Epi Final Exam

Front 1

Core Functions of Public Health

Back 1

Assessment: To regularly and systematically collect,
assemble, analyze, and make available
information on the health of the community,
including statistics on health status,
community health needs, and epidemiologic
and other studies of health problems

Policy Development: To serve the public interest in the development
of comprehensive public health policies by
promoting the use of the scientific knowledge
base in decision-making about public health
and by leading in the developing public health
policy

Assurance: To assure their constituents that services
necessary to achieve agreed upon goals are
provided, either by encouraging actions by
other entities (private or public), by requiring
such action through regulation, or by
providing services directly

Front 2

Primary Prevention

Back 2

Strategies, tactics, and procedures
that prevent the occurrence of
disease in the first place

Examples: safe drinking water,
vaccinations

Front 3

Secondary Prevention

Back 3

Strategies, tactics, and procedures,
including screening tests, that detect
disease as early as possible so that its
progress can be arrested and, if
possible, the disease eradicated

Examples: Pap test for cervical cancer,
HIV test

Front 4

Tertiary Prevention

Back 4

Strategies, tactics, and procedures,
including interventions, that aim to
arrest the progress of established
disease

Examples: chemotherapy for
colorectal cancer, stroke rehabilitation
program

Front 5

Defining a Population

Back 5

Person (Who is getting the disease?)
e.g., age, sex, race/ethnicity, job, SES

Place (Where is it occurring?)
e.g., geopolitical (country, state), natural
geographic features, environment (physical,
biologic, social)

Time (How is it changing over time?
e.g., calendar year, time in the life course, can
range from hours to decades

Front 6

Those who are not at risk include…

Back 6

– people who currently have the disease
– people who lack the organ defining the disease
– people who are immune to the disease

Front 7

Dynamic (i.e., open) population

Back 7

– membership is based on a condition and is transitory
– e.g., Population of the US in 2012

Dynamic population: You usually assume “steady
state” conditions apply for short intervals; the
number leaving are balanced by the number
entering, e.g., use midpoint estimate

Front 8

Fixed (i.e., closed) population

Back 8

– membership based on an event and is permanent
– e.g. Persons living in Chernobyl on April 25, 1986;
persons on a cruise ship during an outbreak of
norovirus

You take account of “persontime” (e.g., person-years) under observation

In studies of closed (i.e., fixed)
populations, it is often important to know
not only who was at risk, but also how
long each person was at risk.

Front 9

Seasonal Variation

Back 9

Many disease cycles occur annually and represent
seasonal variation in disease occurrence
Seasonal variation is a well-known characteristic
for many infectious diseases and is usually based
on…
- characteristics of the infectious agent itself,
- the life pattern of the vector or other animal hosts, and/or
- changes in the likelihood of person-to-person spread.

Front 10

Endemic

Back 10

The constant presence of a disease within a
given geographical area or population group

May also refer to the usual prevalence of a
given disease within such an area or group

Front 11

Epidemic

Back 11

The occurrence in a community or region of
cases of an illness, specific health-related
behavior, or other health-related events clearly
in excess of normal expectancy

Relative to usual frequency of the disease

Front 12

Pandemic

Back 12

An epidemic occurring worldwide, or
over a very wide area, crossing
international boundaries, and usually
affecting a large number of people

Front 13

Demographic Equation

Back 13

Natural change in population size =
Births - Deaths ± Migration

Front 14

Ages in the Epidemiologic Transition

Back 14

The Age of Pestilence and Famine:
Mortality is high and fluctuating, low and variable life expectancy (range 20-40 years)

The Age of Receding Pandemics: Mortality declines
with fewer epidemics, life expectancy increases (from
30 to 50 years), population growth is sustained and
begins to be exponential

The Age of Degenerative and Man-Made Diseases: Mortality continues to decline and approaches
stability at low level, chronic diseases replace
infectious diseases as the primary causes of death

Front 15

Types of Calculations to Measure
Morbidity & Mortality

Back 15

Rate:
How fast is disease occurring?
Events (e.g., cases)/ Population-Time

Proportion:
What fraction of the population is affected by disease?
Number Affected/ Total Population

Ratio: How do two quantities compare?
Estimate in Ratio population #1: Estimate in population #2
Note: Ratios are unitless (when the two estimates are in the same units)

Front 16

Incidence

Back 16

The number of new cases of a disease that occur
during a specified period of time in a population at
risk for developing the disease

Incidence =
Number of NEW cases of a disease occurring in the population during a specified period of time /
Number of persons who are AT RISK of developing the disease during that same period of time

Front 17

Cumulative Incidence

Back 17

Cumulative Incidence (%) =
Number of new cases
of a disease occurring
in the population during
a specified period of time /

Number of persons at risk
for the disease at the
beginning of the period

Note: All individuals are observed for entire period

Front 18

Incidence Rate

Back 18

Incidence expressed as a rate in units of
person-time “at risk”

Note: Some individuals are not observed for
entire period (i.e., different people are observed
for different lengths of time)

# new cases/Person-time at risk
example: 5/11.67 person-yr = 42.8 per 100 person-

Front 19

Prevalence

Back 19

The proportion (%) of the population affected by
the disease at a given time (i.e., point in time or
period in time); not a measure of risk

Prevalence=
(Number of existing cases of disease present in the population at a specified time/
Number of persons in the population at that specified time) x 100

Range for prevalence: [0,1] or [0%, 100%]

Prevalence = Incidence x Duration If there is a steady-state situation (i.e., rates are not changing and in-migration equals out-migration)

Front 20

Annual All-cause Mortality Rate
(per 1,000 population)

Back 20

(Total number of deaths from all causes in 1 year/
Number of persons in the population at mid-year) x1,000

Front 21

Cause-specific Mortality Rate
(per 1,000 population)

Back 21

Annual Mortality rate from “Disease A” per 1,000 population = Number of deaths from “Disease A” in 1 year / Number of persons in the population at mid-year) x 1,000

Front 22

Proportionate Mortality

Back 22

What proportion of all deaths in a given
year are due to a particular disease?

