L2: Study design part 2 Flashcards
(10 cards)
how can study designs be classified?
Study designs can be classified as prospective
and retrospective. If the researcher has collected
information about the exposure before the outcome event, a study is considered to be
prospective. In contrast, if information about the
exposure is collected after the outcome, the study
is retrospective. Thus, this classification refers to
time of information collection.
ecological studies?
Ecological studies
Studies that investigate risk factors of health outcomes in which the unit of analysis is at the group level rather than the individual.
Group measures (exposure and or outcome) can include:
summary measures of a group (mean, average rate)
environmental factors (air pollution, hours of sun-light, fast-food shops)
i.e. something that is not measured at the individual level
Examples:
Time trends, geographic comparisons
Ecological studies use data from a population
level, defined based on geographical or temporal
characteristics. Individual information on the
potential relation between exposure and outcome
is therefore missing. In the context of cancer,
there are many examples investigating dietary
and lifestyle risk factors using geographical or
temporal ecological study designs. A main threat
to the validity of ecological studies is the lack of
individual information. We usually have little or
no control over the distribution of important
exposures and/or confounders in the populations
and a specific sub-group of the population may
be over or underexposed. Therefore, biological
interpretation from ecological data is limited.
Even if one population is at large much more
exposed to a certain factor, it may not be the
exposed ones but rather the un-exposed minority
in that population is the source of most of the
outcomes we are looking for.
Geographical:
Common Use: Many ecological studies compare health outcomes or exposures across different geographical locations (like countries, regions, or cities). For example, researchers might compare the rates of lung cancer between countries with different smoking rates.
Focus: The geographical factor often serves as a proxy for environmental, social, or healthcare differences that might influence health outcomes. In these studies, the unit of analysis is typically the group (e.g., a country or a city) rather than individuals.
Example: Comparing the incidence of heart disease in cities with different air pollution levels.
Temporal:
Common Use: Ecological studies can also look at changes over time within a given population. For example, comparing the incidence of a disease over several decades and relating it to changes in exposure to risk factors like diet, lifestyle, or environmental factors.
Focus: The time aspect is important for examining trends, such as the rise or fall in disease prevalence, and the corresponding shifts in behaviors or environmental factors over time.
Example: Examining changes in the rate of cancer over the years in a country and correlating it with dietary changes or public health policies.
Are Ecological Studies Always Defined by Temporal or Geographical Factors?
Not necessarily. While geographical and temporal factors are commonly used to categorize and analyze data in ecological studies, the term “ecological” primarily refers to the unit of analysis being the group (or population level) rather than the individual level.
Ecological studies are defined by the analysis of group-level data, and while many studies involve temporal or geographical data, ecological studies can focus on other factors, such as:
Socioeconomic groups: Comparing health outcomes among different income or education levels in a population.
Cultural or lifestyle factors: Comparing populations with different lifestyle factors, like diet or physical activity, across various regions or time periods.
advantages and disadvantages of ecological studies?
Advantages:
Easy to do
No individual data necessary
Good to generate ideas about potential associations
Disadvantages:
No information on the individual level
Not able to account for other factors that might explain the association
ecological fallacy
The ecological fallacy occurs where an analysis of group data is used to draw conclusions about the individual.
Example:
The average salary is higher in countries that sell more reading glasses
Therefore if you wear reading glasses you are likely to have a higher salary
Likely to be due to other factors that are not taken into account (confounders). results may differ between ecological and individual-level studies.
cross-sectional studies?
The process
Select a sample (representing the population of interest)
Measure exposure and outcome variables at the same time
Determine prevalence
Measure of association: odds ratios
Cross-sectional studies, also called prevalence
studies, involve all persons (or a subset) in the
population at a specific time of ascertainment of
their health status. These studies do not include a
longitudinal component and are hence often used
to describe the prevalence of a disease since both
newly and previously diagnosed disease will be
enumerated. The American National Health and
Nutrition Examination Survey (NHANES) is a
programme of studies designed to assess the
health and nutritional status of adults and children. It is run by the National Centre for Health
Statistics with the aim to produce vital and health
statistics for the nation [13]. NHANES includes
information from an interview on demographic,
socioeconomic, dietary and health-related issues
as well as an examination involving medical,
dental and physiological assessments. It is thus
used to determine the prevalence of major diseases and risk factors for diseases
screenshot slide 29
strengths and weaknesses of cross sectional studies?
Strengths
fast and inexpensive
immediate answers – no follow-up time
no loss to follow-up (but can have non-responders)
Weaknesses
can’t determine temporal relationship
not good for rare exposures or outcomes
bias can be a problem - measurement bias, survivor bias
cohort study design?
Cohort study design
The Process
Start with the POPULATION of interest
Identify or assemble a cohort
Measure risk factor(s) and potential confounders
Measure the outcome over the follow-up period
Can be
Prospective
Start with assembling a cohort, measure risk factors then follow over time to measure outcomes.
or
Retrospective (historical)
Identify a suitable cohort (from the past), collect risk factor data measured in the past, collect subsequent outcome data . So, the data collection is done in the present, but the analysis looks at the past to determine exposure before the outcomes.
