1. the exam will be 3 hours long2. The exam includes 50 questions: include True/False, Matching type, Fill in the blanks and multiple choice questions.3. exam study guide will be very helpful4. the questions would be in the same style of previous exams and homeworkStudy Design I: Case Reports,
Case Series, Ecological and
Cross-sectional
Learning Objectives
»Reasons for conducting studies
»Definition, characteristics, and analysis of
Descriptive and Analytic studies
»Types of study designs
~ Case reports
~ Case series
~ Ecological
~ Cross sectional
Basic Question in Epidemiology
» To describe burden of disease or prevalence of risk factors, health
behaviors, or other characteristics of a population that influences risk of
disease
» To determine causes or risk factors for illness
» To determine relative effectiveness of interventions
Basic Question in Epidemiology
»Are exposure and disease
linked?
Exposure
Disease
Types of Primary Studies
»Descriptive studies
~describe occurrence of outcome
»Analytic studies
~describe association between
exposure and outcome
Study Types
Epidemiology Study Types
Experimental
Trials
Case Reports
Descriptive
Case Series
Observationa
l
Cross-sectional
Ecological
Analytical
Case Control
Prospective Cohort
Retrospective
Cohort
Case Reports
Time
»A descriptive study that documents the
experience of a single patient.
»Generally provide detailed documentation of a
unique medical occurrence.
»May lead to the generation of a new hypothesis.
»Traditionally, a common type of study published
in medical journals.
»Chief limitation: Sample size of 1.
»Effect measure: Count and percentage.
Case Reports
Time
» Generally report a new or unique finding
• e.g. previous undescribed disease
• e.g. unexpected link between diseases
• e.g. unexpected new therapeutic effect
• e.g. adverse events
Case Reports
Time
»Advantages
~ May lead to the generation of a new hypothesis.
~ Informative for very rare disease with few established risk factors
~ Can follow the case for disease progression
»Disadvantages
~
~
~
~
Cannot study cause and effect relationships
Cannot assess disease frequency
Limited to only a case
No control group
Case Series
Time
»A descriptive study for a collections of individual case reports.
»Often used as an early means to identify the beginning or
presence of an epidemic.
»Experience of a group of patients with a similar diagnosis
»Cases may be identified from a single or multiple sources
»Generally report on new/unique condition
»May be only realistic design for rare disorders
»Effect measure: Count and percentage.
Case Series
Time
»Advantages
~
~
~
~
May lead to the generation of a new hypothesis.
Informative for very rare disease with few established risk factors
Characterizes averages for disorder
Can follow cases for disease progression
»Disadvantages
~
~
~
~
Cannot study cause and effect relationships
Cannot assess disease frequency
Limited sample size
No control group
Ecological Studies
»Ecological studies are studies of risk-modifying factors on
health or other outcomes based on populations defined
either geographically or temporally.
» Compares outcomes between groups, not individuals
» Useful to examine trends over time or to explain differences
between groups
» Ecological studies often provide the first tentative evidence of an
association between a causal factor and disease
» Effect measure: Correlations (r)
Data Used in Ecological Studies
»Common data source for ecologic studies
»Describes disease patterns in entire geographic or
political populations
»Routinely collected information from birth and death
certificates; allow comparisons between countries over
time
»Comparison by age, race, sex, geographic areas and
time period
Ecological Studies
Advantages
~Ecological studies are simple to conduct
~Often using pre-existing data collected for other purposes
~Inexpensive
~Representative of large groups and large geographic areas
~Available over long periods of time
~Uniform coding rules
~Data can be used from populations with widely differing characteristics
Disadvantages
~Usually not possible to adjust for potential confounders
~Links between exposure and disease seen at the aggregate (population) level may
be spurious and not seen at the individual level (ecological fallacy)
~Suitable exposure and outcome data may not be available as the data is usually
pre-existing
The association between mean air pollution
levels and annual mortality rates in 50 US cities (1979-1983)
Pope et al., 1995
Cross-sectional Studies
»Analytical study conducted to measure exposure and
disease at a single point in time or over a short period of
time, like a photo “snap shot”.
