Lydia B. Zablotska, MD, PhD Associate Professor Department of Epidemiology and Biostatistics With...

Lydia B. Zablotska, MD, PhDAssociate ProfessorDepartment of Epidemiology and Biostatistics

With thanks to Erin Richman, ScD

Design features of cohort and case-control studies

Week 2 - Cohort and C-C Studies1

Highlights from Last Week

I. Theories of causal inference: i. Inductiveii. Deductiveiii. Other (refutationist, bayesian)

II. Causal models (heuristics): i. Rothman’s sufficient-component cause model (‘causal pies’)ii. Greenland and Morgenstern’s counterfactual model (‘potential

outcomes’)iii. Graphical models (DAGs)

III. Practice of causal inference:I. A. Bradford Hills causal guidelines (with caveats)II. Surgeon General’s causal criteria


Highlights from Last Week

Sufficient-component cause model Useful for examining multiple risk factors (‘web of causation’), but omits

discussion of origins of causes, focuses on proximal causes, and ignores induction period

Does not address indirect effects Causes of disease in individuals or groups of individuals, not in populations Does not consider factors that control distribution of risk factors Ignores dynamic non-linear relations

Counterfactual model Can be applied at individual or population level Specifies what would happen to individuals or populations under alternative

patterns of exposure Forces researchers to think about operational definitions of cases and

controls, sampling schemes, and other important design questions One of the two conditions in the definitions of the effect measures must be

contrary to fact – exposures or treatment vs. a reference conditionWeek 2 - Cohort and C-C Studies3

Learning Objectives

Classification of epidemiologic studies Cohort Studies

– Closed vs. open cohorts– Classifying person-time– Lag time

Case-control Studies– Selection of cases and controls– Measures of association– Control sampling schemes– Sources of controls

Field methods, i.e. nuts and bolds of conducting observational studies


Classifications by:

• Approach to data collection

• Goal

• Timing and directionality

• Unit of analysis

Epidemiological Methods


• Experimental

–RCTs, field trials, community intervention

and cluster randomized trials

• Quasi-experimental

–natural disaster studies

• Non-experimental or observational

–cohort, case-control, ecological

Classification by approach to data collection


• Descriptive

– ecological correlational studies, case reports,

case series, cross-sectional surveys

• Analytic

– observational studies and intervention studies

(RCTs)

Classification by goal


– Directionality: "Which did you observe first, the exposure or the disease?“

– forward (RCT, cohort) – backwards (case-control)

– Timing: “Has the information being studied already occurred before the study actually began?"

– retrospective and prospective cohort studies

Classification by timing and directionality


Classification by timing and directionality

Retrospective cohort study

past present future

x

RCT

exposed

unexposed

Diseased

Non-diseased

Diseased

Non-diseased

exposed

unexposed

Diseased

Non-diseased

Diseased

Non-diseased

9

– What is a unit?

Observations for which outcome and exposure are

measured

– Individual-level variables are properties of individuals

– ecological variables are properties of groups,

organizations or places

Classification by unit of analysis


Experimental study Cohort study Case‑control study

Each design represents a different way of harvesting information.

Selection of one over another depends on the particular research question, concerns of about data quality and efficiency, and practical and ethical considerations .

Main Epidemiologic Study Designs for Testing Hypotheses


Main Epidemiologic Study Designs for Testing Hypotheses, cont.

Randomized controlled trial = Investigator assigns exposure and follows participants overtime to ascertain the outcome

Cohort study = Analogous to RCT, but investigator does not assign exposure

Case-control study = Analogous to cohort study, but with more efficient sampling


Cohort Studies


1. Randomization of treatment so groups are comparable on known and unknown confounders.

2. Use placebo in order to reduce bias.

3. Blinding to avoid bias in outcome ascertainment.

Principles of experimental studies


1. Randomization of treatment so groups are comparable on known and unknown confounders.

Can't randomize in an observational study so select a comparison group as alike as possible to the exposed group

Principles of experimental studies applied to observational cohort studies


2. Use placebo in order to reduce bias.

Can’t use placebo in observational studies so you must make the groups as comparable as possible.



