COHORT STUDIES
Cohort (follow-up) studies are observational analytic studies, where group (s) of individuals are defined on the basis of presence or absence of the exposure to a suspected risk factor of a disease, then followed for a period of time to assess the occurrence of the disease provided that they should be FREE from the disease at the start of the exposure . Two main types: 1.Follow-up studies (the prospective form) 2.Retrospective cohort study :Follow-up studies (the prospective form) Constitutes the basic observational strategy for testing hypotheses. In a follow-up study, people without the disease are followed up to see who develop it, and disease incidence in persons with a characteristic is compared with incidence in persons without the characteristic. A "cohort" is a defined group of people who share a common characteristic. e.g. born in certain year, have same exposure to a hazard.
Exposed to a risk factor
Developed an outcomeDidn’t develop an outcome Not Exposed to a risk factor
Developed an outcome
Didn’t develop an outcome Present FutureDiagram of Prospective Cohort Studies
Exposed to a risk factor
Developed an outcomeDidn’t develop an outcome Not Exposed to a risk factor
Developed an outcome
Didn’t develop an outcome 1. Retrospective cohort study : The observer looks backward to the disease & exposure because both of them have happened when the study had started"
past investigator
Diagram of Retrospective Cohort Studies
a+b+c+d
b+d (controls )
a+c (cases)
Total
c+d Non Exposed
d
c
(-)ve
a+b Exposed
b
a
(+)ve
Exposure to a risk factor
Absent
Present
Total
Disease
The two by two table :
We start with 2 groups , one exposed to the factors & the other group not exposed but both groups do not have the disease , then follow them up in time. Group 1 Exposed : ( a+b ) Group 2 Non Exposed : ( c+d )
Analysis :The measure of association between the exposure & the development of the disease is calculated by : Incidence of disease among exposed1. Relative risk (RR) = ـــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــ no unit Incidence of disease among non exposed a / ( a+b) Relative risk (RR) = ــــــــــــــــــــــــــــــــــــ c / (c+d) RR = 1 : No association bet exp. & risk of disease RR > 1 : +ve association ( increased risk among exposed) ( risk factor).RR< 1 : -ve association ( decreased risk among exposed) ( protective factor).. RR estimates the magnitude (strength ) of association between exposure & disease.. it indicates the probability of developing the disease in the exposed related to those unexposed The exposure : risk factor The outcome : disease or death
e.g. 1 In a Cohort study for association between usage of oral contraceptive use (OCP) & bacteruria, the following table was formed:
2390
2286
104
Total
1908
1831
77
No
482
455
27
Yes
OCP
No
Yes
Total
Bacteruria
a / ( a+b) 27/482 Relative risk (RR) = ــــــــــــــــــــــــــــ = ـــــــــــــــــــــ = 1.4 c / (c+d) 77/1908 This means that those who are OCP users have risk a 1.4 times the risk to develop the outcome than those no exposed.
e.g.2 suppose the incidence of Hepatitis B sero (+)ve among those having previous blood transfusion is 5/1000 / year & those with no blood transfusion is 1/1000/y so: 5/1000/year RR= ـــــــــــــــــــــــــــــــــــ = 5 times of developing an outcome among exposed 1/1000/year compared to the non- exposed )
2. Attributable Risk (AR) :AR= I exposed minus I non exposed AR = {a / ( a+b)} - {c / (c+d)} ……. unit Also called the risk difference .provides information about the absolute effect of the exposure i.e. the excess risk of dis. among the exposed compared to the non exposed . Now look at the bacteruria –OC table AR = Ie - Ie‾ = 56.02/1000 per year - 40.36/1000 per year=15.66/1000 per yearNote :"AR is only calculated from cohort studies " & cannot be calculated from case-control studies .
Back to hepatitis example so : AR = 5/1000 hepatitis per year -1/1000 hepatitis per year = 4/1000 hepatitis per year ( absolute measure , effect of the exposure). . It quantifies the excess of risk of disease in the exposed group which is attributable to the exposure. . AR is useful as a measure of public health impact of a particular exposure assuming a causal effect f the exposure on the outcome .
140 RR ( lung Ca) = ـــــــــــــــــ = 14 times 10 669 RR ( CHD) = ـــــــــــــــــ = 1.6 times 413So 14.0 : A person who smokes will have a 14.0 times chance to die from lung Ca than a non- smoker . And : 1.6 : 1.6 times chance to die from CHD than a non- smoker Annual Mortality Rate per 100,000
256/105/y
130/105/y
AR
1.6
14.0
RR
413
10
Non-smokers
669
140
Cig. Smokers
CHD ( coronary H. disease)
Lung Ca
3. Attributable Risk Percent (AR%) : Attributable RiskAttributable Risk % = ـــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــ x 100 Incidence among exposedEstimates % of gain, if the factor is removed from population. Gives an idea about the proportion of the disease in the exposed that could be prevented by eliminating the exposure .e.g. of cohort the bacteruria –OC 15.66/1000/year AR% = ــــــــــــــــــــــــــــــــــــــــــــــــ x 100 = 27.95% 56.02/1000/yearIt estimates the proportion of the disease among the exposed that is attributed to the exposure.
Incidence
I among I among non exposed exposedA R%
A R
Strengths -Advantages-of cohort studies: 1. Establish the temporal relationship between disease. i.e. the time sequence between the exposure & the outcome & it is important in determining the causal outcome . 2. determines the risk of getting the disease through the exposure to a factor. 3. Useful for rare exposure. E.g. : Chemical & Radio active exposure is best studied through Cohort. 4. Examines multiple effects of a single exposure. E.g. People exposed to Asbestos & follow them up to develop Lung Ca, pulmonary fibrosis & other effects of the exposure 5. Allows direct measurement of the incidence of the disease among exposed & non- exposed groups.
Limitations: 1.Expensive : Personnel & Finance. 2. Time Consuming : due to the follow up e.g. Framingham Study which started in the 50s studying the exposure of certain factors to development of various heart disease. 3. Problems to follow up : die, run away, disappear ,.etc. 4. Of limited use in rare disease.
Sources of Exposure Data 1. Pre-existing records. 2. Information from the study subjects -interview. 3. Direct Physical examination or an investigation. 4. Direct measurement of the environment. e.g. detection of the exact level of a certain thing in the environment as noise by a sound level meter. Selection of Comparison Group They should be similar to the study group in all the factors related to the disease, except the factor under study. Sources of Outcome Data Fatal Outcome : - death certificates. Non-Fatal Outcome : - Medical Records. - Direct from the participants. - Data from periodic M Exam. The method of assessment of outcome should be the same for both groups.
e.g. "A" a hypothetical cohort study of cigarette Smoking & lung Ca (100exposed, 100nonexposed)
200
100
100
Total
100
70
30
No
100
30
70
Yes
Cigarette Smoking
Control
Case
Total
lung Ca
a / (a+b) 70 / 100
RR = = = 2.3 c/ (c+d ) 30 / 100
Now : same e.g. "B“370 exposed 730 non exposed 1100
1000
100
Total
730
700
30
No
370
300
70
Yes
Cigarette Smoking
Control
Case
Total
lung Ca