»A design that surveys exposures and disease status at a
single point in time (a cross-section of the population).
»Effect measure: Correlations (r) and odds ratio (OR).
E+
ED+
D-
time
Study only exists at this point in time
Cross-sectional Studies
»Often used to study conditions that are relatively frequent
with long duration of expression (nonfatal, chronic
conditions)
»It measures prevalence, not incidence of disease
»Not suitable for studying rare or highly fatal diseases or a
disease with short duration of expression
Cross-sectional Studies
» Advantages
~
~
~
~
~
Simple
Feasible
Quick
Economic
Can adjust for confounders
» Disadvantages
~ No Follow-up
~ Cannot study cause and effect relationships
~ Temporal ambiguity (cannot determine whether the exposure preceded
outcome)
~ Possible measurement error
Cross-sectional Studies – Example 2
»Aortic Pulse Wave Velocity Is Associated With
Measures of Subclinical Target Organ Damage
Hierarchy of Study Designs
High
Cross sectional
Ecological
Case Series
Case Report
LOW
QUESTIONS? COMMENTS?
Clinical Trials
Learning Objectives
»Discuss clinical trials and compare and contrast them
with observational studies
»Discuss the design and types of clinical trials
»Cite strengths and challenges for assessing the results
of randomized trials
»Phases of clinical trials
»Reporting of clinical trials
»Distinguish systematic review from meta-analysis
»Steps to conduct systematic review and meta-analysis
»Limitations of meta-analysis
Study Types
Epidemiology Study Types
Experimental
Trials
Case Reports
Descriptive
Case Series
Cross-sectional
Observational
Ecological
Analytical
Case control
Prospective Cohort
Retrospective
Cohort
What is a Clinical Trial?
»Clinical trials have a forward logical direction: they start
with an exposure and then determine outcome.
»Similar to cohort studies, they have a prospective
temporal direction.
»However, unlike prospective cohort studies, they are
interventional studies.
»This means that the investigators assign subjects to the
exposure groups.
»This is the Gold Standard in testing causal hypotheses.
Overview of Controlled Trials
»Clinical trials could be controlled or non-controlled.
»Clinical trials could be randomized or non-randomized
(Quasi-experiments).
Trials
Non-controlled
(Quasi)
Non-randomized
(Quasi)
Controlled
Randomized
outcome
RANDOMIZATION
Intervention
no outcome
Study
population
outcome
Control
no outcome
Baseline
Future
Time
What is Randomization?
»A method based on chance alone by which study
participants are assigned to a treatment group.
»The critical element of randomization is the unpredictability
of the next assignment.
»Minimizes the differences among groups by equally
distributing people with particular characteristics among all
the trial arms.
RANDOMIZATION
Intervention
Study
population
Control
Why Randomize?
»Recall that confounding occurs when an independent
risk factor is unequally distributed between exposure
groups.
»Randomization helps to ensure that there is no
inequality in the distribution of extraneous disease
determinants.
»The study is called Randomized Clinical Trial (RCT)
Selection of Subjects
»Ensure that participants meet the criteria for the intervention
~ have a specific disease for a therapeutic trial
~ free of disease for a preventive trial etc.)
»Usually eligibility is also defined by age, gender, race, state of
health (absence of contraindications etc.)
»The narrower the eligibility criteria, the less generalizable will
be the results.
Data Collection on Subjects
»Treatment (Assigned and Received)
~ We need to know if the subject that is assigned to receive a treatment actually
received it or decided to switch to the alternative treatment.
»Prognostic Profile at Entry
~ We want to verify that randomization has provided reasonable similarity
between the two groups in terms of risk factors that affect the outcome.
»Outcome
~ The need for comparable measurements in all study groups is particularly true
for measurements of outcome.
Blinding in Controlled Trials
»Measurement bias in outcomes can compromise the
validity of trials.
