KEY POINTS
Studies
Principal Steps
Independent
Variables, X
Potential
confounding, Z
Descriptive studies:
Describe the
Hypotheses Do not have a
frequency,
can be
comparison
Starting at
natural
assessed
the bottom of history, and
group, do not
the research
through more
allow
possible
hierarchy
rigorous
assessment of
determinants
research
associations
of a condition
Cross-sectional study:
a snapshot in time
Sometimes termed
Outcome and
a frequency survey
exposure are
or a prevalence
ascertained at the
same time
study
Temporal relation
Unclear:
Obesity and
Artrithis?
Cohort study:
looking forward in time
Logical
sequence:
from
exposure to
outcome
Follow the
exposed and
unexposed
groups
forward in
time
Exposed
group
develops a
higher
incidence,
then the
exposure is
associated
Enables
calculation of
true
incidence
rates, relative
risks, and
attributable
risks
Case-control study:
thinking backwards
Starting
with an
outcome,
looks
backward
in time for
exposures
Prevalence
Especially
exposure is
useful for
higher among
outcomes
cases than
that are rare
among
or that take a
controls, then
long time to
the exposure
develop
is associated
Controls
should be
similar to
cases, except
for not
having the
outcome
Non-randomised trial:
penultimate design?
Do not randomly
allocate
participants to
exposures
The way a
non-randomised
Designate this
trial is done
research
and analysed
design as class II-1
resembles that of a
cohort study
Measurement of outcomes:
Confusing fractions