Renske de Veer (rdeveer)
MIN17 BMR - Premaster BMW
1. Epidemiologie
Glossary
Age-adjustment Measures of disease frequency- risks, rates, prevalences- are
often age-adjusted when used to compare different groups,
e.g., different categories of exposure, or different locations.
In particular, the comparison of crude death rates is
frequently misleading because such rates do not account for
factors such as age that can influence mortality. This is why
age-adjusted rates are typically used instead of crude rates.
The goal of age-adjustment [rates] is to modify the crude
rates so that any difference in rates cannot be explained by
age distribution of the comparison groups:
- The direct method of age-adjustment re-computes the
rates by substituting a common age distribution for the
separate age distributions of the groups being
compared.
- The common age distribution is determined by
identifying a standard population.
- Potential confounding variables of interest other than
age, e.g., race and sex, can also be adjusted, both
individually and simultaneously. We generally use the
term rate adjustment to describe adjustment involving
any type or number of confounding variables and any
type of measure of disease frequency, whether a risk,
rate, or prevalence.
Association measure Quantifies the relationship between exposure and disease
among the two groups (risk ratio, rate ratio, odds ratio).
Correlation between determinant and frequency of disease
outcome.
Attributable proportion population Part of total disease incidence that is due to the determinant
(APt = (It – I0)/It → (p(RR-1))/(p(RR -1)+1) ) for which APt =
Attributable proportion, It = incidence total population and I0 =
Incidence unexposed. Incidence that would be observed if the
population were entirely unexposed, compared with its
current (actual) exposure pattern.
Attributable proportion exposed Part of disease incidence in exposed group due to determinant
(APe = (I1 – I0)/I1 → (1-(1/RR)).
Attributable risk (excess risk) Difference in rate of a condition between an exposed
population and an unexposed population (AR = Ie – Iu; for
which Ie = Incidence exposed and Iu = Incidence unexposed.
The percentage can be calculated by: (AR/Ie) * 100. Difference
,Renske de Veer (rdeveer)
between two incidences. (Amount of extra cases of a disease
related to exposure to a determinant). Can be calculated in
cohort studies.
Average rate [like speed] Describes how rapidly disease or death has been occurring as
an average over a period of time. Rather than instantaneous
rate.
Bias Bias is to have an opinion, or view, that is often without
considering evidence and information.
Blinding Information about test is masked from the participant to
reduce or eliminate bias, until after a trial outcome is known.
Double blinding: Both tester and subject are blinded. Without
blinding → information bias, influence external variables.
Case (patient)-control study Type of observational study in which two existing groups
differing in outcome are identified and compared on the basis
of some supposed causal attribute. Have two groups of
interest, look at their history and compare them and then
draw a conclusion. Only ODDS ratio can be calculated.
Cohort study Longitudinal study that sample a cohort (group of people who
share a defining characteristic) performing a cross-section at
intervals through time. Can be retrospective (looking back in
time) or prospective (requiring collection of new data).
Retrospective cohort study is suitable to study disease with
long latency periods, because of gain of time. Compare two
groups → follow over time → compare outcomes.
Closed population (cohort) Population in which a specific event is of undetermined time.
All men born between 1950 and 1955 in Nijmegen.
Confounder A confounder is an independent risk factor for the outcome
variable and is simultaneously related to the exposure variable
of interest. The effect or association between exposure of
interest and outcome is therefore distorted by the presence of
this other confounding variable. A confounder should be an
independent determinant of the disease, associated with the
studied determinant and no intermediare between the
determinant and the disease.
Confounding by indication Biased results in intervention studies due to selective
allocation of treatments, based on patient characteristics with
prognostic significant. This means that if doctors are allowed
to decide about treatment allocation, those who are recruited
from the treatment group have an apriori different prognosis
compared to those who are recruited for the control group.
Cross-over trial Intervention study in which patients are sequentially allocated
to all intervention groups under study. Effect of intervention
must be reversible and effect of the first intervention has no
influence on the effect of the second intervention.
Cross-sectional study Type of observational study that analyses data collected from
a population at a specific point in time. Can be used to describe
OR, AR and RR from prevalences.
Cumulative incidence A measure of disease frequency during a period of time
(number of new cases with disease X/total population at risk
at a specific point in time = %). Can be used for closed
populations. Gives the proportion of the persons that
develop a disease.
