64
CASE CONTROL STUDY Seminar Presentation by: Dr. Timiresh Kumar Das Moderator: Dr. D. K. Raut, Director Professor, Dept. of Community Medicine, VMMC & Safdarjung Hospital

Case control study

Embed Size (px)

DESCRIPTION

Epidemiological studies and details of case control study

Citation preview

Page 1: Case control study

CASE CONTROL STUDY

Seminar Presentation by: Dr. Timiresh Kumar Das

Moderator: Dr. D. K. Raut, Director Professor,Dept. of Community Medicine, VMMC & Safdarjung Hospital

Page 2: Case control study

Outline of Presentation Epidemiological study cycle Analytical studies: Types

Case control vs Cohort Case control study

DefinitionsHistoryDesignOutcomesLimitationsAdvantages and ApplicationsNested case control studiesSelected examples of case control studies

Page 3: Case control study

Epidemiological Study Cycle The sequence of events starting with

description of disease or health related event in relation to time, place, personsearching for and finding differences in occurrence in different populationsformulating hypotheses regarding possible causative factors and testing themanalysing the resultsresults may lead to further descriptive studies or new hypotheses.

Page 4: Case control study

Epidemiological Study Cycle

DESCRIPTIVE STUDY

• Ca Lung increasing mostly smokers

• Death rates higher in populations with higher per capita cigarette consumption

Ochsner, 1939

CASE CONTROL STUDY

• Ca Lung patients and non patients

Clarifies if it was smokers who contributed to high Ca Lung

Doll, 1947-52

COHORT STUDY

• Follows a cohort of smokers and non smokers without Ca Lung

•Smokers develop Ca Lung more frequently

INTERVENTIONAL TRIAL (RCT)

•Proves hypothesis conclusively•Gives inputs regarding other factors, control measures.

Hypothesis: Smoking causes Ca

Lung

Hill, 1951-61

Page 5: Case control study

Analytical Studies Observational

Case control (Retrospective) studiesCohort (Prospective) studies

Experimental (Interventional):Animal experimentsHuman studies

• Therapeutic trials• Preventive trials

Difference in study groups is ONLY observed

& analysed, NOT created

experimentally

Difference in study groups is

CREATED EXPERIMENTAL

LY and outcomes observed

Page 6: Case control study

Observational Studies Purpose: To produce a valid estimate of

a hypothesised cause-effect relationship between suspected risk factor and disease.

Case Control Study Cohort Study

Starts with diseased (cases)& not diseased (controls)

Starts with not diseased but exposed & not exposed

Determine if 2 groups differ in exposure to specific factor or factors

Followed up to determine difference in rates at which disease develops in relation to exposure

Called as case control study due to the way in which study group is assembled

Called so because of the use of a “cohort” (a group of people who share a common characteristic or experience)

Page 7: Case control study

Retrospective (Case-Control)

a b

dc

DISEASEpresent absentEXPOSURE

present

absent

cases

controls

Total

Total

Pro

sp

ecti

ve

(C

oh

ort

)

exposed

Not exposed

A fourfold table

Mausner, 1985

Page 8: Case control study

Case Control Studies Cohort Studies

Proceeds from effect to cause Proceeds from cause to effect

Starts with the disease Starts with people exposed to the risk factor or suspected cause

Tests whether the suspected cause occurs more frequently in those with disease than those without disease

Tests whether disease occurs more frequently in those exposed than in those not exposed

Usually the 1st approach to the testing of hypothesis, but also useful for exploratory studies

Reserved for the testing of precisely formulated hypothesis

Involves fewer study subjects Involves larger number of subjects

Yields results relatively quickly Long follow-up, delayed results

Suitable for study of rare diseases Inappropriate when disease or exposure under investigation is rare

Generally, yields only estimate of relative risk (Odds ratio)

Yields incidence rates, relative risk, attributable risk

Cannot yield information about disease other than that under study

Can give information about more than one disease outcome

Relatively inexpensive Expensive

Page 9: Case control study

Definitions: Case control study synonyms:

