ep test 2
Terms
undefined, object
copy deck
- Absolute Risk
- Incidence of disease in a population that does not take into account differences in experience for exposed or non-exposed individuals
- Association
- A statistical relationship between two variables, a determinant and an outcome
- Relative Risk (Measure of association)
- indicates the likelihood of disease in the exposed group “relative†to the unexposed in the study a/(a+b)/ c/(c+d)
- Odds Ratio (measure of association)
- indicates the “odds†or likelihood of exposure for cases “relative†to the controls in the study
- Odds Ratio = 2.54
- This means that in your study, persons with (heart) disease were ~2.54 times more likely to have had exposure than those without the disease
- RR and OR
- Relative Risk (RR) and Odds Ratio (OR) are equivalent in their ability to measure the strength of an association
- When is an OR a good estimate of RR?
- When the cases’ exposure is representative of the diseased population’s exposure; When the controls’ exposure is representative of the healthy population’s exposure; AND The disease is relatively rare (less than ~1 per 100,000)
- Attributable Risk
- absolute measure of the effect of exposure compared to non-exposure; the excess occurrence of disease in the study sample due to risk factor. [Incid(exp) - Incid(unexp)] or [a/(a+b)]-[c/(c+d)]
- Attributable Risk Percent
- Also called the etiologic fraction measure of the proportion of disease in exposed group that is due or “attributed†to the exposure, in the study sample [Incid(exp)-Incid(unexp)]/Incid(exp) x 100
- Attributable Risk Percent (alternate eq)
- Either RR or OR can be used AR%=(OR - 1)/(OR) X 10
- Population Attributable Risk (PAR)
- Estimate of the effect of exposure on the disease experience of a total population. Incid(total) - Incid(unexp) or AR X Prev(exp) 1. Incidence in the total (NOT study) population 2. Prevalence of the exposure in the total population (NOT the study)
- Interpretation of RR and OR
- strength of association between exposure and disease a factor in establishing causality 1.0 means no association < 1.0 suggests protective effect; > 1.0 suggests increased risk
- Interpretation of Attributable Risk and Population Attributable Risk
- estimate of the public health impact of the exposure on disease occurrence
- Study Designs and Measures of Association
- Cohort studies -- RR, AR, AR% contribute to computation of PAR, PAR% Case-control studies -- OR, AR% Cross-sectional studies -- OR
- Rates and Percentages Info
- RR, OR, AR and AR% are all calculated from and for a specific study sample PAR and PAR% require info from outside of the study The generalizability of the results are dependent upon the representativeness of the study sample
- Confidence Intervals
- confidence intervals (CI) define the range of values within which the true value for any estimate lies, with a specified degree of statistical certainty
- CI Details
- confidence intervals (CI) reflect the variability of the estimate and indirectly indicate the sample size can be applied to means, proportions or OR and RR CI combined with the p-value from a statistical test provides the most useful information regarding the role of chance
- Stratified Analysis
- Magnitude of confounding is determined by the difference between crude and adjusted risk measures Calculates unconfounded estimates of risk within each level of potential confounder Evaluates both confounding and effect modification