# deaths due to particular disease/ Total # deaths

Sometimes this is (incorrectly) called the
Proportionate Mortality Rate.

Front 23

Case-Fatality Rate (CFR) (%)

Back 23

# deaths from Disease Y/ # people with Disease Y

Note: Although usually called a “rate”, the
CFR is really a proportion. CFR is a
measure of the severity of a disease.

Front 24

When is a mortality rate a good index
of an incidence rate?

Back 24

When case-fatality rate is high (e.g., untreated rabies)

When the duration of the disease (i.e., survival) is short (e.g., pancreatic cancer)

Front 25

Two methods of adjustment

Back 25

Direct - weighted average of specific rates
• Uses rates of sub-groups of populations
being compared
• Need to know sub-group distributions of
events in both populations

Indirect - compares observed number of events to
expected number
-->Used when sub-group distribution of events
in population of interest is not known

Front 26

Direct age-adjustment
Steps

Back 26

• Step 1 – Calculate rates for each population
overall and by age category
– Divide the number of events by the population

• Step 2 - Identify the standard population
– Most often, add populations of interest by age
categories
– Can select appropriate population with age categories
(e.g., 2010 US population)

• Step 3 – Apply all rates to standard population
by age category to obtain expected number of
events for each population

• Step 4 – Sum the expected number of events for
each population

• Step 5 – For each population, divide the sum of
expected number of events by the total standard
population to obtain the age-adjusted rate

Front 27

Indirect age-adjustment Steps

Back 27

Step 1 – Identify total observed events in
population of interest
• Step 2 – Identify age-specific event rates for
known population of comparison
• Step 3 – Apply age-specific event rates (from
comparison population) to age-specific strata of
population of interest to obtain expected number
of deaths for each age category
• Step 4 – Sum the expected number of events for
the population of interest
• Step 5 – Divide the number of observed events
by the number of expected events to obtain the
standardized mortality ratio (SMR)

Front 28

Standardized mortality ratio
(SMR)

Back 28

Observed number of deaths/ Expected number of deaths

Note: Multiplication by 100 is often done to
yield results without decimals

Front 29

Measures of Prognosis

Back 29

5-year survival rate:
Proportion of persons in a specified group alive at the beginning of the 5 years who
survive to the end of the 5 years

Observed survival rates
– person-time
– life tables
– Kaplan-Meier method

Median survival time:
Length of time that half of the study population survives

Relative survival rate:
Observed Survival Rate/Expected Survival Rate

Front 30

Life Tables

Back 30

Two Assumptions Made In Using Life Tables:
1)No changes have occurred in survivorship over calendar time.
2)Those lost to follow-up experience the same survivorship as those who are followed.

Front 31

Kaplan-Meier Method

Back 31

Requires date last observed or date outcome occurred on each individual (end of study can be the last date observed) The essence of the Kaplan-Meier (KM) method is having the date each outcome in the cohort occurred.

CALCULATING SURVIVAL USING THE KAPLAN-MEIER METHOD
Columns:
(1)Times to deaths from starting treatment (months)

(2) Number alive at each time

(3) Number who died at each time

(4) Proportion who died at that time Column(3)/ Column(2)

(5) Proportion who survived at that time 1-Column(4)

(6) Cumulative proportion who survived to that time (Cumulative Survival)

Front 32

Median Survival Time

Back 32

Length of time that half of the study
population survives

Why median rather than mean?
– Less affected by extreme values
– Only need to observe deaths of half of
the study group rather than the entire
group

Front 33

Relative Survival Rate

Back 33

Observed Survival Rate/
Expected Survival Rate

Front 34

Investigating an Outbreak

Back 34

0. Determine existence of epidemic
1. Case definition
2. Identify cases and population at risk
3. Obtain information on time, place,
and person
4. Collect specimens for lab analysis
5. Analyze data
6. Develop, test, and refine hypotheses
7. Implement control measures
8. Prepare report and disseminate
findings

Front 35

Analyzing Data
in Outbreak Investigations

Back 35

1. Determine whether the observed number
of cases clearly exceeds the expected
number

2. Calculate attack rates
– Rate defined as the cumulative incidence of
infection in a group observed during an
epidemic
– Note, time is implicit

3. Epidemic curve - A graph of the frequency of
occurrence of cases by time of onset
Patterns:
Point source epidemic
Common source epidemic
Propagated epidemic

4. Spot map - A detailed map identifying the
location of cases in a discrete area

Front 36

Attack Rate

Back 36

Number of people at risk in whom a certain illness develops / Total number of people at risk

• AKA cumulative incidence
• Can be designated for a given exposure,
e.g., food-specific
• Time is implicit

Front 37

Point Source Outbreaks

Back 37

Population is exposed briefly /
simultaneously / at a single point in time

Cases occur suddenly, and, after a brief
peak period of time, fall off in a logarithmic
fashion

One incubation period

Front 38

Common Source Outbreak

Back 38

Exposure is not brief / simultaneous, rather it
occurs over time

Cases arise suddenly and continue to arise
as more individuals continue to be exposed
to the source

Peaks in epidemic curve may represent two
or more incubation periods

Front 39

Propagated Outbreak

Back 39

Exposure can be brief or over time

Person-to-person transmission of disease

Two or more peaks in epidemic curve as
secondary cases occur from person-toperson spread

Determinants of Propagated Outbreaks:
Balance between proportion of susceptible
and immune individuals in the population
-Immunity may be:
Innate (genetic) or
Acquired (previous illness, immunization)

Effective contact between agent and
susceptible host

Virulence of the organism

Front 40

Herd Immunity

Back 40

The relative resistance of a group to an
outbreak of a disease when a critical
proportion, but not all, of the group is
immune.

This occurs when the disease is spread by
person-to-person contact.

The greater the proportion of immune
people in the group, the less likely it is that
an infected person will encounter a
susceptible person.

If such encounters are sufficiently rare, the
disease will not propagate in the group.

