Samenvatting HNE31006
Study design and interpretation in epidemiology and
public health
College 1 – Introduction to the course
Learning objectives:
- Understand the characteristics of major epidemiological study designs,
including potential methodological threats with respect to validity and
confounding.
- Critically review and interpret epidemiological studies reported in the
literature
- Design, plan and review studies used in epidemiology and public health,
taking methodological and practical considerations into account
- Understand the principles of a systematic literature review and meta-
analysis (overall result)
College 2 – Refresher
Which results can be trusted? As epidemiologist you need to criticize and able to
say which one you believe, cause there is always ‘inconsistency’ and
‘heterogeneity’ in studies.
How to conduct a good epidemiological study?
Two broad types of epidemiology
Descriptive epidemiology = frequency/amount of disease or other
characteristics in a population (person, time, place)
Typical study design: survey, ecological, cross-sectional study
What? Who? When? Where?
Analytical epidemiology = study of the causes of disease, exposure-
disease associations
Typical study designs:
Observational: ecological, cross-sectional, cohort, case-control study
Experimental: randomized controlled trial
+ Why?
Analytical epidemiology: two study types
- Observational = ecological (or correlation) study, cross-sectional study,
case-control study and cohort study
- Experimental = randomized controlled trial
The main difference is that in a trial the investigator assigns the exposure
(randomly). In observational there is no interfering but just look at what people
are doing.
1
,Observational studies
Cross-sectional study
So information is gathered at the
same time.
2
, Cohort study
Case-control study
A case-control study is back in
time (retrospective) (direction
of inquiry)
So time is the key
Observational study designs:
In cohort
studies
normally
problems with the response rate (only about
30%)
Be careful with confounders: in a study n
pancreatic cancer and coffee: coffee was
concluded to be dangerous to pancreatic
cancer risk while this was confounded and
the effect was in real life not there.
Experimental studies:
- Randomized controlled trial
- Field trials/community
intervention trials
Design of a trial. So
the random allocation
means that the researcher puts people in
a group.
Questions:
1. The Cancer and Steroid Hormone (CASH)
study, in which women with breast cancer and a comparable group of
women without breast cancer were asked about their prior use of oral
3
Study design and interpretation in epidemiology and
public health
College 1 – Introduction to the course
Learning objectives:
- Understand the characteristics of major epidemiological study designs,
including potential methodological threats with respect to validity and
confounding.
- Critically review and interpret epidemiological studies reported in the
literature
- Design, plan and review studies used in epidemiology and public health,
taking methodological and practical considerations into account
- Understand the principles of a systematic literature review and meta-
analysis (overall result)
College 2 – Refresher
Which results can be trusted? As epidemiologist you need to criticize and able to
say which one you believe, cause there is always ‘inconsistency’ and
‘heterogeneity’ in studies.
How to conduct a good epidemiological study?
Two broad types of epidemiology
Descriptive epidemiology = frequency/amount of disease or other
characteristics in a population (person, time, place)
Typical study design: survey, ecological, cross-sectional study
What? Who? When? Where?
Analytical epidemiology = study of the causes of disease, exposure-
disease associations
Typical study designs:
Observational: ecological, cross-sectional, cohort, case-control study
Experimental: randomized controlled trial
+ Why?
Analytical epidemiology: two study types
- Observational = ecological (or correlation) study, cross-sectional study,
case-control study and cohort study
- Experimental = randomized controlled trial
The main difference is that in a trial the investigator assigns the exposure
(randomly). In observational there is no interfering but just look at what people
are doing.
1
,Observational studies
Cross-sectional study
So information is gathered at the
same time.
2
, Cohort study
Case-control study
A case-control study is back in
time (retrospective) (direction
of inquiry)
So time is the key
Observational study designs:
In cohort
studies
normally
problems with the response rate (only about
30%)
Be careful with confounders: in a study n
pancreatic cancer and coffee: coffee was
concluded to be dangerous to pancreatic
cancer risk while this was confounded and
the effect was in real life not there.
Experimental studies:
- Randomized controlled trial
- Field trials/community
intervention trials
Design of a trial. So
the random allocation
means that the researcher puts people in
a group.
Questions:
1. The Cancer and Steroid Hormone (CASH)
study, in which women with breast cancer and a comparable group of
women without breast cancer were asked about their prior use of oral
3
