Lecture 22-23: Cohort Studies Flashcards
1
Q
Describe the Cohort Study Design
A
- Cohort studies are OBSERVATIONAL studies allowing researcher to be a passive observer of natural events occurring in naturally-EXPOSED and UNEXPOSED (comparison) groups
- Group-allocation based on EXPOSURE-STATUS OR Group MEMBERSHIP (SOMETHING IN COMMON)
- Useful when studying a rare exposure
- Cohort studies also termed: Incidence studies/Follow-up studies/Longitudinal studies
- COHORT design useful when studying a RARE EXPOSURE Commonly generates the RISK OF DISEASE/OUTCOME for each, then a Risk Ratio/Relative Risk (RR) as measure of association
2
Q
For what reasons would a cohort study design be appropriate?
A
- Unable to force group allocation (‘randomize’)
- Unethical / Not feasible
- Limited resources
- Time / Money / Subjects
- The EXPOSURE OF INTEREST is rare in occurrence and little is known about its associations/outcomes
- More interested in incidence rates or risks for outcome of interest (more than effects of interventions)
- Can be conducted in a PROSPECTIVE, RETROSPECTIVE (OR HISTORICAL), OR AMBIDIRECTIONAL fashion
- Group assignment is STILL based on EXPOSURE!
3
Q
Describe Prospective Cohort Studies
A
Prospective Cohort Studies:
- Exposure group is selected on the basis of a past or current exposure and both groups (exposure and non-exposure) followed into the future to assess for outcome(s) of interest (which has yet to occur), and then compared
4
Q
Describe Retrospective Cohort Study Designs
Also Called Historical
A
- At the start of the study, both the exposure and the outcome of interest have already occurred
- Retrospectively start at time of exposure (historically) and follow forward to the point of outcome occurrence (known), in the present
- Exposure still has to occur Before outcome of interest and group allocation is based on exposure status, not disease status
5
Q
Describe Ambidirectional Cohort Design
A
- Uses Retrospective design to assess past differences but adds all data collected on additional outcomes Prospectively from start of study
- Looking for outcomes in the past and into the future
6
Q
Describe what a Cohort Is
And what a birth cohort is
A
- A Cohort is also refers to a group with something in common:
- Example “cohorts”:
- -Birth cohort: Individuals assembled based on being born in a geographic region in a given time period. Or Everyone born in KC city limits in 2014
7
Q
Describe an Inception Cohort
A
Cohorts can also be groups with something in common. Like Inception Cohorts.
- Individuals assembled at a given point based on some common factor
- Where people live or where they work, or something they have in common
- Useful for single-group assessments for incidence rate determination. Such as A single health-care system. Or a single payer of health-care coverage
- There was some famous one called in Framingham in 1948. It was selected on being a stable population with updated annual population lists, and other unique attributes… Can also be a wide spectrum of occupations or a single hospital. I know these notes suck, but I don’t get where they are? Do I really suck that hard? NO MORE ABSENCES FROM CLASS DAMNIT!
8
Q
Describe Exposure Cohorts
A
- Individuals assembled based on some common exposure
- Frequency connected to environmental or other one-time events
9
Q
Describe how the size of cohort studies can change over time
A
- Fixed Cohort: A cohort (derived from an irrevocable event) which can’t gain members but CAN have loss-to-follow-ups
- Closed Cohort: A Fixed cohort with NO loss-to-follow-ups
- Open (or Dynamic) Cohort: A cohort with new additions and some loss-to-follow-ups
- Cohorts can increase or decrease over time
10
Q
Describe how to select an Exposed study population
A
- This is the easier part (selection of Unexposed is harder!)
- