10 Sept 2025

CASE CONTROL STUDY

AND
COHORT STUDY
BY GROUP: 5, 22MUG17-21
AMOGH K ADKOL
ANAGHA GOK ARE
ANANTH K K ASB E
ANGADI NANJ UNDAPPA
ANIK ET AGARWAL
CASE Case co ntro l studi es, o ften cal l ed

CONTROL "retro specti ve stu di es" a re a co mmo n

fi rst a ppro a ch to test causal
STUDY h ypo thesi s. I n recent yea rs, th e case
co n tro l appro ach h as emerged as a
metho d of epi demi o l o gi cal
i n vesti ga ti o n.

By defi n i ti o n , a case co ntro l study

i n vo l ves two po pu l ati o n s - cases a nd
co n tro l s. I n ca se co ntro l stu di es, the
u ni t i s the i n di vi du al rath er th an th e
gro u p. The fo cu s i s o n a di sea se o r
so me o ther h ea l th pro bl em that ha s
al rea dy devel o ped.
THREE DISTINCT FEATURES OF
CASE CONTROL METHOD

01 B oth exp osure and outcome ( d isease) hav e occurred b efore

the start of the stud y.

02 T he stud y p roceed s b ackward s from eff ect to cause.

I t uses a control or com p arison g roup to sup p ort or refute an

03
inference.
 T he b asic d esig n of a case control
stud y is shown in Tab le b elow. I t is a
2x 2 tab le which p rovid es a v ery
FRAMEWORK useful fram ework to d iscuss the
v arious elem ents which m ake up a
OF A CASE case control stud y.

CONTROL
STUDY
 To illustrate, if it is our intention to test
using the case control m ethod , the
FRAMEWORK inv estig ation b eg ins by assem b ling a
g roup of lung cancer cases ( a + c ), and a
OF A CASE g roup of suitab ly matched controls ( b +

CONTROL d ) . One then exp lores the p ast history of

these two g roup s for the p resence or
STUDY ab sence of sm oking , which is susp ected to
be related to the occurrence of lung
d ancer. If the freq uency of sm oking ,
a/( a+c) is hig her in cases than in controls
b /( b +d ), an association is said to ex ist
b etween sm oking and lung cancer.
STEPS IN CONDUCTING A CASE
CONTROL STUDY

01 02
S E LE CTI O N OF CASE S M ATCH IN G
A N D CO N TROLS

03 04
M EA S UR E M EN T O F A N A LY S IS A N D
EX P O S UR E IN TERP RETAT IO N
01. SELECTION OF CASES

Defi nition o f a c as e : T he p rio r d e fi nit io n o f w hat c o nstitutes a "case" is

crucial to the c as e c o nt ro l s t ud y. It invo lv e s t w o s p e c ifi c at io ns :

(i) DIA G NOS T IC C R IT E R IA : T he d iag no s t ic c rit e ria t he d is eas e and the s tag e o f
d iseas e, if any (e . g . , b re ast c anc e r S t ag e I) t o b e inc lud ed in t he st ud y m us t
b e sp ecifi ed b e fo re t he s t ud y is und e rt ake n. S up p o s ing w e are investig at ing
cases o f canc e r, w e s ho uld b e q uit e c le ar t ha t w e have, for o ur cas es , a
g roup his tolo g ic a lly t he s a m e . (Onc e t he d iag no s t ic c riteria are es tab lished ,
they sho uld no t b e alt e re d o r c hang e d t ill t he s t ud y is o ver)
(ii) E LIG IBIL IT Y C R IT E R IA : T he s e c o nd c rit e rio n is that of elig ib ility. (A
criterio n cus t o m arily e m p lo ye d is t he re q uire m e nt t hat o nly new ly d iag nos ed
(incid ent ) ca s e s w it hin a s p e c ifi e d p e rio d o f t im e are e lig ib le than old cases
o r cas es in a d v anc e d s t a g es o f t he d is e as e (p re va le nt c ases).
01. SELECTION OF CONTROL

T he controls must b e free from the d isease und er stud y. They m ust b e as
sim ilar as to the cases as p ossib le, exp ect for the ab sence of d isease
und er stud y. As a rule, a com p arison g roup is id entifi ed b efore a stud y is
d one,com p rising of p ersons who hav e not b een ex p osed to the d isease or
som e other factor whose infl uence is b eing stud ied .

S ources of C ontrol :
1) Hosp ital control : T he controls m ay b e selected from the sam e hosp ital
as the cases, b ut with d iff erent illnesses other than the stud y d isease.
2) Relatives
3) Neig hb ourhood controls
4) G eneral p op ulation
02. MATCHING

Th e co ntro l s may diff er from the cases i n a nu mber o f fa cto rs su ch a s a ge, sex,
o ccu pa ti o n , social status, etc. An i mpo rtan t co n si derati o n is to en su re
co mparabi l i ty between cases a nd co n tro l s. Ma tchi n g i s defi ned a s th e pro cess
w hi ch as sel ect co ntrols in such a wa y th at they a re si mi l a r to cases w i th rega rd
to certai n pertinent selected va ri a bl es (e.g., a ge) w hi ch are kno wn to i n fl uence
the o utco me o f disease and w hi ch, i f n o t adequ atel y ma tched fo r co mpara bi l i ty,
co u l d di sto rt o r confound the resul ts)

A " confounding factor " is defi n ed as one w hi ch is asso ci ated bo th wi th

expo sure a nd
di sea se, an d is distributed u nequ al l y i n stu dy an d co ntro l gro ups. eg : I n the
stu dy o f th e ro le of alcohol i n the aeti o l o gy o f o eso ph agea l can cer, smo ki n g i s a
co n fo u ndi n g factor because (i ) i t i s asso ci ated wi th the co n sumpti o n o f a l co h o l
02. TYPES OF MATCHING

1) Group matching : Thi s may be do ne by assi gni n g cases to su b-catego ri es

(strata) based on their cha ra cteri sti cs (e.g., age, o ccupati o n, cl ass) a nd
appro pri a te co ntrols.

2) Matching by pairs : Match i n g i s al so do n e by pai rs. Fo r exampl e, fo r ea ch

case, a co n tro l is chosen whi ch can be match ed qui te cl o sel y. Thu s, i f w e h ave a
50 year o l d mason with a parti cu l ar di sease, we wi l l search fo r 50 year o l d ma so n
w i th o ut th e disease as a co ntro l . Thu s o ne can o btai n pai r o f pa ti ents an d
co n tro l s o f th e same sex, age, du ra ti o n a nd severi ty o f i l l ness, etc. But th ere may
be grea t diffi culties in o bta i ni n g cases a nd co n tro l s matched on all
cha ra cteri sti cs, and it may be n ecessary to wa i t a co n si derabl e peri o d ti me
befo re o bta i ni ng a suffi cient n umber o f match ed pa i rs.
03. MEASUREMENT OF EXPOSURE

Defi n iti o ns and criteria abou t expo sure (o r vari abl es wh i ch may be o f a eti o l o gi cal
i mpo rtan ce) are just importan t as th o se used to defi ne cases a nd co n tro l s.
I n fo rmati o n abo ut exposure sh o u l d be o btai n ed i n preci sel y the sa me man n er
bo th fo r cases and contro l s. Thi s may be o btai ned by i ntervi ews, by
questi o n n ai res o r by studyin g past reco rds o f cases such as ho spi tal reco rds,
empl o yment reco rds, etc. It i s i mpo rtant to reco gn i ze th at wh en case co n tro l
stu di es are being used to test a sso ci ati o ns, th e mo st i mpo rta nt facto r to be
co n si dered, even mo re impo rta nt th an the P. val ues o bta i ned, i s the qu esti o n o f
"bi as". o r systematic error whi ch must be rul ed o u t
04. ANALYSIS AND INTERPRETATION

Th e fi na l step is analysis to fi nd o ut:

1) Expo sure rates amo ng cases a nd co n tro l s to su spected fa cto r
2) Esti mati o n of disease risk asso ci ated wi th expo su re (o dds ra ti o )

1.EXPOSURE RATES:

A: cases = a/ (a+c) = 33/ 35=0.94.2 %

B : Co ntro l s = b/(b+d) = 55/82=0.67%

P <0.001
04. ANALYSIS AND INTERPRETATION

The Table shows that the frequency r ate of lung

cancer was defi nit el y higher among smoker s than
among non-smoker s. The next step wil l be to
ascer t ai n whether t here i s a statist ical association
between exposure status and occurrence of l ung
cancer. Thi s questi on can be resolved by
cal cul ati ng the P. val ue, which in this case i s l ess
than 0.001. Accordi ng to convent ion, i f P is l ess
than er equal to 0.05, it is regarded as
"stat isti cal ly signifi cant" The small er the P . value,
the great er the stati sti cal signi fi cance or
probabil ity t hat the associ ati on al one. However ,
stati sti cal associati on (P. val ue) does not impl y
04. ANALYSIS AND INTERPRETATION