3. Blinding to avoid bias in outcome ascertainment.

In a cohort study, it is crucial to have high follow-up rates and comparable ascertainment of outcomes in the exposed and comparison groups.

You can blind the investigators conducting follow up and confirming the outcomes.



General definition of Cohort Studies

Investigator identifies a group of individuals who are free of the disease, but at risk for the disease

Classifies individuals into 2+ groups based on exposure

Follows over time to determine who develops disease


Closed Cohorts

Fixed population– After start of follow-up, no one can be added– Closely resemble RCT as follow-up starts at randomization

Membership-defining event– Cohort of persons born in 2012 (calendar time)– Cohort formed at entry into UCSF medical school (an event)

Can directly calculate:– Risk ratio– Incidence rate ratio– Odds ratio


Closed Cohorts, cont.

Limitations of incidence proportion (cumulative incidence) analyses:

– Loss to follow-up– Competing risks– For certain outcomes (e.g. death), incidence proportion

tends toward 1 over time– Exposures may change over time

Solution: Calculate Incidence Rates


Open Cohorts (Dynamic Cohorts)

No fixed roster – person-time accrued from a potentially changing roster of individuals

– Residents of California– Members of Kaiser Permanente

Individuals can contribute varying amounts of person-time

Individuals can leave and re-enter the cohort– Connecticut Tumor Registry

Individuals can contribute person-time to multiple exposure categories over time

– Occupationally exposed workers


Open Cohorts, cont.

Rates can be directly measured Incidence proportion cannot be directly measured Can calculate IP from IR assuming:

– The IR is constant over the follow-up periods (j)– No competing risks or loss to follow-up related to disease risk– Number of events in each time interval is small relative to

number of people at risk in that time interval

Use for communicability of results

Incidence Proportion = jj tIRe1


Classifying Person-Time

Each unit of person-time contributed by an individual has its own exposure classification

Must consider the etiologically relevant exposure Exposure may change over time

Exposure Disease Initiation

Disease Detection

Induction period Latent period


Classifying Person-Time, cont.

Time at which exposure occurs ≠ time at risk of exposure effects

– Radiation from an atomic bomb and risk of cancer

Only the time at risk for exposure effects should be counted in the denominator of the incidence rate for that level of exposure

If the induction time is not known, can estimate empirically by calculating the incidence rates for differing categories of time since exposure



How do you classify person-time contributed by exposed subjects before the minimum induction time has elapsed or after the maximum induction time has passed?

Example: – Exposure = Rotavirus vaccine– Outcome = Intussusception– Assume induction period ranges from 1-7 days

Exposure Disease Initiation

Induction periodWeek 2 - Cohort and C-C Studies25


Consider the time unexposed– BUT if the assumed induction time is incorrect, may

misclassify exposed time as unexposed = underestimate exposure effects

Omit person-time of exposed subjects when they are not at risk for the exposure effects

– Must be a large number of unexposed cases

Model exposure effects as depending on time since exposure


Classifying Exposure

Exposure may change over time– Measure exposure constantly and classify each unit of

person-time A given individual can contribute person-time to one or more

exposure category in the same study

– Assume one measure of exposure history is the only aspect of exposure associated with current risk

Current, average, cumulative, etc.


BASELINE QUESTIONNAIRE

Have you smoked 20 packs of cigarettes or more in your lifetime?

What specific brand and type?

If quit, how long ago?

At each age from <15 to 60+, what was the average number of cigarettes you smoked per day?

FOLLOW-UP QUESTIONNAIRES

Do you currently smoke cigarettes?

If so, how many per day?


Different exposure metric = different assumptions

Classifying Exposure, cont.

