ADI Final
hazard function
analyzes the likelihood that an item will survive to a certain point in time based on its survival to an earlier time
number needed to treat
average number of patients who need to be treated to prevent one additional bad outcome inverse of the absolute risk reduction
confounding can be dealt with
by randomization, matching, stratification analysis, or multiple regression analysis
Log-Rank Test
compare survival curves
OR = 1
disease/ outcome and exposure are independent
RR = 1
disease/ outcome and exposure are independent
number needed to harm
how many persons on average need to be exposure to a risk factor over a specific period to cause harm in an average of one person who would not otherwise have been harmed inverse of the absolute risk increase
disadvantage of stratification
inability to control simultaneously for multiple confounding variables
when less effective treatment is applied more frequently to less severe cases,
it can appear to be a more effective treatment
censoring
key feature of survival data occurs when we have some information about individual's survival time but we don't know the exact survival time
Odd Ratio
must be a positive number is symmetric (reversing the roles of disease and exposure makes no difference) (a)(d) / (b)(c)
Relative Risk (Risk Ratio)
must be positive not symmetric
confounding
occurs when the relationship between exposure/ treatment and disease/ outcome is distorted by the influence of a third variable or group of variable (confounders)
Mantel-Haenszel Method
pooling method that combines stratum-specific RRs or ORs
hazard rate
rate of death for an item of a given age
as you increase the number of strata,
sample size becomes a major problem because many of the strata contain few or no people
Cochran-Mantel-Haenszel Test
statistical test for assessing association between exposure/ treatment and disease/outcome, controlling for confounding
survival function
the chance that an individual is still alive after time t
OR is approximately equal to RR when..
the disease is rare
Non-Inferior Margin
the predetermined margin of difference between the new and standard treatments
OR > 1
there is a greater odds of disease/ outcome when exposed than unexposed
RR > 1
there is a greater risk of disease/ outcome when exposed then unexposed
OR < 1
there is less odds of disease/ outcome when exposed than unexposed
RR < 1
there is less risk of disease/ outcome when exposed than unexposed
Why are RR and OR used?
to quantify the relationship between exposure and disease