2. ESTIMATION OF RISK :

It sh ou l d be n ot ed (Tabl e above) t hat if t he exposure r ate was 94.2 per cent i n the
st u dy grou p, i t does n ot mean that 94.2 per cent of those smoked would develop l ung
can cer. t h e est i mat i on of di sease r isk associ ated wi th exposure i s obtai ned by an
i ndex kn own a " Rel at i ve R i sk" (R R) or "r i sk r ati o" , whi ch i s defi ned as the Rat io
bet ween t he i n ci den ce of di sease exposed per sons and incidence among non-exposed.
It i s gi ven by t h e formu l a:

relativ e r isk = In ci den ce among exposed

In ci den ce am ong non-exposed
From a c a s e c on t rol s t u d y , w e c a n
d eri v e w h a t i s k n ow n a s O d d s Ra t i o
(O R) which is a m ea s u re of the
ODDS RATIO
s t ren g t h of t h e a s s oc i a t i on b et w een (CROSS-PRODUCT
r i s k fa c t or a n d ou t c om e. O d d s ra t i o
i s c l os el y rel a t ed t o rel a fi v e ri sk . Th e RATIO)
d eri v a t i on of od d s ra t i o i s b a s ed on
t h ree a s s u m p t i on s :

( a) t h e d i s ea s e b ei n g i n v es t i g a t ed
m u st b e rel a t i v el y ra re.
( b ) t h e c a s es m u s t b e rep res en t a t i v e
of t h os e w i t h t h e d i s ea s e, a n d
(c) the c on t rol s m u st be
rep res en t a t i v e of t h os e w i t h ou t t h e
 BIAS IN CASE CONTROL STUDIES

Bi a s is a ny sy st em a t i c error i n t he d et erm i na t i on of t he a ssoc ia t ion b et w een

t he ex p osure a nd d i sea se.
Ty p es of b ia s d u r i ng ep i d em i ol og i c a l st ud i es :

a) Bia s d ue to c onfound ing : M ent i on ha s a l rea d y b een ma d e a b out

c onfound ing a s a n i m p or t a nt sourc e of b i a s. T hi s b i a s c a n b e remov ed b y
m a t c hing in c a se c ont rol st ud i es.

b )M emory or rec a l l b i a s : When c a ses a nd c on t rol s a re a sked q uest ions a b out

t hei r p a st hist or y , i t m a y b e m ore l i kel y for t he c a ses t o reca ll t he ex ist enc e
of c e rt a in ev ent s or fa c t or s, t ha n t he c ont rol s w ho a re hea l t hy p ersons.

c )S el ec t ion b i a s: Th e c a se s a nd c ont rol s m a y n ot b e rep resent a t iv e .

d ) i nt erv iew er ' s b i a s : Bi a s m a y a l so oc c ur w hen t he i nt erv iew er k now s t he

 EXAMPLE OF CASE CONTROL
STUDY

T h a lidom ide tra ged y :

Tha l i domi de was fi rs t marketed as a s afe, non-ba rbi tur a te hypnoti c i n

B r i ta i n i n 1958. In 1961, at a congress of Gynaecol ogi s ts , a ttenti on wa s
dr a wn to the bi rth of a l arge number of babi es wi th congeni ta l
a bnorma l i ti es , whi ch was previ ousl y rare. In the s ame yea r , i t wa s
s ugges ted that thal i domi de mi ght be responsi bl e for i t. A retros pecti ve
s tudy of 46 mothers del i vered of deformed babi es showed tha t 41 were
found to have thal i domi de duri ng thei r earl y pregna ncy. Thi s wa s
compa red wi th a control of 300 mothers who had del i vered norma l
ba bi es ; none of thes e ha d taken thal i domi de. Laborator y exper i ments
confi rmed that thal i domi de was teratogeni c i n experi menta l s tudi es
THANK YOU
COHORT STUDIES
COHORT
Co h o r t stu d y is a n o th er type of a n al yti c al
STUDIES ( o b ser va ti o n a l )
stu d y w h i c h i s u su a l l y u n d er ta ken to o btai n ad di ti o n a l
evi d en c e to refu te o r su p p o rt th e exi sten c e o f a n
a sso c i a ti o n
b etw een su sp ec ted c a u se a n d di sease.
In ep i d emi o l o g y, th e term "c o h o r t" i s d efi n ed as a
g ro u p of p eo p l e who sh are a c o mmo n
c h a r a c ter i sti c o r exp er i en c e wi th i n a defi n ed ti me
p er i o d ( e. g . , a g e, o c c u p a ti o n , expo su re to a dr u g
o r va c c i n e, p reg n a n c y, i n su red p er so n s, etc )
Co h o r t stu d y i s kn o wn b y a var i ety o f n ames:
p ro sp ec ti ve stu d y, l o n g i tu d i n al stu dy, i n c i den c e
stu d y, a n d fo r w a rd -l o o ki n g stu d y. Th e mo st wi del y
u sed term, h o w ever , i s "c o h o r t stu d y"
DISTINGUISHING FEATURES OF
COHORT STUDIES