Case comparison study Case compeer studyCase history studyCase referent studyRetrospective study

Case control study definitions:The observational epidemiologic study of

persons with the disease (or other outcome variable) of interest and a suitable control (comparison/ reference) group of persons without the disease. (Dictionary of Epidemiology: 3rd ed; John M Last. 2000)

Page 10: Case control study

Case control study definitions: A study that compares two groups of people:

those with the disease or condition under study (cases) and a very similar group of people who do not have the disease or condition (controls). (National Institute of Health, USA)

A case control study involves two populations – cases and controls and has three distinct features : Both exposure and outcome have occurred before

the start of the study. The study proceeds backwards from effect to cause. It uses a control or comparison group to support or

refute an inference.(Park’s Textbook of Preventive and Social Medicine – 20th ed; K. Park.

2009)

Definitions:

Page 11: Case control study

Case : A person in the population or study group identified as having the particular disease, health disorder or condition under investigation. (Dictionary of Epidemiology: 3rd ed; John M Last. 2000)

Control: Person or persons in a comparison group that differs, in disease experience (or other health related outcome) in not having the outcome being studied. (Dictionary of Epidemiology: 3rd ed; John M Last. 2000)

Definitions:

Page 12: Case control study

Bias: Any systematic error in the design, conduct, or analysis of a study that results in mistaken estimates of the effect of the exposure on disease.

Confounding: When a measure of the effect of an exposure on risk is distorted because of the association of exposure with other factors that influence the outcome. It creates data where it is not possible to separate the contribution that any single causal factor has made an effect.

Definitions:

Page 13: Case control study

History of case control studies

LANE CLAYPON’S BREAST CANCER STUDY1926

LUNG CANCER AND SMOKING1950

Early beginnings Establishment and acceptance

PCA Louis (1788-1875) - numerical method; William Augustus Guy (1843);Baker (1862) – case control comparisons of marriage and fertility in breast cancer

Page 14: Case control study

Six essential elements which developed separately over time in medical hiatoryIdea of the caseInterest in disease etiology and preventionFocus on individual, as opposed to group

etiologiesAnamnesis or history taking from patientsGrouping individual cases together into seriesMaking comparisons of the differences between

groups, in order to elicit average risk at the level of individual

History of case control studies

Page 15: Case control study

Concept found in works of Parisian physician PCA Louis (1788-1875) - “numerical method”, a technique whose principal tool was the tabulation of aggregated data about patients with similar pathologic and clinical findings.

First explicit description by William Augustus Guy (1843) – analysis of relationship of prior occupational exposure and occurrence of pulmonary consumption.

Baker (1862) – case control comparisons of marriage and fertility in breast cancer patients.

History of case control studies

Page 16: Case control study

Lane Claypon’s Breast cancer study 1926 -‘‘A further report on cancer of the breast: reports on public health and medical subjects.’’ (Lane-Claypon 1926a).500 hospitalised cases and 500 controls with non-

cancerous illnesses22% lower fertility in the case group.

1950 - Four studies that implicated cigarette smoking in cancer of the lung published in 1950 in the United States (Levin et al 1950; Wynder & Graham 1950; Schrek et al. 1950) and in Britain (Doll & Hill 1950), have established several features of the modern form of the case-control study. Doll & Hill’s study is perhaps the most well known in history.

History of case control studies

Page 17: Case control study

Design of Case Control StudyThe investigator selects

cases with the disease and appropriate

controls without the disease and obtains

data regarding past exposure to possible etiologic factors in both groups.

The investigator then compares the frequency of exposure of the two groups.

Page 18: Case control study

Design of Case Control Study

Hallmark of Case Control Study: Starts from cases and controls and searches for exposure.

Disease No Disease

“CASES” “CONTROLS”

Not ExposedExposed Exposed Not Exposed

Page 19: Case control study

Design of Case Control Study

FIRST: Select CASES CONTROLS

(With Disease) (Without Disease)

THEN: Were exposed a b

Measure

Exposure Were not exposed c d

TOTALS a + c b + d

Proportions a bExposed a + c b + d

Page 20: Case control study

Design of Case Control Study

Selection of CASES:1. Representativeness: Ideally, cases are a random sample of

all cases of interest in the source population (e.g. from vital data, registry data).