The most common observational study design
in epidemiology is a cohort study, which refers to
a cohort of people who are followed overtime to
identify whether they develop an event of interest. Usually a cohort is considered to be a fixed
roster of study participants who stop contributing
person-time either when they develop the outcome of interest, die, are lost to follow-up or at
the end of the study. The cohort members can be defined at one point in time or recruited over
some inclusion time. An RCT is a special case of
a cohort study comparing two (or more) cohorts
based on their exposure status.
The European Prospective Investigation into
Cancer and Nutrition (EPIC) is a cohort study
with more than half a million participants
recruited across ten European countries and followed for almost 15 years. It was designed to
investigate the associations between different
exposures to diet, nutritional status, lifestyle and
environmental factors and the incidence of cancer and other chronic diseases [10]. Another
well-known cohort study is the American Nurses’ Health Study (NHS) [11], which is one of
the largest and longest running investigations of
factors that influence women’s health. It started
in 1976 and the information provided by the
238,000 dedicated nurse-participants has led to
many new insights into health and disease.
strengths and weaknesses of a cohort study?
Cohort studies (in general):
Can establish sequence of events
Can assess risk of multiple outcomes at the same time
Can estimate incidence (how many new events within a certain time)
Able to directly calculate absolute and relative risk
Prospective cohort studies:
Can control who is in the cohort
Lower risk of bias (exposures measured before outcomes)
Retrospective cohort studies:
More efficient: less time, less costly
but…
Cohort studies (in general):
not a controlled experiment - so can’t claim ‘causation’
difficult to control for all other confounding factors
expensive – often require large sample
not good choice for rare outcomes
Prospective:
not timely, long follow-up
potential loss to follow-up (can lead to bias)
Retrospective:
knowing the outcome might lead to bias
limited to data already collected
little control over who is in the cohort
case-control studies?
Select a sample of ‘cases’ (i.e. people who have the condition/disease)
Select a sample of ‘controls’ (i.e. people without the disease but who have the same chance of having the disease)
Measure (past) exposure to risk factors of interest
Case-control studies
Selecting study subjects
Results may be biased if exposures are different due to the selection process
CASES – all those who develop a disease (or random sample)
Cases from a hospital
Cases on a population registry (i.e. cancer registry)
Best to be new cases
CONTROLS must come from a source population with similar chance of being exposed to the risk factor of interest
Hospital-based controls with different disease
Population controls (good if cases are from a registry)
How many controls?
Statistical power of higher for case-control studies so sample sizes are lower
1:1 selection is often OK especially when number of cases is large
Ratio of controls can be increased when number with outcomes is low or when the proportion likely to be exposed is low
1:2 is common
not much power gained by increasing >3
Case-control studies
Matching:
Controls can be matched with cases to ensure they are comparable with respect to other influencing factors (confounders)
age and sex matching is common - strongly linked to many diseases and exposures
sometimes geographical area
individual matching
frequency matching (overall proportions are the same)
Matching is not necessary but can be useful to control for strong confounders
Overmatching can ‘hide’ true effects, and can limit the analysis that can be done
case control studies measurement bias?
Case-control studies measurement bias:
Information about exposures collected after the outcome is known
can introduce bias:
different level of recall/reporting of risk factors between cases and controls
more details about risk factors in cases (because of the disease)
researchers look harder for evidence of exposure in cases
Possible solutions
use data collected before outcome was known
researcher blinded to outcome
subjects asked about multiple possible risk factors
pick controls have a disease that is linked to similar risk factors
Case-control studies are the other typical
design for non-experimental or observational
studies. The design starts with identifying those
who have the outcome (cases) and a set of controls. A case-control study can be seen as the
sampling of an existing or tentative cohort. Cases
and controls should come from the same
person-time experience, i.e. study base. The term
“nested case control study” is often used when
the cases and controls are drawn from an existing
cohort study. Cases and controls are compared
on the basis of an exposure variable for which
information was collected either retrospectively
or prospectively. Most case-control studies are
retrospective because information about the
exposure is collected after the cases and controls
are defined. The Northern California Childhood
Leukaemia Study is an example of a populationbased case-control study of risk factors for
childhood leukaemia [12]. Cases were identified
from paediatric oncology centres in the Northern
California region and controls were identified
from birth certificates. Information on maternal
diet was collected using a food frequency questionnaire to assess the diet 12 months prior to
pregnancy. The statistical analyses showed that
women with a high intake of fruits and vegetables had children with a lower risk of leukaemia.
Case control studies can be very effective in
several situations. When the outcome under
study is rare, it would be extremely expensive
and time-consuming to collect all relevant
information on exposures in a large cohort study
as it would only result in a few cases. Moreover,
when one wants to study an exposure that is very
expensive or time-consuming to measure (i.e.
staining for a specific biomarker in cancer
patients to predict recurrence), it may be more
suitable to conduct a case-control study.
strengths and weaknesses of case control?
Strengths
Useful for rare outcomes (eg specific cancers)
Efficient/less costly than cohort study: smaller sample size; no follow-up required
Weaknesses
Biases if cases and controls come from different populations
Biases due to measuring exposure after the outcome
Confounding due to other influential factors (not measured)
Can only study one outcome