1.The cohorts are identifi ed prior to the appearance of the

disease under investigation.
2.The study groups, so defi ned, are observed over a period
of time to determine the frequency of d isease among
them
3.The study proceeds forward from cause to eff ect
1.Wh en th ere is good eviden ce of
an a ssocia tion bet w een
exposu re a nd disea se as
derived from cl i n i ca l INDICATIONS
observa ti ons a nd support ed by
descri pt i ve a nd ca se con t rol FOR COHORT
stu di es.
2.Wh en exposure is ra re, bu t th e STUDIES
inci den ce of disea se h i gh
a m on g exposed, e. g. speci a l
exposu re groups like th ose i n
indu stri es, exposure to X-ra ys,
etc.
3.Wh en a ttrition of stu d y
popu l a t ion ca n be min i m ized,
e. g. , fol l ow -up is ea sy, coh ort i s
FRAMEWORK OF A COHORT STUDY

We begi n wi t h a grou p or cohor t (a+b) exposed to par ti cul ar factor thought to be

rel at ed t o di sease occurren ce, and a group (c+d) not exposed to that par ticular factor.
The former i s kn own as “st u dy cohor t” and l at ter “control cohor t”
IN ASSEMBLING COHORTS, THE FOLLOWING GENERAL
CONSIDERATIONS ARE TAKEN INTO ACCOUNT:

1.T he c o ho rt s m us t b e fre e fro m t he d is e a s e und er s tud y.

2.Ins o fa r as t he kno w le d g e o f t he d is e as e p erm its , b oth
t he g ro up s (i. e : , s t ud y and c o nt ro l c o ho rt s) sho uld b e
e q ua lly s us c e p t ib le t o t he d is e as e und e r stud y, o r
e ffi c ie nt ly re fl e c t a ny d iff e re nc e in d is e a s e occurrence.
3.B o t h t he g ro up s s ho uld b e c o m p a rab le in resp ect of all
t he p o s s ib le va ria b le s , w hic h m ay infl uence the
fre q ue nc y o f t he d is e as e .
4.T he d iag no s t ic a nd e lig ib ilit y c rit e ria o f the d isease
m us t b e d e fi ne d b e fo re ha nd ; t his w ill d e p e nd up on the
a vailab ilit y o f re liab le m e t ho d s fo r re c o g niz ing the
d is e as e w he n it d e v e lo p s .
TYPES OF COHORT STUDIES

01 02
P R O S P E CTIVE RE TROS PE CTIVE

03
C O M B IN ATIO N O F
P R O S P EC T IV E A N D
R ETR O S P EC T IV E
01. PROSPECTIVE COHORT STUDIES

A pros pect i ve cohort s t udy i s one

which t he out come has not y et
occurred at t he t i me of i nv es t i gat i on
begi ns . Mos t pros pect i v e s t udi es
behi n i n t he erre s ent as a cont i ns e
i nt o For exampl e, t he l ong- t erm
eff ect s of ex pos ure t o urani um w as
ev al uat ed by i dent i fy i ng a group of
uranium mi ners and a compari s on
group of i ndiv idual s not ex pos ed t o
uranium mi ni ng and by as s es s i ng
s ubs equent dev el opment of l ung
cancer i n bot h t he groups .
02. RETROSPECTIVE COHORT
STUDIES
A ret rosp ec t ive c ohort stud y (or
" hist oric al" c ohort s tud y) is one in
whic h t he outc omes have all
oc c urred b efore t he s tart of the
investi g ation. The inves tig ator g oes
b ac k in t ime, s ometimes 10 to 30
years, to selec t his st ud y g roup s
from rec ord s of p as t emp loyment,
med ic al rec ord s and trac es them
forward t hroug h t ime, from a p ast
d at e fi xed on t he rec ord s, usually up
to t he p resent .
03. MIX OF PROSPECTIVE AND
RETROSPECTIVE COHORT STUDIES

