COHORT STUDIES COHORT STUDY Environmental Exposure Association Disease

COHORT STUDIES

COHORT STUDY Environmental Exposure Association Disease or Other Outcome 2

DESIGEN OF A COHORT STUDY EXPOSED NOT EXPOSED 3

DESIGEN OF A COHORT STUDY EXPOSED DISEASE NOT EXPOSED DISEASE NO DISEASE 4

Time-Frame for a Hypothetical Concurrent (Prospective) Cohort Study Begun in 2010 Defined Population Prospective 2010 NON- RANDOMIZED 2020 2030 EXPOSED DISEASE NOT EXPOSED DISEASE NO DISEASE 5

Time-Frame for a Hypothetical Retrospective Cohort Study Begun in 2010 DISEASE NO DISEASE EXPOSED DISEASE NOT EXPOSED NO DISEASE Retrospective 1990 2000 Non - RANDOMIZED 2010 Defined Population 6

Time-Frames for a Hypothetical prospective and a retrospective Cohort Study Begun in 2010 Defined Population Prospective 2010 Retrospective 1990 NON- RANDOMIZED 2020 2028 EXPOSED DISEASE NOT EXPOSED DISEASE NO DISEASE 2000 2010 7

WHY DO COHORT STUDY • Temporal sequence between exposure and disease can be clearly established- because all are disease free at the start • Possible to assess the rare exposures- Nipah virus • Multiple effects of a single exposure- smoking can cause lung cancer, HTN, myocardial infarction etc. 8

Which one we need to select? Look carefully and tell me what is the difference and / or similarities the following two option of Cohort study design 9

OPTION – I THE DESIGEN OF A COHORT STUDY Start with: EXPOSED Then, follow up, for: DISEASE NO DISEASE AND DISEASE NOT EXPOSED NO DISEASE 10

OPTION – II THE DESIGEN OF A COHORT STUDY Start with: DEFINED POPULATION Then, Follow up for: NOT RANDOMLY ASSIGNED EXPOSED DISEASE NOT EXPOSED DISEASE NO DISEASE 11

Example : Option I Atomic Bomb Blast at Japan or Impact of Tannery industry west in Hazaribagh population Example : Option II Children with Exclusive Breast feeding VS Non exclusive in Dhaka A group with sugary drink intake VS No sugary drink intake in BD

Whatever may be the option, what we usually found that is……. 13

DESIGEN OF A COHORT STUDY OF SILICONE BREAST IMPLANTS AND CONNECTIVE TISSUE DISEASE (CT) BREAST IMPLANTS CT DISEASE NO CT DISEASE 14

A COHORT STUDY OF SILICONE BREAST IMPLANTS AND CONNECTIVE TISSUE DISEASE (CT) – IF THERE IS AN ASSOCIATION, WE WOULD EXPECT TO FIND: BREAST IMPLANTS NO CT DISEASE NO IMPLANTS CT DISEASE NO CT DISEASE 15

Types of Potential Bias in Cohort Studies • Selection bias – Who has been exposed and who has not • Bias in assessment of outcome • Non-response bias, lost to follow up • Analytic bias • Information bias – Particularly in retrospective studies 16

COMMON EXPOSURE • Cigarette smoking or coffee drinking It is possible to get a large number of population exposed to these. One can identify the incidence of lung cancer in this population easily 17

RARE/UNCOMMON EXPOSURE • Alcohol consumption / Specific addicting agent is rare in a conservative society. But one can follow a small cohort and subsequently generalize the exposure and outcome relation (liver cirrhosis) for the general population. 18

RARE/UNCOMMON EXPOSURE • This rare/uncommon exposure allows one to identify a rare outcome. • Possibly this outcome is rare in general population but it may be sufficiently common in special exposure population. 19

Cohort Studies – Examples Starting with Exposed and Non-Exposed Individuals 1. Atomic Bomb Casualty commission (ABCC) (Hiroshima and Nagasaki) Starting without (or before) any specific exposure 1. Framingham studies – I, II and III 20

The Framingham Study Objectives To study the impact of several factors on incidence of CVD Several Exposures (Smoking, diet, etc. ) Multiple Outcomes (Coronary heart diseases, stroke, congestive heart failure, peripheral arterial disease)

Framingham Study ORIGINAL COHORT Begin in 1948 5127 men and women ages 30 to 62 Return every 2 yrs. OFFSPRING COHORT 1971 - 2 ND Generation group enrolled 5124 of the original participants’ adult children for similar examination GENERATION III COHORT Currently being recruited. Goal is 3500 grandchildren of original cohort; How genetic factors relate to CV disease

IMORTANT ISSUES (COMPARISON GROUP) • Important and difficult • The principle is that comparison group is Similar in all respect except the determinants in question 23

IMORTANT ISSUES (COMPARISON GROUP) Why? If there is no association between the exposure and disease, the disease rates in the comparison group essentially be same. It is important that information obtained from the non exposed group are adequate to compare with the exposed group. 24

TAKE HOME MESSAGE When is a Cohort Study Warranted? 1. When there is a good evidence of an association of the disease with a certain exposure 2. When exposure is rare, but incidence of disease among exposure is high 25

TAKE HOME MESSAGE When is a Cohort Study Warranted? 3. When the time between exposure and disease is short 4. When attrition of study population can be minimized 5. When ample funds are available 6. When the investigator has a long life expectancy 26

TAKE HOME MESSAGE: Cohort Study Limitations: 1. Inefficient for evaluation of rare disease, unless attributable risk % is high 2. If prospective it is time consuming and expensive 3. If retrospective requires the availability of adequate records 4. Validity of the results can be seriously affected by losses to follow-up 27

Figure it out Some “Hot” Hypotheses • Does diet affect breast cancer risk? • Does passive smoking cause lung cancer? • Do breast implants cause connective tissue disease? • Do cell phones cause brain cancer? 28

Thanks 29