The critical proportion of immune people
varies with the virulence of the disease,
e.g. for measles it is about 94%

Front 41

Risk Difference

Back 41

Risk Difference = (ARexposed) – (ARnon-exposed)

Front 42

Risk Ratio

Back 42

Risk Ratio = (ARexposed) / (ARnon-exposed)

Front 43

Median Incubation Period

Back 43

• 50th percentile
• observation at the middle of the distribution

Example: Time of symptom onset was known for 30 people (N=30)

Median = (30+1)/2 = 15.5 observation

Front 44

Passive Surveillance

Back 44

Routine reporting of disease cases seen in health care facilities

Passive surveillance often gathers disease data from all potential reporting health care workers. Health authorities do not stimulate reporting by reminding health care workers to report disease nor providing feedback to individual health workers.

Front 45

Active Surveillance

Back 45

Special search to find disease cases

An active surveillance system provides stimulus to health care workers in the form of individual feedback or other incentives. Often reporting frequency by individual health workers is monitored; health workers who consistently fail to report or complete the forms incorrectly are provided specific feedback to improve their performance. There may also be incentives provided for complete reporting.

Front 46

Sentinel Surveillance

Back 46

Disease-specific reporting systems in defined catchment areas

Instead of attempting to gather surveillance data from all health care workers, a sentinel surveillance system selects, either randomly or intentionally, a small group of health workers from whom to gather data. These health workers then receive greater attention from health authorities than would be possible with universal surveillance.

Front 47

Real-time Surveillance

Back 47

Alert public health care practitioners in early
phases of outbreak

Promptly institute case finding and control
measures

Improve access to treatment with the goal of
reducing morbidity and mortality

Front 48

Syndromic Surveillance

Back 48

Adapted CDC definition:
Surveillance using already existing health related data that precede diagnosis and signal a sufficient probability of a case or an outbreak to warrant further public health
response

Supplements existing surveillance methods

TWO SOURCES:
Syndromic Surveillance Clinical Data Sources:
Emergency Department (ED) data, e.g., chief
complaint, total patient volume
Emergency medical system call type, e.g., 911
Poison control center call
Unexplained deaths
Insurance claims or billing data
Clinical laboratory or radiology ordering volume

Alternative Data Sources:
School absenteeism
Work absenteeism
Over-the-counter medication sales
Health-care provider database searches
Internet-based health searches by public
Animal illness or deaths

Front 49

Analysis of Web Data

Back 49

Best used for surveillance of epidemics and
disease with high prevalence

Better suited for use in developed countries
with large populations of web search users

Pitfall - Disease publicity may increase use of
disease related search terms

Front 50

mHealth

Back 50

mHealth is the use of mobile hand-held devices, especially
text and web-enabled cell phones, for the delivery of
health information and messagingmHealth

Cell phone market penetration in the developed
world is around 90 percent, and in the developing
world about one-third of that, and growing

Android operating system is open-source

USES OF mHealth
Mobile phones can provide reminders about
healthy activities, sources for disease-specific
information

PROMISE of mHealth
Mobile phones are also being explored as a
means of…
- Tracking patient wellbeing or compliance
with treatments
- Data collection for public health purposes
- Warning of disasters or emergencies (e.g.,
tsunami, JHU text alerts)

CHALLENGES of mHealth
Evaluation is difficult
How to quantify effectiveness as part of an
intervention?
How to transition from pilot projects to scaledup programs?
Sustainability
No standards for mHealth platform integration

Front 51

Magpi

Back 51

Creates web-based data gathering forms that
can be filled out via cell phone from the field

Omits data collection on paper and input into
electronic databases for sharing and analysis

GPS enabled phones allow data to be tagged
with geo-coordinates

Front 52

Validity

Back 52

An expression of the
degree to which a
measurement
measures what it
purports to measure

Front 53

Sensitivity

Back 53

The measurement ability of the
test to identify correctly those
who have the disease

= # positive among those with disease/ total # with disease

= Pr ( + | Disease)

Front 54

Specificity

Back 54

The measurement ability of the
test to identify correctly those
who do not have the disease

# negative among those with no disease/ total # with no disease

= Pr ( - | No Disease)

Front 55

Two-Stage Sequential Screening

Back 55

People are identified/defined as having the
disease only if they test positive twice

Purpose of re-testing those who screened
positive is to have a second chance to rule out
the false positives

Ruling out more false positives increases
specificity

Undesirable consequence is that true positives may become false negatives; this decreases sensitivity

No need to re-test those who initially screened negative, as they won’t meet definition of disease

Front 56

Simultaneous Screening

Back 56

Purpose of re-testing everyone is to have a
second chance to rule out the false negatives
(i.e., identify more positives)

Ruling out more false negatives increases
sensitivity

Undesirable consequence is that true negatives
may be mistakenly labeled as false positives;
this decreases specificity

Those who tested positive the first time really do
not need to be re-tested as they already meet
the definition of disease, but this is done for
convenience with little added cost

Front 57

Sensitivity of test

Back 57

= # identified as positive twice / total # with disease

Front 58

Specificity of test

Back 58

= # identified as negative by either test / total #
with no disease

Front 59

Net sensitivity

Back 59

= sensitivity test 1 x sensitivity test 2

Front 60

Net specificity

Back 60

= specificity of test 1 + specificity of test 2 –
(specificity of test 1 x specificity of test 2)

Front 61

Comparison of Simultaneous
and Sequential Testing
Versus Using Either Test Alone

Back 61

Sequential testing
– “Positive” = positive result on both tests
– Net loss in sensitivity
– Net gain in specificity

Simultaneous testing
– “Positive” = positive result on either test
– Net gain in sensitivity
– Net loss in specificity

Front 62

Positive Predictive Value

Back 62

Of those who test positive, the proportion who actually
have the disease
Pr ( Disease | +)

= # tested positive among those with disease /
total # tested positive

synonyms: predictive value positive, PVP,
PPV, PV+

PPV is not a fixed characteristic of the test

Front 63

Negative Predictive Value

Back 63

Of those who test negative, the proportion of
patients who are actually free of the disease
Pr ( No Disease | -)

= # tested negative among those with no disease / total # tested negative

synonyms: predictive value negative,
PVN, NPV, PV

NPV is not a fixed characteristic of the test

Front 64

Reliability

Back 64

Can the result of the test be replicated
if the test is repeated?

i.e., consistency of results under repeated
examination (by the same people, or similarly
trained people, under the same/similar
conditions)

Synonyms: repeatability, precision

Front 65

Percent agreement

Back 65

Overall percent agreement (concordance)
for test with dichotomous outcomes:
Percent agreement = (a + d)/total x 100

Front 66

Kappa Statistic

Back 66

((Observed Agreement(%))-(Agreement Expected By Chance Alone (%))) / (100% - (Agreement Expected By Chance Alone (%)))

Front 67

Kappa Statistic Numerator:

Back 67

How much better is the observers’ agreement than by chance alone?