»Trials must use blinding/masking to prevent raters, who
classify outcome status, from knowing treatment status.
»Instead of single blinded studies, which means that
participants are blinded, trials should be double blinded,
which means that both participants and the trial staff are
blinded. However, they are only feasible when
intervention is feasible. Ideally, the statisticians are also
blinded to make the study triple blinded.
Intent-to-Treat (ITT) Principle
»Unlike animal studies, investigator cannot
dictate what a participant should do in a clinical
trial.
»A participant may forget to take the pills,
receive dose reduction due to toxicity, drop out
from the study at any point or lost to follow-up.
Ethics of Clinical Trials
»Do no harm. Clinical trials must be reasonably safe to
participants and have a favorable risk-benefit ratio.
»Proper informed consent is essential.
»Rationale for randomization must be sufficiently
justified.
»Fee for recruitment can result in bias. Unreasonable
compensation to participants can also cause
problems.
»Equal opportunities to be recruited and treated, e.g.:
race and gender.
RCT – Advantages
»The “gold standard” of research designs.
»Convincing evidence of relationship between exposure and
effect.
»Blinding.
»Comparable groups (using randomization).
»Controlling for possible confounders (using randomization).
»Realistic data analysis (using ITT).
RCT – Disadvantages
»Large sample size.
»Long term follow-up (possible losses).
»Very expensive.
»Compliance.
»Possible ethical concerns.
~ it may be unethical, for example, to assign persons
to certain treatment or comparison groups.
Three Designs of Clinical Trials
»Simple, parallel design
»Cross-over design
Eligible
Randomised to
A or control
Eligible
Randomised to
A THEN B or
B THEN A
A
Control
A
B
B
A
A
»Factorial design
Eligible
Randomised to
Multiple
Treatments
Control
A + Control
None
Parallel Design
Eligible
Randomised to
A or control
A
Control
» RCT are often parallel design.
» A parallel design includes independent study groups and each
group receives a different treatment regimen or intervention.
» Parallel design is more useful for studying conditions which are
prone to change over time (pain, acute exacerbations of a
disease, remissions).
» Example:
~ In a study to evaluate the efficacy of beta blockers for
hypertension, 24 patients are randomized into two groups of
12 patients.
~ One group is then treated with a beta blocker and the other
treated with placebo.
Crossover Design
Eligible
Randomised to
A THEN B or
B THEN A
A
B
Washout
Period
B
A
» A crossover design includes independent study group receiving both
treatment types wherein the subject becomes control of himself/herself.
» Example:
~ In a study to evaluate the efficacy of beta blockers for hypertension
patients, 12 patients would be enrolled into the study and 6 patients
would be assigned to treatment with the beta blocker first, followed by
placebo treatment and the other 6 patients would receive the same
treatments in reverse order, all having a washout period in-between
treatments.
» It is more statistically sensitive and efficient, using fewer patients.
» Least biased and most conservative.
Crossover Design
Eligible
Randomised to
A THEN B or
B THEN A
A
B
B
A
»Disadvantages
~ Carryover (Washout Period): If a subject is changed from therapy A to
therapy B and observed under each therapy, the observations under
therapy B will be valid only if there is no residual carryover from therapy
A. There must be enough of a washout period to be sure none of
therapy A, or its effects, remains.
~ Period Effect, the order in which the therapies are given may elicit
psychological responses. Patients may react differently to the first
therapy given in a study as a result of the enthusiasm that is often
accorded a new study; this enthusiasm may diminish over time. We
therefore want to be sure that any differences observed are indeed due
to the agents being evaluated, and not to any effect of order.
A
Factorial Design
Eligible
Randomised to
Multiple
Treatments
Control
A + Control
None
» Used if the anticipated outcomes for the two drugs are different,
and their modes of action are independent, one can economically
use the same study population for testing both drugs.
» Answers the question of whether the effects of the two drugs are
additive or multiplicative?