,Renske de Veer (rdeveer)
Can be computed using the formula for simple cumulative
incidence CI = I/N,
- N denotes the size of a disease-free fixed cohort at the
start of follow-up
- I denotes the number of cases of disease that develop over
the follow-up period.
- ^Ci stands for sample , Ci stands for population. Used for a
closed population.
Problems that can occur:
- Size of cohort may be reduced during follow-up: death,
loss, so it is unknown if they have developed the disease
- Dynamic cohort [continually changing, allowed to enter
and leave] rather than fixed cohort [no entries]
- subjects may be followed for different periods of time:
loss to follow up-, withdrawn, enters after the study starts
and are disease-free after the study ends, if the follow-up
times varies for study objects.
Cumulative risk Measure of total risk (between 0 and 100%) to develop a
certain disease during a given period of time. This can be
directly measured in a cohort, indirectly in a dynamic
population.
Effect modification Occurs when the magnitude of the effect of the primary
exposure (recent use of contraceptives) on an outcome (heart
attack) differs depending on the level of a third variable (age).
Epidemics Rapid spread of infectious disease to a large number of people
in a given population within a short period of time.
Epidemiology Study of the patterns, causes, and effects of health and disease
conditions in defined populations
Hawthorne Type of reactivity in which individuals modify and aspect of
their behaviour in response to their awareness of being
observed.
Hazards ratio =1 no effect, <1 the lower the chance, >1 more chance of
getting sick. Adjusted hazards ratio: correcting for the factors
which have influence on getting sick.
Impact measures Quantifies the impact of health risk and benefits in a
population (APt and APe)
Incidence A measure of the probability of occurrence of a given medical
condition in a population within a specified period of time.
(number of new cases during some time period; proportion).
Incidence density (rate) Number of new cases per population at risk in a given time
period in which the dominator is the sum of the person-time
of the at risk population (number of new cases with disease
X/total person-time of observation at specific point in time =
n/tijd). Can be used for open/dynamic populations.
Information bias Bias arising from measurement error. Information bias is also
referred to as observational bias and misclassification.
Nondifferential misclassification is when all classes, groups or
categories of a variable have the same error rate or probability
of being misclassified for all study subjects. Differential
misclassification occurs when the error rate or probability of
being misclassified differs across groups of study subjects.
Prevent by blinding or objective measurement tools.
, Renske de Veer (rdeveer)
Instantaneous rate [like velocity] describes how rapidly disease or death is occurring at a
moment in time.
Intention-to-treat analysis Comparing of groups of patients as they were initially allocated
at randomisation, irrespective of whether a patient complies
to treatment or not. Even patients who never ever started with
the allocated treatment are included in the original group
during analysis. (Alternative: per-protocol analysis).
Mantel-Haenszel procedure Test association between a binary predictor or treatment and
a binary outcome such as a case or control status while taking
into account the stratification (confounding). Sum is for each
of the 2x2 tables. For 1 confounding factor. Regression analysis
for more than 1 confounding factor.
Matching Match the two groups being compared on a couple of
confounding factors.
Mortality rate/risk Measure of the number of deaths in a particular population.
can be risk or rate, depending on the study design used.
Mortality risk or rate can be measured in a number of ways,
including disease-specific mortality risk or rate, all-causes
mortality risk or rate, and case-fatality risk or rate.
Mortality rate can be measured using the general formula for average rate,
namely, IR = I /PT.
I= number of deaths observed notified over a specified study
period [due to specific disease]
PT= person time for initial cohort [with specific disease of
interest]
Depends on specific disease mortality or all causes mortality
Mortality risk The formula for simple cumulative incidence, namely, CI = I /
N, can be used.
I = number of death due to the specific disease of interest.
Depends on specific disease mortality or all causes mortality.
N= size of cohort [regardless of disease status]
Case-fatality risk proportion who die from a particular disease
Odds ratio Quantify how strongly the presence or absence of property A
is associated with the presence or absence of property B in a
given population. Ratio exposure to determinant in diseased
and non-diseased. OR > 1 = risk factor (exposure determinant
more frequent diseased compared to non-diseased), OR = 1
(similar frequency in exposure of determinant) and OR < 1 =
protective factor (exposure determinant less frequent in
diseased compared to non-diseased. (Ratio exposure to
