Vous êtes sur la page 1sur 21

Types of

Epidemiological studies
Epidemiology - Terms
• Epidemiology comes from Greek
– Epidemos (“on people”)

• Related to the words


– Epidemic
– Demographics

• Epidemiology is the study of diseases in


populations of humans or animals
– How
– When
– Where, they occur
Epidemiologists
• Epidemiologists attempt to determine what
factors are associated with disease
– RISK FACTORS

• Also what factors protect people against


disease
– PROTECTIVE FACTORS

• The first epidemiologists looked at possible


causes of contagious diseases
– Smallpox
– Typhoid
– Polio
Modern Epidemiology

• Epidemiology has been expanded to study


non-transmissible disease
– Cancer
– Poisonings due to environmental agents

• Molecular epidemiology focused on the


biochemical and genetic changes associated
with specific diseases

• Specific biochemical “footprints” indicate


different types of exposure of cellular
damage.
– Eg. Cyclobutane pyrimidine dimer DNA damage and
UV exposure
Epidemiology -
Can:Cannot
• Epidemiological studies cannot prove the
causation of any disease.

• They cannot say that a specific risk factor


causes the disease being studied.

• Epidemiological studies can only show that


the risk factor is associated or correlated
with a higher incidence of the disease in the
exposed population.
Examples
• The link between smoking and lung cancer.

• Based upon comparisons of rates of lung cancer


in smokers and non-smokers

• Rates are much higher in populations which


smoke.

• This does not prove smoking causes lung cancer


– It identifies a RISK FACTOR
Example
• To prove smoking causes lung cancer it was
necessary to expose animals to tobacco
smoke and smoke extracts.

• Needs highly controlled conditions with a


defined control group.

• Only difference between controls and


exposed group is the exposure to smoke
etc.

• Such studies provided a CAUSAL link


between smoking and lung cancer.
Types of Epidemiological
Study
• Descriptive Epidemiology

• Distribution of cancer cases within a population


according to certain characteristics.
– Geographic factors
– Occupation
– Age Sex
– Residence etc

• Risk Factor Epidemiology

• Combination of environmental factors associated


with the disease
Descriptive Epidemiology

• INCIDENCE
–A measure of frequency of the cancer in a given
population

• INCIDENCE RATE
Number of people with cancer in a specific time
period
Total population at that time

Incidence is therefore a PROBABILITY that tries to


estimate risk that a person will develop cancer
during a given time.

Often expressed per 100,000 of the population


Incidence Rates – Breast
Cancer
Descriptive Epidemiology
• PREVALENCE: The number of people in the
population with cancer at a specific time.
– A proportion and not a rate

• Number of people with cancer


• Total population

• Annual prevalence is the proportion within


a single year

• Lifetime prevalence: during a complete


lifetime
Descriptive Epidemiology
• Prevalence of cancer with a good
prognosis will be higher than incidence
(lots of people living with disease).

• Low prevalence with poor prognosis


(people die).

• Cancer generally has a high incidence


because it’s a long-term illness

Prevalence = Incidence x duration


Mortality Statistics
• The number of people who die from cancer.

• A useful tool in descriptive epidemiology

• Expressed as the number of deaths per


100,000 of the population.

• The only completely reliable source of


information

• UK in 2002 there were 155,180 cancer deaths.


Analytical Epidemiology
• Descriptive epidemiology observes trends.
Occurrence, prevalence and incidences but
cannot identify risks and potential causative
factors.

• Analytical epidemiology is based upon


comparisons of groups of people with
different characteristics.

• Two methods
– Case controlled studies
– Cohort studies
Case-controlled studies
• Investigation into groups of people with
cancer and groups of people without
cancer

• Determine differences in
– Exposure
– Occupation
– Behaviour
– Lifestyle
– Geography
– Location
– Socio-economic factors
– Genetic susceptibility etc
Case-controlled Studies
• They are retrospective and data is obtained
by various methods (interviews,
questionnaires etc)

• They estimate RISK or RELATIVE RISK.


Exposed Non-exposed
Cancer A C
No Cancer B D

• Relative risk = AxD


BxC
Study Bias
• Many different types of bias can affect
case-controlled studies.

• Difficult to get well matched study groups.

• Difficult to control for all patient


– Behaviour
– Exposure
– Honesty etc.
Cohort Studies
• Cohort studies are prospective studies.

• A cohort of healthy people is selected and


studied for many years.

• The characteristics of those that get cancer


in this group are then closely studied.

• These characteristics can the be identified


and compared to those in the group that do
not get cancer
Cohort Studies
• Cohort studies may be much more
reliable than case-controlled studies.

• They are however


– Time consuming
– Complicated
– Very expensive

• Most of the bias is eliminated because


there is no need for a control group
Cohort Studies
• Fixed cohort studies: people are chosen
due to a specific event occurring at a
specific time
– Cohort does not take on new members

• Open dynamic cohorts are common in


cancer epidemiology

• Cohort allows for in and out migration.

• New members can be added due to


attritional loss.
Other Issues
• Confounding factors
– Unrecognised variable: Smoking often associated
with cirrhosis but does not cause it.
– Caused by excessive smoking (associated with
heavy drinking).

• Molecular epidemiology
– Tumour marker
– Mutational gene profile
– Toxins

Vous aimerez peut-être aussi