Clinical Trial Phases
»Phase I: Clinical pharmacology and toxicity
»Phase II: Initial Assessment of Efficacy
»Phase III: Full-scale Evaluation of Treatment Efficacy
»Phase IV: Postmarking Surveillance
Phase IV: Postmarking Surveillance
Types of Trials
Window of Non-Inferiority Margin
Superiority
5.1
Inferiority
5.5
6.0
6.5
6.9
Hypothetical Control
Treatment Event Rate
» If the experimental treatment event rate is < 5.1, then the experimental treatment would be superior to the active control. » If the experimental treatment event rate falls between the 5.1 and 6.9 range, then the experimental treatment is noninferior to the active control. » If the experimental treatment event rate is > 6.9, then the experimental treatment is inferior to the active control.
Measures of Treatment Effect in RCT
» Any treatment involves tradeoffs
~ Weigh benefits against risks/costs
Benefit
$$
Harm
Example
R*
Free of Heart
Failure (HF)
Metoprolol
(N = 5,123)
HF
(N = 823)
Control
(N = 4,988)
HF
(N = 1,397)
* Randomization
Relative risk (RR) = Risk in Treatment group / Risk in Control group
= (823/5,123)/(1,397/4,988) = 0.57
Interpretation: Metoprolol patients are 0.57 times likely to get HF when compared to control)
Relative Risk Reduction (RRR) = (1- RR)*100%
= (1 – 0.57)*100% = 43%
Interpretation: Metoprolol patients are 43% less likely to get HF when compared to control)
Absolute Risk Reduction (ARR) = (Risk in Control group – risk in treatment group)*100%
Also known as Risk Difference = [(1,397/4,988) – (823/5,123)] * 100% = 12%
Interpretation: For every 100 patients, 12 metoprolol patients will not get HF when compared to control)
Numbers Needed to Treat (NNT) = (1/ ARR
= 1/0.12 = 8.3, rounded to 9
Interpretation: For every 9 patients who receive metoprolol, HF is prevented in one patient.
Example 2
Summary of Clinical Trials
Trials
Non-controlled
(Quasi)
Controlled
Randomized
Blinded
Single Blinding
Double Blinding
Triple Blinding
Non-randomized
(Quasi)
Non-blinded
Systematic Review and Meta-analysis
Systematic Review and Metaanalysis
» A Systematic Review is a review of a clearly formulated question
that uses systematic and explicit methods to identify, select and
critically appraise relevant research, and to collect and analyse data
from the studies that are included in the review.
» A meta-analysis is a statistical analysis that combines the results
of multiple scientific studies.
Why Perform Systematic Review and
Meta-analysis?
» Decisions about the utility of an intervention or the validity of a hypothesis
cannot be based on the results of a single study, because results typically
vary from one study to the next.
» Rather, a mechanism is needed to synthesize data across studies.
» Narrative reviews had been used for this purpose, but the narrative review
is largely subjective (different experts can come to different conclusions)
and becomes impossibly difficult when there are more than a few studies
involved.
» Meta-analysis, by contrast, applies objective formulas (much as one would
apply statistics to data within a single study), and can be used with any
number of studies.
Applied Systematic Review and
Meta-analysis?
»Pharmaceutical companies use meta-analysis to gain
approval for new drugs, with regulatory agencies
sometimes requiring a meta-analysis as part of the
approval process.
»Clinicians and applied researchers in medicine,
education, psychology, criminal justice, and a host of
other fields use meta-analysis to determine which
interventions work, and which ones work best.
Steps
1. Clear answerable question
2. Reproducible search strategy
3. Determine eligibility of studies
▪ Inclusion: which ones to keep
▪ Exclusion: which ones to throw out
4. Abstract data from the studies
5. Analyze data in the studies
6. Summary of the evidence
7. Interpretation
8. Conclusions, recommendations
9. Published protocol and review
Systematic Review
Meta-analysis
Systematic Review Steps
1. Be methodical: plan first
2. List of popular databases to search
1. Pubmed/Medline
2. Embase
3. Cochrane Review/Trials Register
3. Other strategies you may adopt
1. Hand search (go to the library…)
2. Personal references, and emails
web, eg. Google (http://scholar.google.com)
Systematic Review Steps
Results of Meta-analysis?
