Managerial Epidemiology
Chapter 9
Measures of Effect
Learning Objectives
Explain the meeting of absolute and relative effects
Calculate and interpret the following measures: risk difference, population risk difference, etiologic fraction, and population etiologic fraction
Discuss the role of statistical tests in epidemiologic research
Apply Hill’s criteria for evaluation of epidemiologic associations
Effect Measure
A quantity that measures the effect of a factor on the frequency or risk of a health outcome
Three Effect Measures
Attributable Fractions
Measure the fraction of cases due to a factor.
Risk and Rate Differences
Measure the amount a factor adds to the risk or rate of a disease.
Risk and Rate Ratio
Measure the amount by which a factor multiplies the risk or rate of disease.
Absolute vs. Relative Effects
Absolute
Attributable risk is also known as a rate difference or risk difference.
Population risk difference
Relative
Relative risk
Etiologic fraction
Population etiologic fraction
Risk Difference (Attributable Risk)
Risk difference--the difference between the incidence rate of disease in the exposed group (Ie) and the incidence rate of disease in the nonexposed group (Ine).
Risk difference = Ie - Ine
2
Calculation of Risk Difference
For women younger than age 75, the incidence (Ie) of hip fractures per 100,000 person-days was highest in the winter (0.41), and the incidence (Ine) was lowest in the summer (0.29). The risk difference between the two seasons (Ie - Ine) was 0.41 - 0.29, or 0.12 per 100,000 person-days.
3
Population Risk Difference
Measures the benefit to the population derived by modifying a risk factor.
Etiologic Fraction
Defined as the proportion of the rate in the exposed group that is due to the exposure.
Also termed attributable proportion or attributable fraction.
Population Etiologic Fraction
Provides an indication of the effect of removing a particular exposure on the burden of disease in the population.
Also termed attributable fraction in the population.
Statistical Measures of Effect
Significance tests
The P value
Confidence interval
4
Null Hypothesis
Underlying all statistical tests is a null hypothesis, which states that there is no difference among the groups being compared.
The parameters may consist of the prevalence or incidence of disease in the population.
5
Significance Tests
Used to decide whether to reject or fail to reject a null hypothesis.
Involves computation of a test statistic, which is compared with a critical value obtained from statistical tables.
The critical value is set by the significance level of the test.
The significance level is the chance of rejecting the null hypothesis when, in fact, it is true.
6
The P Value
Indicates the probability that the findings observed could have occurred by chance alone.
However, a nonsignificant difference is not necessarily attributable to chance alone.
8
The P Value (cont’d)
Possible meaning of nonsignificant differences: For studies with a small sample size the sampling error may be large, which can lead to a nonsignificant test even if the observed difference is caused by a real effect.
Confidence Interval (CI)
A computed interval of values that, with a given probability, contains the true value of the population parameter.
The degree of confidence is usually stated as a percentage; commonly the 95% CI is used.
Influenced by variability of the data and sample size.
9
Clinical vs. Statistical Significance
While small differences in disease frequency or low magnitudes of relative risk (RR) may be significant, they may have no clinical significance.
Conversely, with small sample sizes, large differences or measures of effect may be clinically important and worthy of additional study.
10
Statistical Power
The ability of a study to demonstrate an association if one exists.
Determined by:
Frequency of the condition under study.
Magnitude of the effect.
Study design.
Sample size.
11
Evaluating Epidemiologic Associations
Five key questions to be asked:
Could the association have been observed by chance?
Determined through the use of statistical tests.
Could the association be due to bias?
Bias refers to systematic errors, i.e., how samples were selected or how data was analyzed.
12
Evaluating Epidemiologic Associations (cont’d)
Could other confounding variables have accounted for the observed relationship?
To whom does this association apply?
Representativeness of sample
Participation rates
Does the association represent a cause-and-effect relationship?
Considers criteria of causality.
13
Types of Associations between Factors and Outcomes
Not statistically associated (independent)
Statistically associated
Statistical Association
When a factor and outcome are statistically associated, the relationship can be:
Non-causal
Causal
Indirect
Direct
Multiple Causality
Also referred to as multifactorial etiology.
“…requirement that more than one factor be present for disease to develop…”
Models of Multiple Causality
Epidemiologic triangle
Web of causation, e.g., in avian influenza
Wheel model, e.g., childhood lead poisoning
Pie model, e.g., lung cancer