HSC7004 Healthcare Analytics
Lecture 2 - Cancer Epidemiology
Cancer Epidemiology
Teaching
contents for
Healthcare Basics of Cancer Genetics
Analytics
-Population
Health Cancer Mutation Data
Informatics
Analysis
Precision Medicine – P4 Medicine
“P4 medicine” standards for
Predictive, Preventative, Personalized, Participatory.
P4 medicine recognizes the physiological state of
each individual and reflects a unique combination of
• genetics
• lifestyle
• environmental factors
• health history
Nat Rev Clin Oncol. 2020;17(3):183-194 [Reading Material 03]
Cancer Epidemiology
Teaching
contents for
Healthcare Basics of Cancer Genetics
Analytics
-Population
Health Cancer Mutation Data
Informatics
Analysis
What is Epidemiology?
What is Epidemiology?
“The study of the distribution “The branch of medicine that
and determinants of disease deals with incidence,
or health status in a distribution, and possible
population”. - Centers for control of diseases and other
Disease Control in Atlanta factors relating to health.” -
Georgia Wikipedia
Epidemiology
Studying the distribution (who, when, and where) and
determinant (what) of health and disease conditions at population level
Identifying risk factors for public health, policy making, and evidence-
based practice for disease prevention
Focusing on epidemic diseases, cancer, and many non-disease, common
health-related conditions, such as blood pressure, depression, obesity
The Origin of
Epidemiology
• Dr. John Snow used non-medical means
to discover the source of the Cholera
outbreak in 1854.
Cholera outbreak in London, 1854
Historical Origins of
Epidemiology
• John Snow was one of the founders of
modern epidemiology
• His Innovation: He used dot maps and
numbers to locate cholera cases and
sources.
• He identified the homes taking water
from sewage-polluted sections had a
cholera rate fourteen times higher than
the other regions.
The Origin of
Epidemiology
“As soon as I became acquainted
with the situation and extent of
this irruption of cholera, I
suspected some contamination of
the water of the much-frequented
street pump in Broad Street.” –
John Snow
Types of Epidemiologists:
public health epidemiologist
Types of Epidemiologists:
clinical epidemiologist
Clinical research
and practice
Individual
professional Clinical Patients Choice
experience epidemiology
Epidemiological
research and
practice
Types of Epidemiologists:
research epidemiologist
Types of studies used in epidemiology
Case-series study: Qualitative study of a single patient over time, or small group of
patients with a similar diagnosis, with the potential to produce illness within given period
Case-control study: A group of disease-positive individuals (the "case" group) is compared
with a group of disease-negative individuals (the "control" group). The control group
should ideally come from the same population that gave rise to the disease cases.
Cohort study: The study subjects should be at risk of the outcome under investigation at
the beginning of the cohort study; they should be disease free when the cohort study
starts. The cohort is followed through time to assess their later outcome status.
Prospective and retrospective study
Prospective study: design the
Retrospective study: looks
study, recruit subjects, and
backwards and examines
collect data on all
exposures to suspected risk
subjects, before any of the
or protection factors in
subjects have developed any
relation to an outcome.
of the outcomes of interest.
Some Terminology in
epidemiological research
Some Terminology: Exposure
Statistics: The variable that predicts/causes an effect is an
independent variable.
Epidemiological research: The variable that predicts/causes an
effect is an exposure.
Some Terminology: Outcome
Statistics: The variable that is affected by the independent variable
is a dependent variable.
Epidemiological research: The variable that is affected by the
exposure is the outcome.
Some Terminology: Exposure and Outcome
Statistics Dependent Variable Independent Variable
Epidemiology Outcome Exposure
E.g. Lung cancer Smoking
Some Terminology: risk factors
Epidemiological Research
Exposures that INCREASE or DECREASE the likelihood of developing
certain disorders, conditions, or diseases are called risk factors.