Front 68

Kappa Statistic Denominator:

Back 68

What is the maximum
the observers could possibly
improve agreement over
chance alone?

Front 69

Experimental Trial

Back 69

Studies prevention and treatment of disease

Investigator actively manipulates which groups receive the exposure/intervention

Front 70

Observational

Back 70

Studies causes, prevention and treatment for diseases

Investigator passively watches as nature takes its course

Front 71

Cohort

Back 71

Examines multiple health effects of an exposure

Subjects defined by exposure levels and follow for
disease occurrence

Front 72

Case-Control

Back 72

Typically examines multiple exposures in relation to a
disease

Subjects are defined as cases and controls and exposure
histories compared

Front 73

Cross-sectional

Back 73

Examines relationship between exposure and disease
prevalence in a defined population at one point in time

Front 74

Ecologic

Back 74

Examines relationship between exposure and disease
with population-level data rather than individual-level data

Front 75

Goals of analytic studies

Back 75

Determine risk factors and causes of disease

Evaluate preventive and therapeutic
interventions that alter the course of disease

Front 76

Populations in the context of study designs:

Back 76

Target – The population about which inferences
are desired
Source – The source of subjects for a particular
study, subset of target population, can be
enumerated
Study – Subset of source population, the
subjects who actually participate in study

Front 77

Unit of Observation

Back 77

1. Group (of individuals) as unit
– Examples: Lecture hall of students, city,
country
– Ecologic study
2. Individual as unit
– Examples: Person or event (e.g.,
pregnancy)

Front 78

Allocation of Exposure Under Study

Back 78

1. Experimental
– Performed by researcher
– May or may not be random
– Example: Clinical trial
2. Non-experimental (i.e., observational)
– By nature, self-selection, imposed by
others, etc
– Researcher does not want to interfere
with exposure allocation

Front 79

Timing of Observations

Back 79

1. Longitudinal recording of exposures and
outcomes over time (i.e., data collection at
least two time points)
2. Cross-sectional
- Information on exposures, outcomes,
other factors collected at the same time
(i.e., data collection at only one time point)
- Prevalence data
- No attempt to reconstruct exposure
history

Front 80

Sampling of Source Population

Back 80

1. By exposure status
– Exposed versus unexposed
– Cohort studies
2. By outcome status
– Disease (i.e., case) versus
no disease (i.e., control)
– Case control studies

Front 81

Type of Data

Back 81

1. Primary data – Collected for the purpose of
a study
– Example: Questionnaires
– Large studies may have various exposures
and outcomes
2. Secondary data – Collected primarily for
other purposes
– Example: Medical records
– Use with caution!

Front 82

Ecologic Study

Back 82

Association based on sample of
groups, rather than sample of
individuals

Distinguished by the fact that
groups are the “unit of
observation”

Front 83

Ecologic Studies

Back 83

Examine rates of disease in relation to a
population-level exposure factor
-Summaries of individual population
members, environmental measures, global
measures (e.g., health care system)

Populations/groups may be defined by
time or place or both

Measure of association is often the
correlation coefficient (r) - not an
estimate of risk

Front 84

Advantages in Ecologic Studies

Back 84

Can usually be completed relatively quickly if
based on available data

Inexpensive if data are available

Good for hypothesis generation

Can examine wide range of exposure levels

Front 85

Limitations in Ecologic Studies

Back 85

Individuals who are exposed may not be
the same as those with relevant outcome

Cannot adjust for other extraneous factors
(i.e., lack of information on important
variables)

Can be difficult to interpret, as data are
crude by nature

Cannot establish temporality or causation

Key Limitation: Ecologic Fallacy:
The bias that may occur because an
association observed between variables
on the aggregate/group level does not
necessarily represent the association that
exists at the individual level.

Therefore, we cannot ascribe group
characteristics to individual members of
the group.

Front 86

Cross-sectional Study

Back 86

Observational / non-experimental study

Individual is unit of observation

Data on exposures, outcomes, other
factors collected from defined population
at the same time, and at a single point in
time

Participants are often selected without
regard to exposure or outcome status

Advantages:
- Generate inferences and hypotheses
- Quick
- Low cost
- Highly generalizable, when based on sample of
general population

Often used as basis for public health policy
and programming decisions
Example: NHANES (National Health and
Nutrition Examination Survey)



Limitations of Cross-sectional studies
Cannot establish temporality

Cannot determine causation

May miss prevalent cases of short duration

Healthy worker effect

Front 87

Internal Validity

Back 87

Was the study well done?
Are the findings valid?

Need to consider…
- If there are major methodological problems
- If findings could be due to bias,
confounding, random error

**Important – Need to establish sound internal
validity before you consider generalizing the
results beyond the study population

Front 88

External Validity

Back 88

Aka “generalizability” to target population

To what extent are the participants you have
studied representative of all people with the
outcome of interest?

