Epidemiology Flashcards
What is mortality rate?
Estimate of portion of population that dies during specified time period
What do meaningful statistics in mortality rates need?
Denominator population
Time frame
What is case fatality rate?
Proportion of cases of a specified condition that are fatal within a specified time
What’s CFR?
Case fatality rate
Weaknesses of CFR estimation:
Underestimation in early days of outbreak
Overestimation if denominator is limited (CFR only based on sub-group of cases)
Assumes all cases have been tested
What does more testing allow with CFR rate?
Higher accuracy
What is infection fatality rate?
Proportion of cases of a specified condition that die divided by total infected people
What is IFR?
Infection fatality rates
What does IFR accuracy depend on?
Detection and reporting of a symptomatic or mild symptom cases
What is R0?
Reproduction number - number of cases that can be potentially infected by one case
R0 < 1
Less likely spread of disease
R0 > 1
More likely spread of disease
Characteristics of R0:
Not biologically constant
Give an example of calculating mortality rate:
50 people died of flu x100
——————————————-
Population of fife
Give an example of person-time:
10 deaths per 10000 person years
What can person years mean?
10000 people for 1 year
5000 people for 2 years
Give example of n-year follow up:
5-year mortality of 10 per 10000 people
When are death rates meaningless?
Without denominator population and time
What is incidence?
Number of new cases
Calculating incidence rate:
Number of new people with outcome over time period x100000
——————————————————————————————————
Total number of people (or person-time for people) in group at risk
What’s prevalence?
Proportion of population that has disease
What is point prevalence rate?
At a specified time
Calculating point prevalence rate:
Number of people with outcome at a point in time x100
—————————————————————————————
Total number of people in group
What’s period prevalence rate?
Over specified period
Calculating period prevalence rate:
Number of people with outcome during time period x100
——————————————————————————————-
Average number of people in group
Differences between incidence and prevalence:
Incidence can be rate or proportion whereas prevalence can only be a proportion
Sporadic definition:
Occasional cases occurring irregularly
Endemic definition:
Persistent background level of occurrence
Low-moderate levels
Outbreak definition:
Greater than expected cases of endemic levels
Pandemic definition:
Occurring in more than one continent
What are the different types of exposures?
Non-modifiable (e.g. age)
Modifiable (e.g. weight)
Interventions (e.g. drug therapy)
How do you calculate risk?
Number of outcomes in group
———————————————— x100
Number if people in group
How do you calculate relative risk?
Risk in exposed
—————————
Risk in unexposed
How do you calculate relative risk reduction?
(1 - Relative risk) x 100
How do you calculate absolute risk reduction?
Risk in unexposed - risk in exposed
How do you calculate number needed to treat?
1
————————————
Absolute risk reduction
What are confidence intervals?
Range of plausible values
What does it mean if the confidence interval is wider?
Greater the uncertainty
Cross-sectional study:
Sample population
Estimate proportion + use data
Case-control study:
Select cases with outcome and controls without outcome
Explore exposures + compare
Identify association
Cohort study:
Select people with outcome + classify according to exposure
Follow-up
Compare risk of disease in exposed and unexposed
Randomised controlled trial:
Random allocation
Compare risk of outcome in intervention and control groups
Objectives in cross-sectional:
Prevalence
Objectives of case-control:
Cause
Objective of cohort:
Cause
Prognosis
Incidence
Objective of randomised controlled trial:
Treatment effect
Time-frame of cross-sectional:
Past
Time-frame of case-control:
Past
Time-frame of cohort:
Prospective = future
Retrospective = past
Time-frame of randomised controlled trial:
Future
What is confounding?
True relationship confused by third factor
What may bias involve?
Systematic error (how data is used)
What does bias lead to?
Wrong conclusion about effectiveness and causation