Prevalence, Incidence, Confidence Intervals
• Prevalence
The number of existing cases of disease in a population in a given time
• Incidence
The number of new cases in a population in a given time
• Confidence Intervals
A range of value likely (95%) to contain the true value in a population
Presence of
Morbidity disease in a
population
Morbidity and
Mortality
Occurrence of
Mortality death in a
population
Mortality Rate
• Crude rates (i.e., mortality
rates)
• Specific rates (i.e., disease-
specific mortality rates)
Crude rate
Crude rates are based on actual
(unadjusted) number of events in a
population over a given time period
= Number of events/total
population
Limitations of Crude Rates
Use crude rates with caution when comparing disease
frequencies between different populations
Observed differences in crude rates may be the results
of systematic factors (e.g., age distribution) within the
population rather than the true variation in rates
Specific rates
Age-specific rate =
Disease-specific rate =
Rates within a number of events of a
number of events in a
specific age group / total
specific specific type / total number in that age
subgroup of population at risk group
the population
or a specific Example: Number of
Example: Number of
deaths among 0-65
cause cancer caused deaths
year-olds / total number
/ total cancer patients
of 0-65 year-olds in
population
Age-standardized incidence and mortality rates
Odds ratio
• > 1: Risk
• = 1: No risk
• < 1: Protection
Observational Epidemiology
Many times we do not know the cause of disease, but we
can associate it with certain exposure.
Epidemiology based on observation data leads us to
understand the association between the development of a
disease and certain exposures, habitats, lifestyle choices.
Bradford Hill criteria
Observed Association Causation
Examples
Ice creams Sunburns
Ice cream consumption does NOT cause sunburns.
Sunburns does NOT cause ice cream consumption.
Correlation
Bradford Hill criteria
Observed Association Causation
Several things that should be true if an observed
association is causal.
9 ’criteria’ developed in 1965
• To evaluate human epidemiologic evidence to determine if causation
can be deduced: strength, consistency, specificity, temporality,
biological gradient, plausibility, coherence, experiment, and analogy
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Bradford Hill Temporality: The effect has to occur after the cause
criteria in Biological gradient: Greater exposure should generally lead to greater
epidemiological incidence
study Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
Strength of Association: the larger the association, the more likely that it is
causal
Consistency of Data: Consistent findings observed by different persons in
different places with different samples strengthens the likelihood of an effect
Specificity: Causation is likely if a particular exposure only leads to the
observed outcome but not many other outcomes
Which criteria Temporality: The effect has to occur after the cause
do you think are Biological gradient: Greater exposure should generally lead to greater
the most incidence
important? Plausibility: The proposed causal mechanism should be consistent with
existing knowledge
Coherence: The proposed causal relationship should be consistent with
existing knowledge
Analogy: similar causal relationships in other contexts are supportive
Experimental evidence: evidence from experimental studies that support the
causal relationship
3 very important 6 less important
• Strength of Association • Consistency of Data
• Temporality • Plausibility
• Biological Gradient • Specificity
• Coherence
• Analogy
• Experimental Evidence
• What is epidemiology?
Take home
messages • Terminology in epidemiology
• Bradford Hill criteria
Cancer Epidemiology
The Global Cancer
Burden in 2018
• The global cancer burden:
incidence - 18.1 million new cases
mortality - 9.6 million deaths
• One in 5 men and one in 6 women develop
cancer during their lifetime
• One in 8 men and one in 11 women die
from the disease
• Total number of people who are alive within
5 years of a cancer diagnosis is 43.8 million
The Global Cancer Burden in 2020
The Global Cancer Burden in 2020
The Global Cancer Burden in 2020
The Global Cancer Burden in 2020
World-wide cancer data
Between 2007 and 2017, the average annual age-standardized incidence rates
(ASIRs) for all cancers combined increased in 123 of 195 countries.
In contrast, the average annual age-standardized mortality rates for all cancers
combined decreased within that timeframe in 145 of 195 countries.
Top-ranked cancers by absolute mortality in males, 2020
Top-ranked cancer by absolute mortality in females, 2020
Breast cancer is the most common cancer in women
in almost all countries worldwide
Female breast cancer incidence and mortality rate, 2018
Breast cancer incidence rates decline or stabilize in high-incidence
countries, but increase in lower-incidence countries
Breast cancer mortality rates are declining in high-income countries
Cancer in China, 2020
[Link]
[Link]
Which cancers were the commonest in 2021?
Incident cases
Death cases
Incident
change in
males
Incident
change in
females
Mortality
changes
in males
Mortality
changes in
females
Cancer Epidemiology
Teaching
contents for
Healthcare Basics of Cancer Genetics
Analytics
-Population
Health Cancer Mutation Data
Informatics
Analysis
End of Today’s Lecture