Week 2 - Cohort and C-C Studies

Smoking–Current v. Not

•Assume former = never–Never v. Former v. Current

•Assume intensity & duration are not important–Pack-years

•Assume ½ pack for 10 y = 1 pack for 5 y = 5 packs for 1 y

29

Classifying Exposure, cont.

Pack-years is a composite of duration and intensity of smoking

If the biologic effects of these differ, a study examining pack-years would not detect these differences, and could result in a misleading dose-response pattern

Ideally, present results for pack-years and individual components of the primary exposure metric (e.g. smoking intensity and duration)


Timing of Outcome Events

Events and the person-time being accumulated at the moment of the event are assigned to the same category of exposure

Must clearly define the outcome event, including a protocol for determining the timing of the event

Goal is to detect events at onset, however this may not be possible

– Date of diagnosis for cancer– Hospitalization for end-stage renal disease


Immortal Person-Time

Entry criteria in a cohort may depend on survival– Study of pre-term birth and mortality in adulthood –

Individuals had to survive at least 1 year.

– Study of occupationally exposed workers – Workers had to be employed for at least 1 year.

First year of life = immortal person-time During this time, no one is at risk to become an

event in the study Exclude immortal person-time from the

denominator of incidence rates


Lag exposure to account for induction time between exposure and disease initiation

– i.e. using Sufficient-component-cause model: time that it takes for the causal mechanism to be completed by the action of the complementary component causes

When lag period is not clear, need to hypothesize various induction times and reanalyze the data under each separate hypothesis. Also, can estimate the most appropriate time in the study by various statistical methods.

Lag Period (Induction Time)


1. Background stratified Poisson regression analysis of cohort data. Richardson DB, Langholz B. Radiat Environ Biophys. 2012 Mar;51(1):15-22. 2.

2. Lagging exposure information in cumulative exposure-response analyses. Richardson DB, Cole SR, Chu H, Langholz B. Am J Epidemiol. 2011 Dec 15;174(12):1416-22.

3. Hierarchical latency models for dose-time-response associations. Richardson DB, MacLehose RF, Langholz B, Cole SR. Am J Epidemiol. 2011 Mar 15;173(6):695-702.

Recommended References


Retrospective: both exposure and disease have occurred at start of study

Exposure------------------------Disease

*Study starts

Timing of cohort studies


Prospective: exposure has (probably) occurred, disease has not occurred

Exposure----------------------Disease *Study starts

Ambi-directional: elements of both



How do you choose between a retrospective and a prospective design?

Retrospective:

• Cheaper, faster• Efficient with diseases with long latent period• Exposure data may be inadequate



How do you choose between a retrospective vs. prospective design?

Prospective:

• More expensive, time consuming• Not efficient for diseases with long latent

periods • Better exposure and confounder data• Less vulnerable to bias



Selection of exposed population

Choice depends upon hypothesis under study and feasibility considerations

Issues in design of cohort studies


Examples of exposed populations:

• Occupational groups• Groups undergoing particular medical treatment• Groups with unusual dietary or life style factors• Professional groups (nurses, doctors)• Students or alumni of colleges• Geographically defined areas (e.g.

Framingham)



For rare exposures, you need to assemble special cohorts (occupational groups, groups with unusual diets etc.)

Example of special cohort study• Rubber workers in Akron, Ohio• Exposure: industrial solvent• Outcomes: cancer



If exposure is common, you may want to use a general cohort that will facilitate accurate and complete ascertainment of data (Doctors, nurses, well-defined communities)



• Framingham Study• Exposures: smoking, hypertension, family

history• Outcomes: heart disease, stroke, gout, etc.

Example of general cohort study


Selection of comparison (unexposed) group

Principle: You want the comparison (unexposed) group to be as similar as possible to the exposed group with respect to all other factors except the exposure. If the exposure has no effect on disease occurrence, then the rate of disease in the exposed and comparison groups will be the same.



Selection of comparison (unexposed) group (cont’d)

Counterfactual ideal: The ideal comparison group consists of exactly the same individuals in the exposed group had they not been exposed. Since it is impossible for the same person to be exposed and unexposed simultaneously, epidemiologists much select different sets of people who are as similar as possible.