More commonly they are a selection of available cases from a medical care facility. (e.g. from hospitals, clinics)

Information: can be collected from cases themselves, or from a respondent by proxy (relative/ friend), from records or a combination of the above.

Page 21: Case control study

Design of Case Control Study

Selection of CASES:

2. Method of Selection Selection may be from incidence or

prevalence case:

• Incident cases are those derived from ongoing-ascertainment of cases over time.

• Prevalent cases are derived from a cross-sectional survey.

Page 22: Case control study

Selection of CASES:2. Method of SelectionSelection of INCIDENT CASES is OPTIMAL.

These should be all newly diagnosed cases over a given period of time in a defined population.

However we are excluding patients who died before diagnosis. A difficult problem ???

Prevalent cases do NOT include patients with a short course of disease. So patients who recovered early and those who

died will not be included.

Additional protection against bias by including deceased cases as well as those alive

Design of Case Control Study

Page 23: Case control study

Design of Case Control Study

Selection of CASES:

3. Diagnostic criteria for case studies

a) Specificity b) Diagnostic bias c) Validation

Diagnostic criteria regarding diagnosis of cases, types of cases and stage of disease to be included should be predefined.

Validity is more important than generalizability i.e. the need to establish an etiologic relationship is more important than to generalise results to the population.

Page 24: Case control study

Selection of CASES:3. Diagnostic criteria for case studies Example:

In a study on breast cancer – we can include all cases OR we can include only premenopausal women with lobular cancer.

If we take the later group as cases; we can elicit the etiology better.

Design of Case Control Study

Page 25: Case control study

Selection of CONTROLS:

(i) Should the controls be similar to the cases in all respects other than having the disease? i.e. COMPARABLE

(ii) (ii) Should the controls be representative of all non-diseased people in the population from which the cases are selected? i.e. REPRESENTATIVE

Design of Case Control Study

Page 26: Case control study

Selection of CONTROLS: Comparability vs

Representativeness

The control group should be representative of the general population in terms of probability of exposure to the risk factor

AND they should also have had the same opportunity to be exposed as the cases have. Not that both cases and controls are

equally exposed; but only that they have had the same opportunity for exposure.

Design of Case Control Study

Page 27: Case control study

Selection of CONTROLS: Comparability vs Representativeness

Usually, cases in a case-control study are not a random sample of all cases in the population. And if so, the controls must be selected in the same way (and with the same biases) as the cases.

If follows from the above, that a pool of potential controls must be defined. This is a universe of people from whom controls may be selected (study base).

Design of Case Control Study

Page 28: Case control study

Selection of CONTROLS:

The study base is composed of a population at risk of exposure over a period of risk of exposure.

Cases emerge within a study base. Controls should emerge from the same study base, except that they are not cases.

For example, if cases are selected exclusively from hospitalized patients, controls must also be selected from hospitalized patients.

Design of Case Control Study

Page 29: Case control study

Design of Case Control Study

“Total” Population

Reference Population

Cases Controls

Page 30: Case control study

Selection of CONTROLS: Criteria

Comparability is more important than representativeness in the selection of controls

The control should be at risk of the disease

The control should resemble the case in all respects except for the presence of disease (and any as yet undiscovered risk factors for disease)

Design of Case Control Study

Page 31: Case control study

Selection of CONTROLS: Sources

Design of Case Control Study

Source Advantage Disadvantage

Hospital based Easily identified.Available for interview.More willing to

cooperate.Tend to give complete and accurate information (recall bias).

Not typical of general population.Possess more risk factors for disease.Some diseases may share risk factors with disease under study. (whom to exclude???)Berkesonian bias

Population based(registry cases)

Most representative of the general population.Generally healthy.

Time, money, energy.Opportunity of exposure may not be same as that of cases. (locn, occu,)

Neighbourhood controls/ Telephone exchange random dialing

Controls and cases similar in residence.Easier than sampling the population.