Need to examine…
- Who did not participate in the study
- Characteristics of study participants that might
preclude you from generalizing the study results
to others who were not in the study

Front 89

Clinical Trial

Back 89

Controlled study that prospectively evaluates
the effect of an allocated exposure (i.e.,
intervention) on the outcome of interest

Effects in which we’re interested: Safety,
efficacy, effectiveness

Considered “gold standard” of epi studies

Individual is unit of observation
Experimental design

Follow participants over time
-Collect data from at least two time
points (e.g., before exposure, after
exposure)

Clinical trials are justified when uncertainty
exists regarding the effectiveness of a
treatment (aka, EQUIPOISE)

EQUIPOISE: Legitimate uncertainty or
indecision as to choice or course of action…
because of an unknown balance of benefits
and risks

The researcher must believe that…
(1) what a study proposes to accomplish has
an excellent chance of being helpful (i.e.,
will contribute to generalized knowledge)
and
(2) he/she must have justified doubt about
the relative benefits of the comparison
treatment (which may be the “standard of
care” treatment)

When Clinical Trials Are Impossible
(or Nearly Impossible):
Adverse Exposures (e.g., cigarettes, other
toxins)

Rare Outcomes (e.g., Reye’s Syndrome)

Intervention Already in Wide Use (e.g.,
intensive care unit (ICU) medical care)

Front 90

Basic Protocol in a Clinical Trial

Back 90

1. Obtain approval of Institutional Review
Board (IRB)
2. Enroll participants
3. Gather “baseline” data from participants
4. Allocate exposure to participants
5. Follow-up participants to collect data on
outcome
6. Conduct data analyses
7. Report findings

Front 91

Randomized Assignment

Back 91

Unstratified by any variables
– Assignment is completely random
– Balanced in the long run, but may be
unbalanced in the short run

Stratified by key variables
–Ensures balance within subgroups defined by
key variables before randomization
–Stratification variable should be strongly
related to outcome (e.g., gender, risk level)

Front 92

Why Randomize Exposure Allocation?

Back 92

Ensure that exposure assignment is unbiased

Produce similar groups at baseline by known
and unknown factors
Goal: any difference between the groups at
the end of the study will be the result of the
exposure / treatment / intervention

Minimizes the threat of selection bias

Avoids confounding by indication

Front 93

Confounding by Indication in Observational Studies

Back 93

A bias when patients with the worst prognosis
are allocated preferentially to a particular
treatment.

High risk hypertensive patients are more likely to have
adverse outcomes.

High risk hypertensive patients are more likely to be
prescribed calcium channel blockers (than other drugs
hypertensive drugs).

Observational studies show that calcium channel
blockers are associated with more adverse outcomes

Front 94

Factorial Design

Back 94

Potentially economical way to test two
treatments simultaneously, if their modes of
action are independent

OR

Method to test for treatment synergy
- Is the effect of the combined treatment different
than expected based on the effects of the treatments
alone?

Front 95

“Cross Over"

Back 95

Crossing from one treatment group to the other

Unplanned crossover:
treatment non-adherence
procedures/protocol should be designed to minimize

Planned crossover design:
administration of treatments one after the other in random (or specified) order
treatment may be followed by a “washout” period

Planned “Cross Over”
-Each participant serves as his/her own control
- creates comparability between treatment groups

Feasible only if…
- Outcomes are recurrent, and
- No “carryover” treatment effect after “washout” period

Randomize order of treatments

Front 96

Who to Mask/Blind in the Study and Why

Back 96

Participants: Quantify placebo effects
Physicians: Uniform care apart from study
Data Collectors: Uniform outcome ascertainment
Data Analysts: Reduce threat of analytic bias

Front 97

Partial Masking

Back 97

In some circumstances masking of participants
and/or physicians may be impossible or unethical
(Surgery, behavior modification)

In this setting, others can generally still be
masked:
Data collectors
Adjudicators
Laboratory measurements
Data analysts

Front 98

Approaches to Non-compliance

Back 98

Run-in period / pilot study – randomize subjects
after a trial period assessing compliance

Monitor noncompliance:
- Interview patients, count pills
- Medication bottle devices
- Blood or urine tests
- Directly observed treatment

In the setting of non-compliance, the observed
effect will likely be smaller than the true effect

Front 99

Intention to Treat (ITT) Approach

Back 99

Analysis by assigned treatment regardless of
the observed course of treatment

Maintains initial balance from randomization

Highlights problems from adverse effects

Conservative approach

Strongly recommended as primary approach

Front 100

Number needed to treat (NNT)

Back 100

Number of patients who would need to be treated
to prevent one outcome

NNT = 1 / (outcome frequency in untreated group
– outcome frequency in treated group)

Small NNT is good

Estimates often presented with 95% confidence
intervals

Front 101

Number needed to harm (NNH)

Back 101

Number of patients who would need to be treated
to cause one patient to be harmed (by treatmentrelated adverse events or side effects)

NNH = 1 / (adverse event frequency in treated
group – adverse event frequency in untreated
group)

Large NNH is good

Estimates often presented with 95% confidence
intervals

Front 102

Safety and Stopping

Back 102

“Stopping rule”

A rule set before the start of the trial that
specifies a limit for the observed treatment
difference for the primary outcome which, if
exceeded, automatically leads to the termination
of the treatment or control arm (depending on
direction of the difference)

Front 103

When to stop a clinical trial before
its scheduled end?

Back 103

1. Clear evidence of benefit
2. Clear evidence of harm

--> Importance of plans to monitor the
progress of a trial

Front 104

Advantages of CTs

Back 104

"Gold standard” (Randomization) of epi studies

Designed to minimize bias

“Highest quality evidence available”

Results may be combined into systematic
reviews

Front 105

Why CTs Can Be Difficult

Back 105

Hard to find and recruit the right people

Great responsibility on the investigator(s), need
for tremendous documentation, cost

May take years for outcomes to develop

People are free to do as they please:
- Some assigned to treatment don’t adhere
- Some assigned to control seek treatment
- Some drop out of the trial completely (loss-tofollow-up)

Front 106

Limitations of CTs

Back 106

Cost

Limited external validity
- Country, patient characteristics, study
procedures, outcome measures

Time to conduct and to publish findings

Difficult to study rare events

Difficult to study distant events

Narrowing of the studied question

Front 107

Phases in Clinical Trials

Back 107

I Evaluate safety, dosage-->10-20 healthy volunteers -->Unexpected side effects may occur