In this typ e of st ud y, b oth the retrosp ec tive and prosp ec tive elements are
c omb i ned . The c ohort is id entifi ed from p ast rec ord s, and is assessed of d ate
for t he outc ome. T he same c ohort is followed up p rosp ec tively into future for
furt her ass es sment of outc ome.
 ELEMENTS OF A COHORT STUDY

01 02 03
S E LE CTION O F S TU DY OBTAIN IN G DATA ON S EL E C T IO N O F
S U BJECTS EXPOS U RE C O M PA R IS O N
G R O UP S

04 05
F O LLO W- U P AN ALY S IS
01. SELECTION OF STUDY SUBJECTS

The s ubj ect s of a cohort s t udy are us ual l y as sembl ed in one of t w o w ay s -

ei t her f rom general popul at i on or s el ect group of populati on t hat can be readi l y
s t udi ed.
eg: pers ons w i t h di ff erent degrees of exposure t o t he sus pect ed caus al f act or.

G en era l p op u la tion : When the ex pos ure or caus e of deat h i s f ai rl y f requent i n

t he popul at i on, cohort s may be as sembl ed from t he general popul at i on, res i di ng i n
w el l - defi ned geographi cal , poli t i cal and admi ni st rat iv e areas .

Sp ecia l g rou p s : T hes e may be s pecial groups or ex pos ure groups t hat can
readi l y be s t udi ed.

a) Sel ect groups : t hes e may be prof es si onal groups eg doct ors , l aw y ers , t eachers .
b) E x pos ure groups : I f t he expos ure i s rare, economi cal procedure i s t o s el ect a
02. OBTAINING DATA ON EXPOSURE

I nf ormat i on about ex pos ure may be obt ai ned di rect l y from t he

• C oh ort memb ers : t hrough personal intervi ews or mail ed ques t i onnai res . Si nce
cohort s t udi es i nv ol v e l arge numbers of popul at i on, mai l ed ques t i onnai res off er
a s i mpl e and economi c w ay of obt ai ni ng i nformati on.

• Review of record s : Cert ain ki nds of i nf ormati on (e:g., dos e of radi at i on, ki nds
of s urgery , or det ai l s of medical t reat ment ) can be obt ai ned onl y f rom medi cal
records .

• M ed ica l ex a min a tion or s p ecia l tes t : some t ype of i nf ormat i on can be

obt ai ned onl y by medi cal examinat ion or speci al t es t s , e. g. , bl ood
pres s ure, s erum chol es t erol , etc.

• En viron men ta l s u rveys : Thi s i s t he bes t s ource f or obt ai ni ng i nf ormat i on on

03. SELECTION OF COMPARISON
GROUPS
M any w ay s o f a s s e m b l i n g c o m p ari s o n g ro up s :

1.I nte rnal co m p ar is o ns : I n s o m e c o ho rt s t u d i e s ,

no o ut s i d e c o m p ari s o n g ro up i s re q ui re d . T he
com p ari s o n g ro up s are i n- b u i l t . T h at i s , s i n g l e
coh o rt e n t e rs t h e s t u d y , a nd i t s m e m b e rs m ay ,
o n t he b a s i s o f i nf o rm a t i o n o b t a i ne d . I t s ho w s
t hat m o rt a l i t y fro m l u ng c an c e r i nc re a s e s w i t h
i ncre a s i ng n um b e r of ci g are t t e s s m o ke d
re i nf o rc i ng t he c o n c l u s i o n t ha t t he re i s va l i d
as s o c i at i o n b e t w e e n s m o k i ng a nd l un g c a nc e r.
2.E xte rnal co m p ar is o n : W he n i nf o rm a t i o n o n
d e g re e o f ex p o s u re i s no t av ai l ab l e , i t i s
ne c e s s ary t o p ut u p an ext e rna l c o nt ro l , t o
e val ua t e t he ex p e ri e n ce o f t he ex p o s e d g ro up ,
e .g ., s m o ke rs an d no n- s m o ke rs , a c o ho rt o f
rad i o l o g i s t s c o m p a re d with a co ho rt of
03. SELECTION OF COMPARISON
GROUPS
3 . C o m p a ris o n wi th g e ne ral p o p ula tio n rat e s I f no ne i s av ai l ab l e , t he m ortal i t y exp e ri e nce of
t he ex p o s e d g ro up i s co m p are d w i t h t he m o rt al i t y ex p e ri e nc e o f t h e g e ne ral p o p ul at i o n i n t he
s am e g e o g ra p h i c a re a as t he ex p o s e d p e o p l e , e .g ., c o m p a ri s o n o f fre q ue ncy o f l ung cance r
a mo ng u ra ni um m i ne w o rke rs w i t h l u ng c an c e r m o rt al i t y i n t he g e ne ral p op ul at i on w he re the
m i ne rs re s i d e d w i t h co m p ari s o n of fre q u e nc y o f c a nc e r a m o n g a s b e s t o s w orke rs w i t h t he rat e i n
g e ne ra l p o p u l at i o n i n t he s am e g e o g ra p h i c a re a .

