage cancer
- removal of precancerous polyps
USPSTF grade A recommendation for CRC screening
ages
50-75
USPSTF grade B recommendation for CRC screening
ages
45-49
USPSTF grade C recommendation for CRC screening
ages
76-85
choices for screening tests
4
stool test at home
colonoscopy
sigmoidoscopy
CT colonography
PICO
population
intervention
control
outcome
clusters of individuals are randomly allocated to intervention group
cluster randomization
positive predictive value is dependent on
prevalence
the more sensitive a test, the higher the incidence of
false positive
what test performance characteristic are risk-based screening recommendations looking to improve?
positive predictive value
- plots the true positive rate (sensitivity) on the y axis
- and the false positive rate (specificity) on the x axis
- test accuracy is measured via the area under the curve
ROC curve
probability that a negative test correctly indentifies an individual who does not have that disease
negative predictive value
calculation for negative predictive value
true negatives/
true negatives + false negatives
most important performance characteristic for a screening test
sensitivity
highly sensitive tests produce few…
false negatives
ability of test to correctly identify people without disease
specificity
calculation for specificity
true negatives/
true negatives + false positives
calculation for ARR
mortality in control group - mortality in screened group
deaths/total in control
deaths/total in screened
subtract
calculation for NNT
1/ARR
occurs when an increased ability to detect disease increases the survival time after diagnosis without changing the course of the disease
lead time bias
reason why survival time should not be used as an outcome when evaluating screening tests
lead time bias
calculation for relative difference in mortality
mortality in control - mortality in intervention/
mortality in control
can mask the actual frequency of the outcome in control in intervention groups
relative difference
allows for some understanding of the frequency even if actual numbers aren’t provided
absolute difference