Non cooperation.Security issues.Not representative of general population.

Best friend control/ Sibling control

Accessible, Cooperative.Similar to cases in most aspects.

Overmatching.

Page 32: Case control study

Selection of Controls : Number

o Large study: Cases: Control :: 1:1o Small study: Cases: Control :: 1:2, 1:3, 1:4.o Use of multiple controls

1. Controls of same type: Cases: Control :: 1:1 ( for rare diseases, cases

cannot be increased in that time), ( increases power of the study).

2. Multiple controls of different types: controls- 1 hospital, 1 neighborhood e.g. case-

Children with brain tumor, control- children with other cancer, normal children, risk factor- h/o radiation exposure.

Design of Case Control Study

Page 33: Case control study

Multiple controls of different types are valuable for exploring alternate hypothesis & for taking into account possible potential recall bias.

(From Gold EB, Gordis L, Tonascia J, Szklo M; Risk factors for brain tumors in children. Am J Epidemiol 1979)

Design of Case Control Study

Children with brain

tumours

Children with other cancers

Children without cancer

Radiation causes

cancers

Radiation causes brain

cancers only

Page 34: Case control study

Design of Case Control Study

Selection of Controls: Objectives

Elimination of selection bias - Selection

Minimization of information bias - Blinding

Minimization of confounding - Matching

Page 35: Case control study

Problems in control selection – Confounding variables.Confounding variables are factors associated

with the exposure of interest and causally with the disease of interest.

May lead to a spurious/ biased relationship between risk factor and disease.

Common confounding variables are : age, sex, educational status, socioeconomic level, etc.

These can be adjusted by : Designing the study through Matching Statistical techniques like Stratification and

Regression

Design of Case Control Study

Page 36: Case control study

Matching:Definition: It is the selection of controls

so that they are similar to the cases in specified characteristics. (Epidemiology: An Introductory Text; Mausner & Bahn, 1985)

Matching is defined as the process of selecting controls so that they are similar to cases in certain characteristics such as age, sex, race, socioeconomic status and occupation. (Epidemiology; Leon Gordis, 2004)

Design of Case Control Study

Page 37: Case control study

Matching:Matching variables (e.g. age), and matching criteria

(e.g. within the same 5 year age group) must be set up in advance.

Controls can be individually matched (most common) or Frequency matched. Individual matching (Matched pairs): search

for one (or more) controls who have the required matching criteria, paired (triplet) matching is when there is one (two) control (s) individually matched to each cases.

Group matching (Frequency matching): select a population of controls such that the overall characteristics of the case, e.g. if 15% cases are under age 20, 15% of the controls are also.

Design of Case Control Study

Page 38: Case control study

Matching:Avoid over-matching, match only on

factors KNOWN to be cause of the disease.

Obtain POWER by matching MORE THAN ONE CONTROL per case. In general, N of controls should be ≤ 4, because there is no further gain of power above that.

Obtain Generalizability by matching by matching more than one type of control.

Design of Case Control Study

Page 39: Case control study

Matching: Problems – Individual matching on too many

variables – is time consuming, costly, cumbersome and may lead to too less controls.

Cannot explore possible association of disease with any variable on which cases and controls have been matched. Therefore only factors which are known to be associated with the disease are studied. Suppose we know that breast cancer rates are

higher among single women than in married women; then matching cases for marital status would spuriously NOT detect any relation regarding this factor.

Design of Case Control Study

Page 40: Case control study

Matching: Problems –Overmatching: Matching on variables other than

those that are risk factors for the disease under study, either in a planned manner or inadvertently.

Example: In a study on OCP use as a risk factor for cancer, if we use “best friend controls”, it is most likely that the controls would also be OCP users. In effect we would have matched for the very factor we want to study.

Example: If we use neighbourhood controls in a study on nutrition and tuberculosis, we would be inadvertently matching for socioeconomic status and thus nutrition.