Three possible sources of comparison group

1. Internal comparison: unexposed members of same cohort

• Ex: Framingham study



Three possible sources of comparison group

2. Comparison cohort: a cohort who is not exposed from another similar population

• Ex: Asbestos textile vs. cotton textile workers



3. General population data: Use pre-existing data from the general population as the basis for comparison. General population is commonly used in occupational studies. Usually find healthy worker effect

Ex. A study of asbestos and lung cancer with U.S. male population as the comparison group



Which of the three comparison groups is best?


Sources of exposure information:

* Pre-existing records - inexpensive, data recorded before disease occurrence but level of detail may be inadequate. Also, records may be missing, usually don't contain information on confounders



Sources of exposure information:

Questionnaires, interviews: good for information not routinely recorded but have potential for recall bias

Direct physical exams, tests, environmental monitoring may be needed to ascertain certain exposures.



Sources of outcome information:

Death certificates Physician, hospital, health plan records Questionnaires (verify by records) Medical exams



Goal is to obtain complete follow-up information on all subjects regardless of exposure status. You can use blinding (like an experimental study) to ensure that there is comparable ascertainment of the outcome in both groups.

Various indices describing study quality: tracing, recruitment, retention and loss to follow-up rates.



Approaches to follow-up

In any cohort study, the ascertainment of outcome data involves tracing or following all subjects from exposure into the future.



Approaches to follow-up:

Resources utilized to conduct follow-up: town lists, Polk directories, telephone books; birth, death, marriage records; driver's license lists, physician and hospital records; relatives, friends.

This is a time consuming process but high losses to

follow-up raise doubts about validity of study.



Efficient for rare exposures, diseases with long induction and latent period

Can evaluate multiple effects of an exposure

If prospective, good information on exposures, less vulnerable to bias, and clear temporal relationship between exposure and disease

Strengths of Cohort Studies


Inefficient for rare outcomes If retrospective, poor information on exposure

and other key variables, more vulnerable to bias

If prospective, expensive and time consuming, inefficient for diseases with long induction and latent period

Keep these strengths and weaknesses in mind for comparison with case-control studies

Weaknesses of Cohort Studies


Case-Control Studies


Case-Control Studies

Imagine a population in which a cohort study could be conducted

– Identify cases as you would in the cohort study– Sample from the study base (person-time) to determine

exposure distribution in the population that gave rise to the cases = Controls

– Controls must be sampled independent of exposure!

More efficient version of cohort study Sampling creates new opportunities for bias


This disparaging term was given to case-control studies because their logic seemed backwards (trohoc is ?? spelled backwards) and they seemed more prone to bias than other designs.

No basis for this derogation. Case-control studies are a logical extension of

cohort studies and an efficient way to learn about associations.

“TROHOC” STUDIES


A method of sampling a population in which cases

of disease are identified and enrolled, and a sample

of the population that produced the cases is

identified and enrolled. Exposures are determined

for individuals in each group.

General Definition of a Case-Control Study


Where do you get information on the numerators?

Where do you get the information for the denominators?

Case-Control Studies as Approximation of Cohort Studies


Where do you get information on the numerators? Cases = numerators of the rates of disease being compared you would have measured in exposed and unexposed groups in the corresponding cohort study

Where do you get the information for the denominators? If this were a cohort study, you would have the total population (if you were calculating cumulative incidence) or total person-years (if you were calculating incidence rates) for both the exposed and non exposed groups, which would provide the denominators for the compared rates. Controls = relative size of the exposed and unexposed person-time in the study base ≈ person-time denominators you would have measured in the corresponding cohort study

Case-Control Studies as Approximation of Cohort Studies


A case-control study can be considered a more efficient form of a cohort study.

Cases are the same as those that would be included in a cohort study.

Controls provide a fast and inexpensive means of obtaining the exposure experience in the population that gave rise to the cases.