II Evaluate efficacy --> About 200 patients-->Most drugs fail in Phase II due to being less efficacious than anticipated

III Evaluate effectiveness--> More than 1,000 patients-->Likelihood to detect rare side effects increases with number of patients

IV Evaluate long-term safety and effectiveness--> 1,000s of patients, “real life” evaluation outside of research environment-->Previously untested groups may show
adverse reactions, postmarketing surveillance

Front 108

Cohort Study

Back 108

Observational epidemiologic study that follows
groups with common characteristics over time

Terms associated with cohort studies: followup, incidence, longitudinal study

Participants defined by exposure status, then
followed for outcomes of interest

Front 109

Key Parameters of Cohort Studies

Back 109

Individual is unit of observation

Observational design

Follow participants over time
-Collect data from at least two time points

Participants selected based on exposure
status, and all are “at risk” for the main
outcome at baseline

Front 110

When is a Cohort Study Warranted?

Back 110

Good evidence of an association of the disease
with a certain exposure

Exposure is rare, but incidence of disease
among exposed is high

Time between exposure and disease is short

Attrition of study population can be minimized

Front 111

Timing of Cohort Studies

Back 111

Prospective – Looking forward in time
Participants grouped based on past or current
exposures and followed forward for outcome

Retrospective – Looking back in time
Both exposures and outcomes have already
occurred when study begins, data collection is
based on existing records (historical)

Ambidirectional – Looking both forward and
back in time

Front 112

Cohort Sources of Information

Back 112

Interviews

Medical and employment records

Direct physical exams

Lab tests and biological specimens

Environmental monitoring

And remember…
Each source has advantages and disadvantages
Need comparable procedures for data collection in
exposed and unexposed groups, including standard
outcome definitions and masking

Front 113

Losses to follow-up (LTF)

Back 113

Losses to follow-up (LTF) decrease sample size
LTF may be more like to develop outcome!

Collection of data at baseline on participant,
friends, relatives, physicians

Regular contact via mail, phone, home visits

If possible LTF – Then, “Address Correction
Requested,” contacts provided at baseline,
directories, national registries, commercial
companies

Front 114

Cohort Study Data Analysis Approach

Back 114

Primary objective – Compare disease
occurrence in exposed and unexposed groups
-->Incidence rates, cumulative incidence

Person-time

Induction period – Interval between action of a
cause (e.g., exposure) and disease onset

Latent period – Interval between disease onset
and clinical diagnosis

Front 115

Disadvantages of Cohort Studies

Back 115

Inefficient for rare outcomes

Poor info on exposures and other key variables
(retrospective)

Expensive and time consuming (particularly
prospective)

Inefficient for diseases with long induction and
latent periods (prospective)

More vulnerable to bias (retrospective)

Front 116

Advantages of Cohort Studies

Back 116

Efficient for rare exposures

Good information on exposures (prospective)

Can evaluate multiple effects of an exposure

Efficient for diseases with long induction and
latent periods (retrospective)

Less vulnerable to bias (prospective)

Can directly measure disease incidence or risk

Clear temporal relationship between exposure
and outcome (prospective)

Front 117

Case-Control Study

Back 117

Observational epidemiologic study of persons
with the outcome of interest (“cases”) and
without (“controls”) that examines the presence
of particular attributes (“exposures”) in the two
groups

Participants defined by outcome status, then
exposures of interest are assessed

Highly efficient study design

Front 118

Key Parameters of Case-Control Studies

Back 118

Individual is unit of observation

Observational design

No follow-up of participants over time
(i.e., the investigator does not directly
collect data from the participant over time)
- Collect data at one time point

Participants selected based on outcome
status, then exposures are assessed

Front 119

Types of Case-Control Studies

Back 119

Population-based:
-Participants identified from within a source population
-No pre-existing study infrastructure
-Example: Inpatients at Johns Hopkins Hospital today

Nested:
-Source population is ongoing cohort study
-Benefits of cohort and case-control study designs
-Example: Participants in ALIVE cohort study

Front 120

Selection of Cases
Incident versus prevalent cases(case control)

Back 120

Incidence, if studying causes of disease
Prevalence, if study duration of disease
Might not have a choice, so prevalence

Front 121

Ratio of Controls to Cases

Back 121

Can increase the statistical power of the study
to detect an association by increasing the size
of the control group

Up to a ratio of 4 controls : 1 case will increase
power

Beyond 4:1, not considered worthwhile due to
costs

Front 122

Challenge in Case-Control Studies

Back 122

Cases and controls may differ in characteristics or
exposures other than the one targeted for study
-Is study finding due to exposure, or due to differences
between cases and controls?

Solution via study design: Match cases and control
for factors about which you’re concerned

Matching: Process of selecting the controls so they
are similar to the cases in certain characteristics
(e.g., age, race, sex, etc)

Front 123

Matching (Case Control)

Back 123

Group matching (frequency matching)
Proportion of controls with a certain characteristic
is identical to the proportion of cases with the
same characteristic
# of controls may be less than # of cases

Individual matching (matched pairs)
For each case, at least one control is selected
who is similar to the case for the characteristic of
interest

Problems with Matching:
Practical –A lot of matching may make it
impossible to find a suitable control

Conceptual – Once you match controls to cases
by a certain characteristic, then you cannot
study that characteristic in your analysis

So, only match on factors you are convinced are
risk factors for the disease (and you therefore
don’t need to investigate)

Front 124

Case-Control Study
Data Analysis

Back 124

Challenge: Often, investigators do not know
the size of the total population that produced
the cases

Assumption: Cases and Controls Originate
From Same Hypothetical Source Population

So, we don’t know how many people were “at
risk” for becoming a case (i.e., we don’t know
the denominator), so we can’t calculate
incidence, prevalence, associated measures

But, we can calculate the odds!