Design of Case Control Study

Page 41: Case control study

Bias in Case control Study Definition: Any systematic error in the design, conduct, or analysis of a study that results in mistaken estimates of the effect of the exposure on disease.

Types of bias in case control studies: Selection bias Information bias Confounding bias

Page 42: Case control study

Selection Bias: Sources –

1. Selective loss to follow-up 2. Incomplete ascertainment of cases

(Detection or Diagnostic bias)3. Inappropriate control group4. Differential motivation to participate

Bias in Case control Study

Page 43: Case control study

Selection Bias:

Selective survival - only surviving subject available to be studied; those surviving differ from those dying in potentially important ways.

Solution: :Rapid case ascertainment and interview

Bias in Case control Study

Page 44: Case control study

Information Bias:Occurs due to -

1. Imperfect definitions of study variables

OR2. Flawed data collection procedures.

Leads to – Misclassification of disease and exposure.

Types of Information bias – Recall bias Interviewer bias

Bias in Case control Study

Page 45: Case control study

Misclassification Some of the cases or controls who were actually exposed will be

erroneously classified as unexposed, and some who were actually not exposed will be erroneously classified as exposed.—this generally results in an underestimate of the true risk of the disease associated with the exposure.e.g. cervical cancer with sexual intercourse with uncircumcised menComparison of patients’ statements with examination findings concerning circumcision status, Roswell Park Memorial Istitute, New York

Patients statement regarding circumcision

Examination finding

Yes (no.) Yes(%) No (no.) No(%)

circumcised

37 66.1 47 34.6

not-circumcised

19 33.9 89 65.4

Total 56 100.0 136 100.0

Page 46: Case control study

Bias in Case control Study

Recall bias (usually in case-control studies): Cases who are aware of their disease status may be more likely to recall exposures than controls e.g. congenital malformation with prenatal infections

Results in misclassification

Solution• Achieving similarity in the procedures used to obtain information from cases and controls

• Verify exposure with existing records• Objective measure of exposure• Use of information recorded prior to the time of diagnosis.

Page 47: Case control study

Interviewer bias: When interviewer is not blinded (knows) case status of subjects there is potential for interviewer bias.Leads to –

If interviewer knows case status – differential misclassification likely.

If interviewer does not know case status – non differential misclassification is still possible.

Solution – Blinding of interviewer as to case status Equal interview time for all participants

Bias in Case control Study

Page 48: Case control study

Confounding: When a measure of the effect of an exposure on risk is distorted because of the association of exposure with other factors that influence the outcome.

Not possible to separate the contribution that any single causal factor has madeConfounding Factor: is one which is associated

with both exposure & disease , and is distributed unequally in study & control groups.

E.g.: Alcohol & Esophageal Ca ; confounding factor- smoking

Solution: Study design : Matching Analysis: Stratification &

Regression

Bias in Case control Study

Page 49: Case control study

Outcomes of Case Control Study

On analysis of case control study we find out

Exposure rates: the frequency of exposure to suspected risk factor in cases and in controls

Estimation of disease risk associated with exposure: (Odds ratio)

Page 50: Case control study

Exposure rates:A case control study provides a direct

estimation of the exposure rates (frequency of exposure) to the suspected factor in disease and non-disease groups.

Doll R. and Hill AB. (1950) Brit. Med. J.

Exposure rates Cases = a/ (a + c) = 33/ 35 = 94.2% Controls = b/ (b + d) = 55/82 = 67.0%

Outcomes of Case Control Study

Cases (lung cancer)

Controls (without lung

cancer)

Smokers 33 (a) 55 (b)

Non Smokers 2 (c) 27 (d)

TOTAL 35 (a + c) 82 (b+d)

Page 51: Case control study

Outcomes of Case Control Study

Odds Ratio / Relative odds (estimate of relative risk).

Odds: Odds of an event is defined as the ratio of the number of ways an event can occur to the number of ways an event cannot occur. (Epidemiology; Leon Gordis. 2004)

If the probability of event X occurring is P, then odds of it occurring is = P/ 1-P.

Odds ratio: Ratio of the odds that the cases were exposed to the odds that the controls were exposed.

