Epidemiology (V) Descriptive Flashcards
What is descriptive epidemiology?
A first step that allows us to make comparisons between groups in order to better understand determinants of disease
(set up questions to be investigated further in formal studies that test hypotheses)
What might it mean if morbidity/mortality trends are changing over time?
Some cause of the disease or the effectiveness of treatment is changing over time
Why could a change in morbidity/mortality over time be an artifact of the data?
Disease definition changing, diagnostic capabilities are changing, screening is changing (e.g. lead-time bias), reporting is changing, the population at risk is changing, etc.
What type of bias is the result of a new screening tool diagnosing a disease earlier than before (and thus artificially increasing the disease survival time)?
Lead-time bias
What kind of questions does descriptive epidemiology answer?
What is happening; who it is happening to; where is it happening?
What kind of questions does analytic epidemiology answer?
Why/how is it happening?
Is case fatality a true rate?
No
When are case fatality rates not very useful?
For chronic diseases where competing causes of death may skew the data
What is a 5-year survival rate?
The number of people alive 5-years post-diagnosis / total number of people diagnosed
What is the 5-YSR analogous to?
1 - cumulative incidence of death
What are three caveats of 5-YSRs (follow-up, bias, survival curves)?
- 0% attrition
- No lead-time bias
- Does not take into account when individuals died within the 5-year period
What does it mean that bias is systematic?
It is an error that will result in repeatedly skewed results in a predictable way
What is lead-time bias?
Earlier identification of the disease leads to a perceived increase in disease survival time from diagnosis to outcome.