Odds of event = probability (p) the event will
occur divided by the probability the event will
not occur = p / (1-p)

Front 125

Disadvantages of Case-Control Studies

Back 125

Inefficient for rare exposures

May have poor info on exposures because of
retrospective

Vulnerable to bias because of retrospective

Cannot establish temporal relationship between
exposure and disease

Front 126

Advantages of Case-Control Studies

Back 126

Efficiency
-Less time, less money than cohort studies,
experimental studies

Efficient for rare diseases

Efficient for disease with long induction and
latent periods

Can evaluate multiple exposures in relation to
outcome (so, good for diseases about which
little is known)

Front 127

When is the OR a good estimate of the RR? (CC study)

Back 127

When the cases are representative of all people
with the disease in the population from which the
cases were drawn, with regard to history of the
exposure.

When the controls are representative of all
people without the disease in the population from
which the cases were drawn, with regard to
history of exposure.

When the disease is not frequent (i.e., rare).

Front 128

Necessary Causes

Back 128

If a disease is defined by the presence of an
agent, that agent is necessary by definition.

Example: Tuberculosis can only be caused by
the tubercle bacillus.

Contrast: Hepatitis can be caused by many
viruses, but Hepatitis C is caused only by the
Hepatitis C virus.

Front 129

Any Given Cause May Be Necessary,
Sufficient, Both, or Neither

Back 129

Necessary and sufficient = cause is always
present with disease; nothing but cause is
needed to result in disease
–Example: measles virus and measles

Necessary and not sufficient = cause is
always present with disease, but disease is
not always present with cause
–Example: HPV and cervical cancer

Not necessary and sufficient = cause may or
may not be present with disease, nothing but
cause is needed to result in disease
–Example: High-dose exposure to pesticides
or ionizing radiation and sterility in men

Not necessary and not sufficient = cause may or
may not be present with disease; if cause is
present with disease, then some additional
factor must also be present
–Example: sedentary lifestyle and coronary
heart disease

Front 130

Necessary Conditions

Back 130

“X is a necessary condition for Y” =
If we don't have X, then we won't have Y
OR
Without X, you won't have Y
To say that X is a necessary condition for Y
does not mean that X guarantees Y.

Front 131

Sufficient Conditions

Back 131

“X is a sufficient condition for Y” =
if we have X, we know that Y must follow
OR
X guarantees Y

Front 132

Epidemiologic Guidelines for Establishing
a Cause-Effect Relationship

Back 132

Temporal sequence
Strength of the association
Dose response relationship / biologic gradient
Consistency of the association / replication
Coherence (biologic plausibility)
Specificity of the association
Experiment (cessation of exposure)
Analogy
Consideration of alternate explanations

Front 133

Temporal Sequence

Back 133

Study designs that can establish the potential
“cause” (risk factor or treatment) precedes the
disease include:
-Clinical trial
-Cohort study

Study designs that cannot establish that the
potential “cause” preceded the disease include:
-Cross-sectional study
-Population-based case-control study
-Ecologic study

Front 134

Probabilistic Causality

Back 134

The strength of a causal relationship is
assessed by the magnitude of its measures
of association.

The greater the RR or OR, the closer the
cause is to being necessary and/or
sufficient.

Front 135

Confidence Interval (CI)

Back 135

A computed interval that, upon repeated
sampling, has a given probability (e.g., 95%)
of containing the true value of a statistical
parameter (e.g., ratio, proportion, rate).



In other words…

For a 95% confidence interval, if a single
population is repeatedly sampled, then 95%
of the samples would capture the true value
of the population parameter.

Expresses the precision of the point estimate
- More narrow interval = more precision
- Less narrow interval = less precision
Calculated with predetermined significance
level, α (alpha), which is often set at 0.05

Front 136

Selection Bias

Back 136

Error due to systematic differences in
characteristics between those who take part
in a study and those who do not

The problem is that the association between
exposure and outcome may differ between
those who participate in the study and those
who do no

The measure of association is distorted due
to procedures used to select subjects and
from factors that influence study
participation

Usually inferred, rather than observed

Self-selection bias
• Selection of controls
– Healthy worker effect
• Post-entry exclusion bias

Front 137

Information Bias

Back 137

A flaw in collecting or measuring exposure
or outcome data that results in different
quality/accuracy of information between
comparison groups

Can result in distortion of the measure of
association

Information Bias Types
Misclassification
- Differential and non-differential with
respect to exposure and outcome
status
• Recall bias
• Reporting bias
• Interviewer bias
• Surveillance bias / biased follow-up

Front 138

Distinguishing between random error (i.e.,
chance) and systematic error (i.e., bias)

Back 138

Imagine that a given study could be
increased in size until it was infinitely large

Some errors would be reduced to zero; these
are the random errors

Other errors would not affected by increasing
the size of the study; these are systematic
errors or bias

Front 139

Confounding

Back 139

A situation in which the measure of association
is distorted because of the relationship
between the exposure and a third factor that
also influences the outcome.

It is a true phenomenon, and not an error in
the study.

Distortion in a measure of association due to a
third variable that:
1. Is associated with the exposure
2. Influences the outcome
3. Is not in the causal pathway (i.e., not an
intermediate step between exposure and
outcome)

Front 140

Confounding can be controlled for…

Back 140

1. In the study design
• Matching in a case-control study
• Randomization in a clinical trial
2. In the data analysis
• Stratification
• “Adjustment” (e.g., age adjustment)
• Multivariate regression models
BUT, we must have collected the data!