Meta-Analysis of Effects of Vitamin D Treatment on Withdrawals Due to Adverse Events
Meta-analysis Steps
➢ Choose appropriate outcomes:
Outcomes measured in meta-analysis
could be:
➢ Odds ratio, Relative risk or hazards
ratio for qualitative variables
➢ Correlation coefficient ( r ) or mean
difference for quantitative variables
➢ Weigh studies by quality/sample size
and calculate the weighted average
➢ It is crucial that these results are
interpreted in the context of the studies
used to calculate them
Limitations of Meta-analysis
➢Flaws can come from missing or under utilizing any steps in checklist
➢ Was search strategy comprehensive?
➢ Was publication bias assessed?
➢ Was quality of individual studies assessed using appropriate
checklist?
➢ Was combined treatment effect calculated appropriately?
➢Meta-analyses seek new knowledge from existing data
➢ Limited to what other researchers have done
➢ New techniques not evaluated
➢“Garbage in, garbage out”
➢ Quality of meta-analysis only as good as the quality of the previous
research
Example
For a patient with a painful sore throat, you wonder whether
corticosteroids will help with pain relief?
» You do a search and find several studies:
~ some suggest that steroids reduce pain; some do not
» What do you do?
~ Ask a consultant? Peer? Patient?
~ Ask research student to find all studies & select the
best?
» How do you know which study to believe?
Example
Example 2
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Formulate the Review Question
Is combination therapy superior to monotherapy in obesity?
Define inclusion and exclusion criteria
Population: BMI: ≥27 kg/m2
Outcomes: ≥3 month-weight loss outcomes in lbs or kg
Document combination treatment vs. monotherapy
e.g. phentermine/topiramate vs. phentermine or phentermine/topiramate vs. topiramate
RCTs or RCTs and quasi-experimental designs: RCTs
Published vs unpublished studies: Published studies
Language restriction: English language only
Exclusion: behavioral therapy/interventions as the other intervention
Develop search strategy using key words and locate studies
Select studies
Extract data
Assess study quality
Analyze and interpret results
Disseminate findings
Strategy for Literature Search
A.
1.
2.
3.
4.
5.
6.
B.
1.
2.
3.
4.
Search Terms
Use search terms “combination therapy vs. monotherapy, obesity, randomized controlled trials”
Use search terms “combination therapy versus monotherapy, obesity, randomized controlled trials”
Use search terms “combination therapy, monotherapy, obesity, randomized controlled trials
Use search terms “co-administration, monotherapy, obesity, randomized controlled trials
Use search terms “adjunctive therapy, monotherapy, obesity, randomized controlled trials.
