Week 3- Compare Risks And Role Of Chance Flashcards
Attributable Risk
Risk that is attributed to some harmful exposure
(Incidence in exposed)-(Incidence in unexposed)=AR
Attributable Risk Percent
(Ie-Iu)/Ie=AR/Ie=AR%
A quantity that allows you to say X% of cases of lung cancer in smokers could have been prevented by not smoking
Absolute Risk Reduction
Proportion of patients spared of an adverse outcome due to treatment
ARR = (Incidence in Control)-(Incidence in treated)
Relative Risk Ratio
RR = (Incidence exposed or Incidence Treated)/(Incidence unexp. Or incidence not treated)
Allows for the comparison between treated and untreated groups.
Ex: The risk of atrial fibrulation in those treated with Warferin is 1/3 the risk of those without treatment
Relative Risk Reduction
RRR = ARR/Ic = (Ic-Irx)/Ic Or RRR = 1 - RR
Percentage of baseline risk removed by a therapy
Ex: 67% of the baseline risk of 12 strokes per 100 persons per year can be removed as a result of Warfarin treatment.
Number Needed to Treat
NNT = 1/ARR
Number of patients that need to be treated to prevent 1 outcome event. Reciprical of ARR
Standard Error
SE = SD/(n)^.5
Tells you how different the groups means are
Null Hypothesis
H0 is the hypothesis that there is no diffference in outcomes of comparison populations
Type 1 Error
Alpha error, concluding there is a difference when there is not
Type 2 Error
Beta, failing to observe a difference when there is
P-value
The chance that results as extreme or more extreme will occur again (with a different population sample) assuming the H0 is TRUE.
Cutoff to reject H0 is generally 0.05 (researcher has a 5% chance of committing type 1 error)
Point Estimate
Specific numerical result estimate of a study
Confidence Interval
Calculated range of values surrounding the point estimate that are consistent with the true effect.
CI = mean +/- 2 stdev
Statistical Power
Study’s ability to detect a difference assuming that a difference does exist.
Probability of NOT making a type II error (1-B)
Clinical Importance
The salience of a finding for clinical practice and is related to the magnitude of the finding, taking into account the seriousness of the outcome, and the prevalence of the condition.