Front 141

Interaction (i.e., Effect Modification)

Back 141

If the size of the association between an
exposure and an outcome is changed or
modified by the level of a third variable,
interaction is said to be present
Interaction is also called “effect measure
modification”
Classic examples: age, immunization

Front 142

Model for Additive Effect

Back 142

Combined Total Risk of A and B =
Baseline Risk
+ Attributable Risk (A)
+ Attributable Risk (B)

Combined Effect of A and B =
Attributable Risk (A)
+ Attributable Risk (B)

Front 143

Model for Multiplicative Effect

Back 143

Combined Total Risk of A and B =
Baseline Risk
x Relative Risk (A)
x Relative Risk (B)

Combined Effect of A and B =
Relative Risk (A) x Relative Risk (B)

Front 144

Possible Types of Effect Modification

Back 144

Antagonism: Combined effect less than
predicted by the model (negative interaction)

Synergism: Combined effect greater than
predicted by the model (positive interaction)

Front 145

Comparison of
Confounding and Effect Modification

Back 145

Confounding – Association between exposure and
outcome is distorted by a third variable related to
the exposure and outcome

Effect modification – The association between
exposure and outcome is modified by levels of a
third variable

Front 146

Distinguishing between
confounding and effect modification

Back 146

1. Make list of potential confounders and effect
modifiers (literature review, data collected)
2. Calculate “crude” measure of association for
exposure and outcome of interest
3. Stratify association by levels of potential
confounder or potential effect modifier
4. Compare crude vs. stratum-specific
associations…

If stratified associations are relatively similar
across strata AND different from crude, then you
have confounding

If stratified associations differ across strata AND
crude association seems to be weighed-average of
stratum-specific associations (i.e., crude measure
is between stratum-specific measures), then you
have effect modification

Front 147

meta-analysis

Back 147

a way of combining data across research studies

Aim is to integrate the findings of separate
studies and to identify overall patterns in the
studies

Statistical analysis of results, examining sources
of differences in results among studies, leading
to quantitative summary of the results

Pooling of results from a set of studies may
increase statistical power

Quantitative and qualitative components

Front 148

Systematic review

Back 148

A summary of the literature

Also known as systematic research synthesis
Focuses on peer-reviewed publications about
specific health problem

Rigorous, standardized methods for selecting
and assessing articles

Does not include a quantitative summary of
the results across studies (unlike metaanalyses)

Front 149

Years of Potential Life Lost (YPLL)

Back 149

Measures premature mortality

Also known as potential years of life lost
(PYLL)

Estimate of the average years a person
would have lived if he/she had not died
prematurely

Alternative to mortality rate that gives more
weight to deaths that occur among younger
people

Front 150

Disability-adjusted life year (DALY)

Back 150

Measures disease burden

Expressed as a year lost due to ill-health,
disability or early death in a defined
population

Extends the concept of YPLL to include
equivalent years of “healthy” life lost by virtue
of being in states of poor health or disability

Essentially, mortality and morbidity are
combined into a single, common metric

Front 151

Quality adjusted life year (QALY)

Back 151

measures disease burden as well

Includes both the quality and the quantity of
life lived

Adjustment of life expectancy that takes into
account existence of chronic conditions
causing impairment, disability, and/or
handicap

Used for assessing the monetary cost value
of a medical intervention

Front 152

Cost-effectiveness analysis
focuses on economic costs of an
intervention to achieve desired outcomes

Back 152

Seeks to estimate the costs and effectiveness
of an activity (or between similar alternative
activities) to determine the degree to which it
(they) will obtain the desired outcomes.

The preferred action is one that requires the
least cost to produce a given level of
effectiveness.

Front 153

Cost-utility analysis
focuses on “utility-based” outcomes

Back 153

Economic evaluation in which outcomes of
alterative interventions are expressed in terms
of a single “utility-based” unit of measurement
(e.g., QALY).

Often used in health technology assessment.

Utility – The value of a particular health state,
usually expressed on a scale from 0 to 10.

Front 154

Kauffman Best Practices Project Final
Report Inclusion Criteria

Back 154

1. The treatment has a sound theoretical basis
in generally accepted psychological principles
indicating that it would be effective in treating at
least some problems known to be outcomes of
child abuse.
2. The treatment is generally accepted in
clinical practice as appropriate for use with
abused children, their parents, and/or their
families.
3. A substantial clinical-anecdotal literature
exists indicating the treatment’s value with
abused children, their parents, and/or their
families from a variety of cultural and ethnic
backgrounds.
4. There is no clinical or empirical evidence, or
theoretical basis indicating that the treatment
constitutes a substantial risk of harm to those
receiving it, compared to its likely benefits.
5. The treatment has at least one randomized,
controlled treatment outcome study indicating
its efficacy with abused children and/or their
families.
6. If multiple treatment outcome studies have
been conducted, the overall weight of evidence
supports the efficacy of the treatment.

Front 155

The National Registry of Evidence-based
Programs and Practices (NREPP)

Back 155

Registry of reviewed mental health and substance
abuse interventions
• General info on the intervention
• Description of research outcomes
• Ratings for quality of research and readiness
for dissemination
• List of studies and materials reviewed
• Contact info to obtain more information

Front 156

The Cochrane Collaboration prepares
systematic reviews to inform healthcare

Back 156

Prepares systematic reviews to help
healthcare providers, policy-makers, patients,
their advocates and carers, make wellinformed decisions about health care
• Prepares records of randomized trials across
the world

Front 157

“The Community Guide”

Back 157

a collection of reviewed intervention findings

Official collection of all Community Preventive
Services Task Force findings and the systematic
reviews on which they are based.

Resource with many uses because it is based
on a scientific systematic review process.

Front 158

Comparative Effectiveness
Research (CER)

Back 158

Conduct and synthesis of research comparing
the benefits and harms of different
interventions and strategies to prevent,
diagnose, treat, and monitor health conditions
in "real world" settings.

Purpose: To improve health outcomes by
developing and disseminating evidence-based
information to patients, clinicians, and other
decision-makers, responding to their
expressed needs, about which interventions
are most effective for which patients under
specific circumstances.

Front 159

CER involves several key elements

Back 159

Assess a comprehensive array of health-related
outcomes for diverse patient populations and
subgroups.

Interventions may include medications,
procedures, medical and assistive devices and
technologies, diagnostic testing, behavioral
change, and delivery system strategies.

Necessitates the development, expansion, and
use of a variety of data sources and methods.

Front 160

The three Eff’s

Back 160

• Efficacy: does the agent or intervention work
under ideal laboratory conditions?
• Effectiveness: does the agent work in real life
conditions?
• Efficiency: what is the cost-benefit ratio?