L im itatio ns : T h e l i m i t i ng f ac t o rs i n us i ng g e n e ra l p o p u l at i o n ra t e s f o r com p ari s on are :

1.Non - av a i l ab i l i t y o f p o p ul a t i o n rat e s f o r t h e o u t c o m e re q ui re d ;
2.T he d i ffi cu l t i e s o f s e l e ct i ng t he s t ud y an d co m p a ri s o n g ro up s w h i c h are re p re s e nt at i ve o f the
exp o s e d no n- exp o s e d s e g m e n t s o f t he g e ne ra l p o p ul at i o n .
04. FOLLOW UP
T he p ro ce d ure s re q u i re d co m p ri s e :

1. Pe ri o d i c m e d i c al ex am i na t i o n o f e ac h m e m b e r o f t he c o h o rt
2. Re v i e w i n g p h y s i ci a n a nd h o s p i t al re c o rd s
3. Ro ut i ne s urv e i l l an ce o f d e a t h re co rd s ,
4. M a i l e d q u e s t i o nn ai re s , t e l e p ho ne c al l s , p e ri o d i c ho m e vi s i t s - p re fe rab l y al l t hre e o n an
annu al b a s i s .

05. ANALYSIS
T he d at a are an al y ze d i n t e rm s o f :

1.Inci d e nc e ra t e s o f o u t c o m e am o n g exp o s e d an d no n- exp o s e d

2.E s t i m a t i o n o f ri s k .
 ANALYSIS - INCIDENCE RATES

In a coh or t st u dy, we can determine i nci dence r ates di rectl y in those exposed and
th ose n ot exposed.

a ) a mong smokers = 70/7000 = 10 per

1000

b) a mon g non smokers = 3/3000 = 1 per

1000

Sta ti sti c a l si g ni fi ca n ce : P< 0. 001

ANALYSIS - ESTIMATION OF RISK

Thi s i s don e i n t erms of t wo wel l -known i ndi ces:

(a) Relativ e Risk : Rel at i ve r i sk (R R ) i s the r ati o of the i ncidence of the disease (or
deat h ) am on g exposed an d t h e i nci dence among non-exposed. Some author s use the
term " r i sk r at i o" t o refer t o rel ative r i sk.
R R = In ci dence of di sease am ong exposed
In ci den ce of di sease amon g non exposed

Rel at i ve r i sk of 1 i n di cat es n o associ at ion; rel ati ve r isk greater than one suggests
"posi t i ve" associ at i on bet ween exposure and the disease under st udy.

A rel at i ve r i sk of 2 i n di cat es t hat the i nci dence r ate of di sease is 2 ti mes hi gher in the
exposed groups compared wi t h t he unexposed.
ANALYSIS - ESTIMATION OF RISK
(b) Attr ibutable r isk : At t r i butable r i sk (AR) is the diff erence in inci dence r ates of
di sease (or deat h ) bet ween an exposed group and non - exposed group. Some author s
use t h e t erm " r i sk di ff eren ce" t o attr i butable r isk.

AR = Inci den ce of di sease r at e among exposed - i nci dence of di sease r at e among non
exposed Inci dence r ate among exposed.

At t r i bu t abl e r i sk i n di cat es t o what ext ent the disease under st udy can be att r ibuted to
th e exposu re.

POPULATION ATTRIBUTABLE RISK

It i s t h e i nci den ce of t h e di sease (or death) i n the total populat ion minus t he incidence
of di sease (or deat h ) amon g those who were not exposed to t he suspected causal
fact or. Th e con cept of populati on attr ibutabl e r i sk i s useful i n that it provides an
est i mat e of t h e amou n t by which the disease could be reduced i n that popul at ion i f the
 TO SUMMARISE:
DIFFERENCES BETWEEN CASE CONTROL STUDIES AND
COHORT STUDIES
 THANK YOU
Sou rce: Par k Text book of Preventi ve and Soci al Medicine, 28t h Edi tion