Use the same search terms, except for “obesity.” Replace “obesity” with “weight-loss”
Databases: Using search terms 1-6, search for literature in:
Embase
Pubmed
Clinicaltrials.gov
Cochrane Controlled Clinical Trials Register Database
Inclusion Criteria
• Population BMI 27 kg/m2,
• Outcomes  3 month-weight loss in
percentage
• Type of intervention and control groups:
combination vs. monotherapy
Exclusion Criteria
• Behavioral therapy/interventions as the
other intervention
• Non-RCTs only
• Non-published, Non- English language
Study Flowchart
Author
Year
Percent Change
Jensterle et al.; LIRA3
2017
5.76
Jensterle et al.; LIRA3/MET2000
2017
3.51
Sari et al.; Orlistat
2004
5.28
Sari et al.; Orlistat plus Metformin
2004
6.17
Chirinos et al.; Simvastatin
2010
3.04
Chirinos et al.; Simvastatin/Ezetimibe
2010
3.13
Greenway et al.; Nal
2009
1.2
Greenway et al.; BUP
2009
2.7
Greenway et al.; NB16
2009
5.4
Greenway et al.; NB32
2009
5.4
Greenway et al.; NB48
2009
4.3
Kaya et al.; Sibutramne
2004
12.45
Kaya et al.; Orlistat
2004
10.23
Kaya et al.; Sibutramine/Orlistat
2004
13.41
Paniagua et al.; Orlistat
2016
5.41
Paniagua et al.; Resveratrol
2016
4.06
Paniagua et al.; Orlistat/Rserveratrol
2016
6.34
Smith et al.; LOR
2017
3.3
Smith et al.; LOR/Phen BID
2017
7.2
Smith et al.; LOR/Phen QID
2017
6.7
Hollander et al.; Canagliflozin
2017
1.3
Hollander et al.; Phen
2017
3.5
Hollander et al.; Canagliflozin/Phen
2017
6.9
Ravussin et al.; Pramlintide
2009
8.4
Ravussin et al.; Metereptin
2009
8.2
Ravussin et al.; Pramlintide/Metereptin
2009
12.7
Overall Mean Change
5.67
-2
0
2
4
6
8
10
12
Percent Mean Difference from Baseline
14
16
18
Hierarchy of Study Designs
High
Meta-analysis
Clinical trials
Prospective Cohort
Retrospective Cohort
Case control
Cross sectional
Ecological
Case Series
Case Report
Low
QUESTIONS? COMMENTS?
From Association to Causation:
Bias and Confounding
Learning Objectives
»Define random error and bias
»Discuss types of bias and how to control them
»Define confounding and how to control confounding
»Criteria for causal inference
Causality
» Main application of epidemiology is to identify etiologic (causal)
associations between exposure and outcome.
» If an association is observed, the first question asked must always be
Exposure
Outcome
From Association to Causation
The following conditions have been met:
1. The study has an adequate sample size
2. The study is free of bias
3. Adjustment for possible confounders has been done
If there is an association between exposure of interest and the disease
outcome, is the association causal?
From Association to Causation: Sources of Error
Causal
Effect?
Random
Error
Bias
Confounding
RRcausal
RRassociation
The Role of Random Error
» Also known as chance error
» Results from variability in the data, sampling
» Characteristics of subjects in a sample may vary
from sample to sample. As a result, an
association between an exposure and outcome,
or lack thereof, may be the result of chance.
» Sample size is directly related to chance
» To minimize chance, increase the sample size
» If the results are statistically significant, it is an
indication that there is most likely no random
error
The Role of Random Error – Example
Bias
» “Any systematic error in the design, conduct, or analysis of a study that
results in a mistaken estimate of an exposure’s effect on the risk of
disease”
The Role of Bias
» The deviation of the results (Bias) from the
truth can explain an observed association
between exposure and outcome variables
Bias
Selection
Bias
Information
Bias
Confounding
The Role of Bias – Selection Bias
» This is a bias due to differences in characteristics between those
who take part in a study and those who do not
~ Example: In a study of alcoholism and risk of pneumonia from
hospitalized patients, alcoholics with pneumonia are more likely
to be admitted than non-alcoholics with pneumonia.
» Controlling selection bias
~ Define criteria of selection of diseased and non-diseased
participants independent of exposures in a case-control study
~ Define criteria of selection of exposed and non-exposed
participants independent of disease outcomes in a cohort study
The Role of Bias – Information Bias
» Information bias (Misclassification bias) occurs when information is collected
differently between two groups, leading to an error in the conclusion of the
association
~ Interviewer bias (Observer bias):
• Interviewer may probe differently about exposures in the past if he or she
knows the subjects as cases
• Controlling interviewer bias: Proper training of the those that will conduct
the interview to standardize the interviewing process. Make sure
interviewers and study personnel are unaware of exposure/disease
status.
~ Recall/Report bias:
• Subjects may recall/report past exposure better or in more detail if he or
she has the disease.