Case-Control Studies as Approximation of Cohort Studies, cont.


Cases are the same people that would be cases in the underlying cohort study

– Can randomly sample cases if sampling is independent of exposure

Controls are a random, or conditionally random within strata, sample of study base

Exposure distribution in the controls is the same as in the population that gave rise to the cases, conditional on matching factors

Definitions of Cases and Controls


Selection of Controls

Controls should be selected from the same population – the source population – that gave rise to the cases.

Controls should be selected independently of exposure, within strata of factors that will be used for stratification in the analysis.

Persons are eligible to be selected as a control as long as they are at risk for disease a person can be both a control and a case in the same study!


Sources of Controls

Primary study base– Base population known

Study of Kaiser Permanente members Study conducted within existing cohort, “Nested case-control”

Secondary study base– Identify cases first, base population not known

Hospital-based case-control study

– Must identify person-time contributed by persons who would have become a case in your study had they developed the disease


Advantages of general population controls

Because of selection process, investigator is usually assured that they come from the same base population as the cases.

Disadvantages of general population controls

Time consuming, expensive, hard to contact and get cooperation; may remember exposures differently than cases

Selecting Controls


Neighborhood Controls

Sample residences May individually match cases to one or more

controls residing in the same neighborhood If neighborhood is associated with exposure,

must control for matching in the analysis Neighbors may not be the source population of

the cases– Cases at a VA hospital


Friend/Family Controls

Being named as a friend control may be related to exposure

– Reclusive people are less likely to be named

Investigator dependent on cases for identifying controls

Friend groups often overlap, so persons with more friends are more likely to be selected as a control


Random Digit Dialing

Case eligibility should include residence in a house with a telephone

Probability of calling a number ≠ probability of contacting an eligible control

– Households vary in the number of people, amount of time a person is at home, and the number of operating phones

Method requires a great deal of time and labor


Random Digit Dialing, cont.

Answering machines, voicemail, and caller ID reduce response rates

Cell phones reduce validity of assuming source population can be randomly sampled using this method

– Recent CDC survey showed 2% increase in binge drinking compared to 2009 data – more cell phone numbers included, and average age of respondents decreased

May not be able to distinguish business and residential numbers - difficult to estimate proportion of non-responders


Hospital-Based Controls

If not randomly selecting controls, must be cautious that control selection is independent of exposure

May not represent exposure distribution in source population if exposure is associated with hospitalization, other diseases, or both

– Example: Hospital-based study examining smoking and pancreatic cancer where controls are selected from persons admitted to the hospital for other conditions.


Hospital-Based Controls, cont.

Limit diagnoses for controls to conditions with no association with the exposure

– May exclude most potential controls– Exclusion criteria only applies to the cause of the current

hospitalization

Reasonable to exclude categories of potential controls on the suspicion that a given category might be related to exposure

Imprudent to use only a single diagnostic category as a source of controls


Exposure Disease Hospitalization

Hospital-Based Controls, cont.

Bias will occur if the exposure directly affects risk of being hospitalized, even if exposure is unrelated to the study disease or control diseases

Berkson’s Bias


Advantages of hospital controls Same selection factors that led cases to hospital led controls to

hospital

Easily identifiable and accessible (so less expensive than population-based controls)

Accuracy of exposure recall comparable to that of cases since controls are also sick

More willing to participate than population-based controls

Selecting Controls


Disadvantages of hospital controls

Since hospital based controls are ill, they may not accurately represent the exposure history in the population that produced the cases

Hospital catchment areas may be different for different diseases

Selecting Controls


Deceased Controls

Not members of the source population for the cases

If exposure is associated with mortality, dead controls will misrepresent exposure distribution in source population

Even if cases are dead, generally better to choose living controls

Do not need a proxy interview for living controls of dead cases


Comparability of Information

Comparability of information is often used to guide control selection and data collection