Page 52: Case control study

Odds ratio:Using the four-fold table –

Odds that case was exposedOdds ratio =

Odds that control was exposed

= (a/c)/ (b/d) = ad / bc

Outcomes of Case Control Study

Diseased/ Cases

Not diseased/ Controls

Exposed a b

Not exposed c d

Page 53: Case control study

Odds ratio ( = cross products ratio) can

also be viewed as the ratio of the product of the

two cells that support the hypothesis of an

association (cells a & d – diseased people who

were exposed and non diseased people who

were not exposed), to the product of the two

cells which negate the hypothesis of an

association (cells b & c – non diseased people

who were exposed and diseased people who

were not exposed).

Outcomes of Case Control Study

Page 54: Case control study

When is Odds ratio a good estimate of the relative risk in the population?

Cases studied are representative Regarding history of exposure of all people with

the disease in the population from which cases are drawn.

Controls studied are representative Regarding history of exposure of all people

without the disease in the population from which cases are drawn

When the disease being studied does NOT occur frequently

Outcomes of Case Control Study

Page 55: Case control study

Limitations/ Disadvantages 1. Susceptible to bias if not carefully

designed

2. Especially susceptible to exposure misclassification

3. Especially susceptible to recall bias

4. Restricted to single outcome

5. Incidence rates not usually calculate

6. Cannot assess effects of matching variables

Page 56: Case control study

Advantages

1. Only realistic study design for uncovering etiology in rare diseases

2. Important in understanding new diseases

3. Commonly used in outbreaks investigation

4. Useful if inducing period is long

5. Relatively inexpensive

Page 57: Case control study

ApplicationsRare disease:

Case-control approaches are the most efficient for rare diseases, e.g idiopathic pulmonary fibrosis, most cancers. Cohort approaches would require large populations and prohibitive expense and follow-up time.

Page 58: Case control study

Case ascertainment system in place: The conduct of a case-control study may be facilitated by the availability of a case-ascertainment system. a) Population-based cancer registry b) Hospital-based surveillance systems c) Mandated disease reporting systems

When funding and time constraints are not compatible with a cohort study.

Applications

Page 59: Case control study

Nested Case-Control Studies

Study Population

TIME 1

YEARS

TIME 2DevelopDisease

Do NotDevelopDisease

CASES CONTROLS (Subgroup)

CASE-CONTROL STUDY

Obtain interviews, blood, urines, etc.

Page 60: Case control study

X

X

OX

Time

t1

O

Consider the following hypothetical cohort:

X = lung cancer caseO = loss to follow-up

t2 t3

Nested Case-Control Studies

Page 61: Case control study

Advantages:

1. Possibility of recall bias is eliminated, since data on exposure are obtained before disease develops.

2. Exposure data are more likely to represent the pre-illness state since they are obtained years before clinical illness is diagnosed.

3. Costs are reduced compared to those of a prospective study, since laboratory tests need to be done only on specimens from subjects who are later chosen as cases or as controls.

Nested Case-Control Studies

Page 62: Case control study

Important findings of case control studies 1950’s

Cigarette smoking and lung cancer

1970’s Diethyl stilbestrol and vaginal adenocarcinoma

Post-menopausal estrogens and endometrial cancer 1980 ’s

Aspirin and Reyes sydromeTampon use and toxic shocks syndrome

L-tryptopham and eosinophilia-myalgia syndromeAIDS and sexual practices

1990’sVaccine effectivenessDiet and cancer

Page 63: Case control study

References: Park’s Textbook of Preventive and Social

Medicine – 21st ed; Park JE. 2010. Mausner & Bahn Epidemiology: An

Introductory Text – 2nd ed; Mausner JS, Kramer S. 1985.

A Dictionary of Epidemiology – 3rd ed; Last JM. 2000.

Epidemiology – 3rd ed; Gordis L. 2004. Origins and early development of the

case-control study by Nigel Paneth, Ezra Susser, Mervyn Susser. Available from www.epidemiology.ch/history/papers.

Page 64: Case control study

Thank You