• Controlling recall/report bias: Surrogates, such as relatives and friends to
provide exposure information.
The Role of Confounding
»Occurs when an extrinsic factor is associated with a disease
outcome and, independent of that association, is also
associated with the exposure
»An extraneous factor that wholly or partially accounts for the
observed effect of risk factor on disease.
»Adjustment for the confounder provides an undistorted estimate
of the relationship between the independent and dependent
variables.
Exposure
Outcome
Confounder
Confounding – Example
Understanding Confounders
A factor is a confounder if 3 criteria are met:
1. A confounder must be causally or non-causally
associated with the exposure in the source population
(study base) being studied;
C
E
2. A confounder must be a causal risk factor (or a C
surrogate measure of a cause) for the disease in the
unexposed cohort; and
3. A confounder must not be an intermediate cause
(not an intermediate step in the causal pathway
between the exposure and the disease)
E
C
X
D
D
Understanding Confounding: Conceptual
Example
When evaluating the association between total number of births and breast
cancer risk, should maternal age at first birth be controlled?
1. Age at first birth (POT. CONF) is associated with total number of births
(EXP)
2. Age at first birth (POT. CONF) is a known risk-factor for BrCa (DIS).
3. Age at first birth (POT. CONF) is not a result of total number of births
(EXP).
Therefore, we need to control for age at first birth (POT. CONF).
Controlling for Confounding
»At the study design phase
~ Randomization
• assures equal distribution of confounders between study and control groups
~ Restriction
• subjects are restricted by the levels of a known confounder
~ Matching
• potential confounding factors are kept equal between the study groups
Impact of Randomization
Controlling for Confounding
»At the analysis phase
~ Stratified analysis
• analysis for various levels of potential confounders
~ Multivariable regression
• analysis of multiple variables as part of the statistical model
Exercise
Heart Disease
Gender
Gender Stratified Analyses of Suicide
Unadjusted and Adjusted Findings for
COPD
Summary
» Biases reflect inadequacies in the design or conduct of a study and clearly
affect the validity of the findings.
» Biases therefore need to be assessed and, if possible, eliminated.
» Confounding on the other hand, describe the reality of the interrelationships
between certain factors and a certain outcome.
Hill’s Criteria for Causal Inference
The first complete statement of the epidemiologic criteria of a
causality is attributed to Austin Hill (1897 – 1991). They are:
1. Strength of association
2. Consistency
3. Dose-response relationship
4. Temporality
5. Biological plausibility
6. Coherence
7. Analogy
8. Experiment
Hill’s Criteria – Strength of Association
» Before even considering causality, it is first necessary to establish that there is a
valid association.
» However, one should also consider the strength of the association.
~ Strong associations (e.g., risk ratio of 20 for heavy smoking and lung cancer) are
unlikely to be entirely explained by bias or confounding.
» While strength of the association should be considered, it is not a requirement, since
weak associations can also be causal.
» Effect measure (OR, RR): away from unity (the higher, the stronger the association).
» P-value (at 95% confidence level): less than 0.05 (the smaller, the stronger the
association).
Hill’s Criteria – Consistency
» Associations are more likely to be causal if they are observed
repeatedly by different investigators, in different populations, and
with different study designs.
» While replication increases one’s confidence that the relationship is
causal, it is not a requirement for a judgment of causality.
» Meta-analysis is an good method for testing consistency.
Hill’s Criteria – Dose-response
» Dose-response means that there is a dose-response relationship
between the cause and the outcome, i.e. the probability of the
outcome increases as the exposure level increases.
» For example, the risk of lung cancer increases as the number of
cigarettes smoked per day increases. This criterion is not a
necessary condition for a judgment of causality, because some
causal relationships have threshold doses, or exhibit other nonlinear relationships to risk of the outcome.
Hill’s Criteria – Temporality
» In order for a causal factor to result in an outcome, it must precede
the occurrence of the outcome in time.
» This is the only criterion that is necessary for a judgment of
causality.
