Unit 4-5 Flashcards
What is the difference between prevalence and incidence?
prevalence is a proportion while incidence is a risk/rate (dynamic)
Number of students in a course who are sick vs number of students in a course who reported sick at any point over the semester describes what types of prevalence?
point vs period prevalence respectively
Incidence x Duration = ?
prevalence (existing cases)
The true rate. The speed at which new diseases or health outcomes occur is called the?
incidence rate
What is the formula for incidence risk?
IR = (#of new cases in a specified time period) / [initial number at risk (NAR) - 1/2 (withdrawls)]
To calculate incidence rate (IRR - incidence rate), there are two types of denominators. What are they?
Exact Denominator.
Approximate denominator.
What is the exact denominator?
net time individuals in a population are at-risk during time
What is the approximate denominator?
average number at risk (NAR) from start (NARinitial) to (NARfinal) - endof the follow up period
The average of these x internal time component *ITC)
this is the “true” rate
What is the ITC?
internal time component:
equal to or less than time between initial and final measurements
Crude vs Cause specific
to describes morbidity and mortality risk and rates
Crude
captures all causes and types of disease and death
Cause-specific
express level of disease or death caused by particular exposure/factor
Crude risk
number developing any disease (morbidity) or dying (mortality) within specific period of time / number at risk minus 1/2 withdrawls
Crude rate
number developing any disease (morbidity) or dying (mortality) within a specified period of time divided by average number at risk multiplied by internal time component
Cause specific mortality rate
number dying of a specific disease within specified period of time divided by average number at risk multiplied by internal time component
Case fatality “rate”
= (#who die from a particular disease) / (#who are sick from a disease)
it is used to measure mortality of particular disease
Attack “rate”
= (#who get sick following a specific exposure) / (#who have that specific exposure)
What are 3 measures of association?
- Relative Risk
- Odds Ratio
- Incidence Risk Ratio
If association is <1, how would you interpret this association?
negative association <1
no association = 1
positive association >1
remember these are ratios, so no difference between groups = 1
What is the equation for relative risk?
= (risk of disease outcome in exposed group) / (risk of disease or outcome in non-exposed group)
= risk of disease E+ / risk of disease E-
= (a / (a+b)) / (c/(c+d))
What is the equation for odds ratio?
= (odds of disease or outcome in exposed) / (odds of disease or outcome in non-exposed group)
= odds of disease in E+ / odds of disease in E-
=(a/b)/(c/d) = ad / bc
What is the equation for IRR?
= (incidence rate of outcome in exposed group) / incidence rate of the non-exposed group)
= (a1/t1) / (a0/t0)
What is risk or rate difference?
RD is the subtraction of background level of risk in the nonexposed group from exposed group
RD = risk or rate in exposed - risk or rate in non-exposed
What is risk difference for rates?
RD = a1/t1 - a0/t0
rate in E+ - rate in E-
What is the null value for RD?
0
if less than 0 then exposure is negatively associated
if RD >0 positively associated
RD = 0 then exposure has no differential effect on outcome
What does attributable proportion exposed (APe) express?
the proportion of risk/rate of outcome in exposed group that is attributable to specific exposure
APe is expressed in a percentage true or false.
true
Population Attributable Risk or Rate (PAR)
prevalence or rate of outcome in population - prevalence or rate of outcome in non-exposed group
Population Attributable Fraction
(PAR: prevalence of outcome in population - prevalence in nonexposed group) / (prevalence of outcome in population)
What are 4 reasons that census are impractical?
- time consuming
- costly
- tedious
- difficult if participation is voluntary
What are 3 stages to sampling?
- Who/what to sample
- How you’re going to choose
- How many you’ll need to be confident
What are 2 criterias before sampling to choose subjects?
- inclusion criteria
- exclusion criteria
What is the difference between target, source, study/sample population?
Target population: population that you can extrapolate data from
Source population: population from which study of subjects drawn
Study/sampled population: group of sampling units that was selected from sampling frame
What is the sampling frame?
list of all study subjects in source population
What is the difference between internal and external validity?
Internal: degree to which observed findings in study lead to correct inferences about outcome of interest in source population
External: degree to which inferences drawn from study can be generalized or extrapolated to broader population of interest
When every member of a source does not have equal probability of being selected, this is called…
non-probability sampling
What are the 3 types of non-probability sampling?
convenience sampling
judgement sampling
purposive sampling
What is purposive sampling?
when sampling units are chosen because of their exposure/disease status
What is judgement sampling?
investigator chooses what they think is representative of population
What are 5 types of probability sampling?
- simple random sampling
- systematic random sampling
- stratified random sampling
- cluster sampling
- multistage sampling
What is the difference between cluster and multistage sampling?
they are similar in which clusters/groups are chosen, but in cluster sampling all individuals are sampled whereas multistage, only a proportion of individuals are randomly selected and measured
What is systematic random sampling?
sampling at sampling intervals (every nth subject)
How do you calculate estimated variance for proportions?
p*q
= p is proportions that have the disease * q is proportion of individuals that do not have the disease (1-P)
How do you calculate variance for means?
σ2 = Σ(x-x̄)2 / n
note variance is equal to standard deviation squared
Zα
is the level of confidence
typically choose to have 95% confidence and 5% significance
The width of confidence intervals or how tight your confidence intervals are is the
precision.
range of values around sample estimate that includes the true value of sample proportion
What is “L”
allowable error = precision
i.e. accurate within 10% of true proportion
If there is a large association, it will be _____ to detect. If there is a small association, it will be _____ to detect
large association = easier to detect
small association = will be more difficult
Required sample size generally increases as….
- size of difference between two means or proportions decrease (smaller difference)
- level of power to detect difference between two groups increases
- number of confounding variables you are controlling for study increases
- number of hypotheses testing increases
“cumulative incidence” is known as
incidence risk
