Cohort Study
The Cohort study design
The important elements of a cohort study
- The exposure has already occurred
- A cohort of people- with and without the exposure- are followed up over a period of time to see who develops the condition(s) of interest
- A cohort study helps in establishing causation
When do we do a cohort study?
We do a cohort study when
- We want to establish a causal link between the exposure of interest and the condition(s) of interest.
- We want to establish multiple outcomes from a single exposure
Examples:
a) Is obesity a cause for gestational diabetes mellitus?
b) What are the maternal and fetal outcomes of tobacco smoking in pregnant women?
STOP. TAKE A MOMENT TO CONSIDER YOUR RESEARCH QUESTION.
IS A COHORT DESIGN THE MOST APPROPRIATE DESIGN?
Is your study focused on determining multiple outcomes associated with an exposure?
Is your study focused on determining what causes a condition of interest?
| ITEM | COMMENTS |
| Define the exposure | The exposure or prognostic variable is the variable of interest |
| How is the exposure defined? | |
| Describe the method(s) of ascertainment of the exposure | Show evidence that these are standard and accepted methods of ascertaining the exposure |
| How often is exposure measured and at what intervals | A hospital, population, screening program etc |
| How are missing values or ascertainments dealt with? | |
| Define the non exposure | Need to show evidence that it is possible to correctly classify non exposure and minimize misclassification |
| How is the non exposure defined? | |
| Describe the method(s) of ascertainment of the non exposure | |
| Establish that exposure and non exposure groups do not overlap | |
| Define study subjects | |
| Describe the population from which the two groups are sourced | The two groups should be selected from similar populations that preferably differ only in the exposure of interest |
| Establish that the subjects in the study do not have the outcome of interest at time of enrollment | |
| Define the variables of interest | |
| How are the variables defined? | Describe all variables- exposure, disease, predictor, outcomes, potential confounders and effect modifiers. Give diagnostic criteria if applicable |
| How are the variables ascertained? | Describe how the variables are measured and ascertained Assessment of outcome is blind to exposure status |
| Who and how many people ascertain the variables? | |
| Describe comparability of assessment methods or assessors if there are more than 1 | |
| Describe standardization procedures for assessment methods | |
| Describe standardization training for assessors | |
| Describe the process of masking | |
| Describe time period of ascertaining each variable | How many times and at what intervals |
| Are the two groups matched? | |
| Describe the variables matched for and justify matching | |
| Describe potential sources of bias | |
| Describe potential sources of bias and how bias is minimized | |
| Describe the study size | |
| Assumptions used to estimate the sample size | |
| The sample size including specific subgroup sample size if any | |
| What is the ratio of exposure to non exposure? | |
| Describe statistical analysis | |
| How are missing data dealt with? | Loss to follow up, lack of all measures at all intervals |
| Describe groupings of quantitative variables with justification | |
| Describe all statistical methods including sub group analysis and interactions, sensitivity analyses, and comparisons between the exposure and non exposure | |
| How are confounders dealt? | |
| Describe how causal links will be established | |
| Mention software package used for analysis | |
| Ethics committee and IRB approval |