BUT– Non-differential exposure measurement error does not

guarantee that bias will be toward the null– Efforts to ensure equal accuracy of exposure data tend to

produce equal accuracy of data on other variables– Overall bias due to non-differential error in confounders and

effect modifiers can be larger than error produced by unequal accuracy of exposure data from cases and controls


Measures of Association

IRR

Tx

Tx

Tfx

Tfx

TTyx

TTyx

yx

yxOR

00

11

00

11

0000

1111

00

11

Person-time Data

Exposed Unexposed

Cases x1 x0

Person-time T1 T0

Case-control Data

Exposed Unexposed

Cases x1 x0

Controls y1 y0

• Can also estimate the risk ratio or incidence odds ratio from case-control studies

• The measure of association estimated by the OR depends on the control sampling scheme


Measures of Association

IRR

Tx

Tx

Tfx

Tfx

yx

yxOR

00

11

00

11

00

11

• Controls must be sampled independent of exposure: f1 = f0

• Generally, control sampling rate is not known, so cannot calculate incidence rates in exposed and unexposed

• Generally, rare disease assumption is NOT needed• As with cohort studies, the incidence odds ratio and rate

ratio are only good approximations of the risk ratio if the incidence proportion is less than 0.1


Control Sampling Schemes

Control Sampling Method Description Measure of effect estimated by the OR

Case-cohort Persons at risk of disease at baseline

Risk ratio*Rate ratio

Density sampling Proportional to person-time accumulated by persons at risk of disease during follow-up

Rate Ratio

Cumulative case-control Persons at risk of disease who are non-cases at the end of follow-up

Incidence Odds Ratio Risk Ratio*

* Only need rare disease assumption when estimating the risk ratio from the odds ratio.


Full Cohort Analysis

X

X

X

X

X

Time83

Control Sampling Schemes

Draw the sampling schemes on the open and closed cohort diagrams (X = case):

– Density Sampling– Incidence Density Sampling (“Risk Set Sampling”)– Case-Cohort Sampling– Cumulative Case Sampling


Density Sampling

X

X

X

X

X

Time85

Density Sampling

X

X

X

X

X

Time86

Density Sampling

Sample controls at a steady rate per unit time over period in which cases are sampled

Probability of being selected as a control is proportional to amount of time person spends at risk of disease in source population

Individual may be selected as a control while they are at risk for disease, and subsequently become a case

Incidence density sampling or “risk set sampling” is a form of density sampling in which you match cases and controls on time


Risk Set Sampling

X

X

X

X

X

Time88

Risk Set Sampling

X

X

X

X

X

Time89

Case-Cohort Sampling

X

X

X

X

X

Time90

Case-Cohort Sampling

X

X

X

X

X

Time91

Closed Cohort Analysis

X

X

X

X

X

Time92

Cumulative Sampling

X

X

X

X

X

Time93

Cumulative Sampling

X

X

X

X

X

Time94

Exposure Classification

Same principles as discussed for cohort studies Cases’ exposure should be classified as of the

time of diagnosis or disease onset, accounting for induction time hypotheses

Controls should be classified according to their exposure status at the time of selection, accounting for induction time hypotheses


Timing of Exposure Classification

Selection time does not necessarily refer to the time at which a control is first identified

– For hospital-based controls, selection time may be date of diagnosis for the disease that resulted in the current hospitalization

– Date of interview is often used if there is not an event analogous to the cases’ date of diagnosis

Interviewers should be blinded to case-control status whenever possible


Case-cohort Nested case-control Case-control studies without controls

– Traditional case series– Case-crossover – Case-specular

Variations in case-control study designs


1. Sample the population at risk at the start of the observation period

*-------------------------------------------------------------------------*Start FU End FU ^^

2. Sample population at risk as cases develop*-------------------------------------------------------------------------*Start FU End FU ^ ^ ^ ^^^ ^

3. Sample survivors at the end of the observation period*------------------------------------------------------------------------*Start FU End FU ^^

Sampling a cohort population for controls: nested case-control study


When exposure data are expensive or difficult to obtain- Ex: Pesticide and breast cancer study

When disease has long induction and latent period- Ex: Cancer, cardiovascular disease

When the disease is rare

– Ex: Studying risk factors for birth defects

When little is known about the disease– Ex. Early studies of AIDS

When underlying population is dynamic– Ex: Studying breast cancer on Cape Cod

When is it desirable to conduct a case-control study?