» Prospective cohort studies and prospective clinical trials provide
stronger evidence of temporality than retrospective studies or crosssectional studies.
Hill’s Criteria – Biological Plausibility
» This criterion is met if there is a known biological explanation or a
plausible explanation for how the exposure of interest might result
in or contribute to the outcome of interest.
» For example, we now know that there are many carcinogens in
tobacco smoke, so it is certainly plausible that inhalation of tobacco
smoke might cause lung cancer.
» Moreover, carcinogens and free radicals in tobacco smoke can be
absorbed from the lungs and enter the blood stream, so it is
plausible that tobacco smoke might cause other adverse outcomes
such as heart disease or other cancers.
Hill’s Criteria – Coherence
» Coherence refers to the fact the causal relationship does not conflict
with other facts regarding the disease.
» Theoretical: compatible with pre-existing theory.
» Factual: compatible with pre-existing knowledge.
» Biological: compatible with current biological knowledge from other
species or other levels of organization.
» Statistical: compatible with a reasonable statistical model (e.g. doseresponse).
Hill’s Criteria – Analogy
» Analogy in this setting means that there are similar cause-effect
relationships.
» For example, if it is widely accepted that certain drugs taken during
pregnancy can cause birth defects, then it is easier to accept the
possibility that a new drug might also cause birth defects.
Hill’s Criteria – Experiment
» This means that interventions (treatments or risk factor
modifications) have predictable effects on the occurrence of
disease.
» For example,
~ getting smokers to quit smoking will reduce their risk of getting a
variety of smoking-related diseases.
~ getting an obese person to exercise and lose weight will reduce
their risk of type II diabetes.
Use of Guidelines
» Causal association is mostly a judgmental process based on available
information
» Epidemiologic process is continual
~ New evidence may be discovered that supports or refutes the current
understanding of the relationship between exposure and disease
Criteria Used by the U.S. Preventive Services Task
Force for Grading the Quality of the Overall Evidence
GOOD
FAIR
POOR
Evidence includes consistent results from well-designed,
well-conducted studies of representative populations that
directly assess effects on health outcomes.
Evidence is sufficient to determine effects on health
outcomes, but the strength of the evidence is limited by the
number, quality, or consistency of the individual studies,
generalizability to routine practice, or indirect nature of the
evidence on health outcomes.
Evidence is insufficient to assess the effects on health
outcomes because of limited number or power of studies,
important flaws in their design or conduct, gaps in the chain
of evidence, or lack of information on important health
outcomes.
Summary
» Although causal guidelines discussed in this chapter are often referred to
as criteria, this term does not seem entirely appropriate.
» Although it may be a desirable goal to place causal inferences on a firm
quantitative and structural foundation, at present we generally do not have
all the information needed for doing so.
» The preceding list should therefore be considered to be only guidelines that
can be of most value when coupled with reasoned judgment about the
entire body of available evidence, in making decisions about causation.
QUESTIONS? COMMENTS?

Purchase answer to see full
attachment




Why Choose Us

  • 100% non-plagiarized Papers
  • 24/7 /365 Service Available
  • Affordable Prices
  • Any Paper, Urgency, and Subject
  • Will complete your papers in 6 hours
  • On-time Delivery
  • Money-back and Privacy guarantees
  • Unlimited Amendments upon request
  • Satisfaction guarantee

How it Works

  • Click on the “Place Order” tab at the top menu or “Order Now” icon at the bottom and a new page will appear with an order form to be filled.
  • Fill in your paper’s requirements in the "PAPER DETAILS" section.
  • Fill in your paper’s academic level, deadline, and the required number of pages from the drop-down menus.
  • Click “CREATE ACCOUNT & SIGN IN” to enter your registration details and get an account with us for record-keeping and then, click on “PROCEED TO CHECKOUT” at the bottom of the page.
  • From there, the payment sections will show, follow the guided payment process and your order will be available for our writing team to work on it.