Lecture 3 Flashcards
risk
- the probably that the outcome will occur given a particular set of circumstances
- also called measures of association
risk difference
- the risk (incidence) of disease in one group minus the risk (incidence) of disease in another group
attributable risk
- risk difference when referring to factors that increase the risk of disease
- I_E-I_U
I_E = incidence in exposed I_U = incidence in unexposed
attributable risk percent
- the percent of the risk in the exposed group that is attributable to the exposure
- percentage of risk among the exposed group that would be eliminated if the exposure had not occurredI_E-I_U
________
I_E
absolute risk reduction
- the proportion of patients who were spared an adverse outcome due to the treatment
I_C - I_Rx
I_C=incidence in control group
I_Rx=incidence in treated group
relative risk
- the probability of an outcome given an exposure (risk factor or treatment) compared to the risk without the exposure
RR = I_E/I_U OR
I_Rx/I_C for treatment
when the relative risk is greater than 1.0
- the exposure or treatment increases the risk of disease
when the relative risk is less than 1.0
- the exposure or treatment decreases the risk of disease
when the relative risk = 1.
- the exposure or treatment does not increase or decrease the risk of disease
relative risk reduction
- the percentage of baseline risk that is removed as a result of a given therapy
RRR = ARR/I_c = (I_c - I_Rx)/I_c
RRR=1-RR
number needed to treat
- how many patients need to be treated to prevent one outcome event
NNT = 1/ARR
standard deviation
- a measure of the degree of variability in individual measurements in a study
- how much variability there is in measurements from individuals in a sample
for a variable that is distributed normally
- 68% of the values will be between one standard deviation
95% will be between two standard deviations
standard error
- a measure of the dispersion of a sample means around the population mean
- how much uncertainty there is in the group values
standard error = SD/ sqrt (sample size)
null hypothesis
- there is no difference in outcomes between comparison populations
- Ho is never accepted
type I error
- incorrectly concluded that there is a difference when there is not
- alpha
type II error
- fail to find a difference when a true difference exists
- false negative
p value
- the probability of finding an outcome as extreme as or more extreme than the one we found, assuming that the null hypothesis is true
the lower the p value
- the less compatible the data are with the null hypothesis
The greater the test statistic
- the lower the p value
the lower the test statistic
- the greater the p value
limitations of p values
- mix together
- size of an effect
- sample size
- variability of the data
point estimate
- the specific numeral result of a study
confidence interval
- the calculated range of values surrounding the point estimate that are consistent with the true effect
if null value is outside of the 95% confidence interval
p < 0.05
if the null value is within the confidence interval
p > 0.05
values near the point estimate
- more consistent with the population value
values near the limits of the confidence
- less consistent with the population value
degree of precision of an estimate
- indicated by the width of the confidence interval
the narrower the confidence interval
- the more precise the estimate
statical power of a study
- the study’s ability to detect a difference assuming that a real difference exists
- the probability of not making a type II error
clinical importance
- the salience of a finding for clinical practice
- related to the magnitude (size) of the finding
- taking into account the seriousness of the outcome, and the prevalence of the condition