Summary

Causal models inform decisions made in the design phase of a study

Every decision made by the investigator comes with its own set of assumptions

Cohort study = Analogous to RCT, but investigator does not assign exposure

Case-control study = Analogous to cohort study, but with more efficient sampling


Comparisons between Case-control and Cohort study design

Characteristics Case-control Cohort studySelect subjects based on Disease status Exposure Status

Exposure good for common exposures Good for rare exposures

Cost-effectiveness Cheaper and less time consuming

Expensive and time consuming

Disease Frequency Good for rare diseases Good for common diseases

Establish temporal order Temporality generally not clear Temporality generally clear

Incidence calculation Can not calculate incidence/risk/rate

Can calculate incidence risk or rate depending on study design

Study more than one outcome

No Yes

Examine >1 exposure Yes Generally no

Inherent Study Selection problem

Difficult to ascertain appropriate control group

Not applicable since start with a source population

Subject to biases Susceptible to more biasesParticularly recall bias

Less subject to biases-except to loss to follow-up (Loss of subjects due to migration, lack of participation, withdrawal & death)

Directionality Prospective/retrospective Prospective/retrospective


In theory, it's possible to use each design to test a hypothesis

Example: Suppose you want to study the relationship between dietary Vitamin A and lung cancer….

Which study design to choose?


Subjects are chosen on the basis of exposure status and followed to assess the occurrence of disease

High Vitamin A consumption ‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑> lung cancer

or not

Low Vitamin A Consumption ‑‑‑‑‑‑‑‑‑‑‑‑‑‑> lung cancer or not

What are the advantages and disadvantages of this option?

Cohort Study Option


Special type of cohort study in which investigator assigns the exposure to individuals, preferably at random

Investigator assigns exposure to:

High Vit A consumption ‑‑‑‑‑‑‑‑‑‑‑‑‑-‑‑> lung cancer or not

Low Vit A consumption ‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑> lung cancer or not


Experimental Study Option


Cases with the disease and controls who generally do not have the disease are chosen and past exposure to a factor is determined

Prior Vitamin A consumption <‑‑‑‑------- lung cancer cases Prior Vitamin A consumption <‑‑‑‑‑‑‑‑‑‑ controls


Case‑Control Study Option


State of knowledge

Frequency of exposure and disease

Time, cost and other feasibility considerations

Each study design has unique and complementary advantages and disadvantages

In practice, choice of study design depends on:


1. H. Checkoway H, N. Pearce N, D. Kriebel. Selecting appropriate study designs to address specific research questions in occupational epidemiology. Occup Environ Med 2007 Sep;64(9):633-8.

2. J. H. Fowke. Issues in the design of molecular and genetic epidemiologic studies. J Prev Med Public Health 2009 Nov;42(6):343-8.

Recommended References


Efficient for rare diseases and diseases with long induction and latent period.

Can evaluate many risk factors for the same disease so good for diseases about which little is known

Strengths of case-control studies


Inefficient for rare exposures Vulnerable to bias because of retrospective

nature of study May have poor information on exposure

because retrospective Difficult to infer temporal relationship between

exposure and disease

How do these strengths and weaknesses compare to cohort studies?

Weaknesses of case-control studies


Field methods

o Operations Manuals

o Staff training and re-training, certification

o Quality control measures

o Assessment of progress


Lydia B. Zablotska, MD, PhD Associate Professor Department of Epidemiology and Biostatistics With...

Documents

Transcript of Lydia B. Zablotska, MD, PhD Associate Professor Department of Epidemiology and Biostatistics With...