Formula Sheet
Formula Sheet
Formula Sheet
Epidemiologic measures are used to quantify the frequency of diseases in a population, measure
the association between exposures and diseases, and address the potential impact of an
intervention. Here’s the outline of Chapter 3:
3.1 Measures of Disease Frequency 1
Incidence Proportion (Risk, Cumulative Incidence)
Incidence Rate (Incidence Density, Average Hazard, Person-Time Rate)
Prevalence
Odds 2
3.2 Measures of Association between an exposure and outcome (disease)
Absolute Measures of Associations (“Risk Difference”)
o Risk Difference
o Rate Difference
o Prevalence Difference
Relative Measure of Association (“Relative Risk”)
o Incidence Proportion Ratio (Risk Ratio)
o Incidence Rate Ratio (Rate Ratio)
o Prevalence Ratio
o Odds Ratio
3.3 Measures of Potential Impact
Attributable Fraction in the Population
Attributable Fraction in Exposed Cases
Preventable Fraction
3.4 Rate Adjustment
Direct adjustment
Indirect adjustment (and the SMR)
1
Measures of disease frequency are often called “rates” (even though only the “incidence density” is a true
rate, mathematically speaking).
2
Incidence proportion and prevalence “count data” may also be expressed as odds.
No. of onsets
Incidence Rate =
person-time
In a cohort, Σperson-time can be estimated by summing individual person-time.
In an open population, Σ person-time ≈ (average population size) (duration of follow-up).
Actuarial adjustments may be needed when the disease outcome is not rare.
Examples of vital rates:
births
Crude birth rate (per m) = m
mid-year population size
deaths
Crude mortality rate (per m) = m
mid-year population size
deaths 1 year of age
Infant mortality rate (per m) = m
live births
No. of cases
Prevalence Proportion =
No. of individuals in the study
Also called prevalence proportion and point prevalence.
The prevalence odds = prevalence / (1 – prevalence). When the disease is rare (or at least not too
common), prevalence proportion ≈ prevalence proportion.
One may hear reference to a period prevalence. The period prevalence should be avoided because
it confuses the concepts of incidence and prevalence.
Prevalence is dependent of the incidence and average duration of disease according to this
formula: Prevalence (incidence rate) (average duration of illness)
Reporting
To report a risk or rate as a unicohort (i.e., number of individuals expected to produce one case),
take its reciprocal and report the risk as 1 per “unicohort.” For example, an incidence proportion of
1
0.0025 = 1 in or “1 in 400.”
0.0025
To report a risk or rate “per m” simply multiply by m. For example, an incidence proportion of
0.0010 = 0.0010 × 10,000 = 10 per 10,000. Use a population multiplier that derives a nice whole
number numerator (e.g.,
Rates should be reported with 3 significant digit accuracy. (Intermediate calculations should use at
least 4 significant digits.)
Notation and terminology: These concepts apply to incidence proportions, incidence rates, and prevalence
proportions, all of which are loosely called “rates.” Let R1 represent the rate or risk of disease in the
exposed group and let R0 represent the rate or risk of disease in the non-exposed group.
Absolute Measure of Association/Effect (Risk or Rate Difference) is captured in the rate or risk
difference (RD)
RD R1 R0
R1
RR
R0
Note: The relative effect of an exposure can also captured by the SMR (see section on Rate
Adjustment)
Measures of Potential Impact are used to quantify the effect of removing a hazardous exposure. The two
main measures of potential impact are:
R1 R0 RR 1
The attributable fraction among exposed cases is AFe , or equivalently, AFe .
R1 RR
R R0
The attributable fraction in the population is AFp , or equivalent, AFp AFe pc
R
where pc represents proportion of population cases that are exposed
D+ D− Total
E+ (Group 1) A1 B1 N1
E− (Group 0) A0 B0 N0
M1 M0 N
For person-time data, delete cells B1 and B0 and let N1 and N0 represent the person-time in group 1
and group 0, respectively.
A1
For cohort and cross-sectional data, the risk (count data) or rate (person-time data), R1 and
N1
A0 A1 B0
R0 . For case-control data. ignore the formulas for risks and rate and use OR as a
N0 A0 B1
“Relative Risk” measure. For matched-pair data, see text. For matched tuple sets, use WinPEPI’s
“PairsEtc” program E. ‘Yes-no’ variable: compare subjects with 2 or more controls.
Validity of an Ascertainment
Test + TP FP TP + FP
(those who test positive)
Test − FN TN FN + TN
(those who test negative)
Total TP + FN FP + TN N
(those with disease) (those w/out disease)
Bayesian equivalents for determining predictive values based on prior probabilities and
test parameters are available in the text.
TEN
COMMANDMENTS
FOR DEALING
WITH
CONFOUNDING
Source: EPIB-601 McGill University, Montreal, Canada [email protected],
http://www.teachepi.org/documents/courses/Ten%20Commandments%20for%20Dealing%20with%20Con
founding.pdf
I. Always worry about confounding in your research, especially at the design/protocol stage. Try to use
design elements (e.g. randomization) that will help reduce potential confounding.
II. Prior to the study, review the literature and consider the underlying causal mechanisms (e.g. draw
causal diagrams such as directed acyclic graphs [DAGs]). Then make sure you collect data on all
potential confounders; otherwise you will not be able to adjust for them in your analyses.
III. Know your field or collaborate with an expert who does! Subject-matter knowledge is important to
recognize (e.g. draw causal diagrams) and adjust for confounding.
IV. Use a priori and data-based methods to check if the potential confounders are indeed confounders that
should be adjusted for.
V. Use stratified analyses and multivariable methods to handle confounding at the analysis stage. Choose
the multivariate model that best suits the type of data (e.g. dichotomous vs. continuous) you collected
and the design you employed (e.g. case-control vs. cohort).
VI. Do not adjust for covariates that may be intermediate causes (on the causal pathway between the
exposure and disease). Do not adjust for covariates that may not be genuine confounders. And beware
of time-varying covariates that will need special approaches.
VII. Use matching with great caution. Use analytic methods that are appropriate for the design used; for
example, if matching was done, use methods that take matching into account (e.g. conditional logistic
regression, matched pairs analyses).
VIII. Always consider effect measure modification, but perform and interpret subgroup analyses with
caution. The subgroup analysis should be one of a small number of hypotheses tested, and the
hypothesis should precede rather than follow the analysis (i.e. subgroups must be pre-specified).
IX. Always remember that adjustment for confounding can be inadequate due to residual confounding
because of unmeasured confounders, misclassification of confounders, and inadequate adjustment
procedures (e.g. model misspecification, categorization of continuous covariates).
X. If conventional methods prove to be inadequate, consider using newer approaches such as propensity
scores, matched sampling, instrumental variables and marginal structural models. However, make sure
you work with statisticians who understand these new methods (not many do).
When all else fails, pray! If prayer fails, consider changing professions!!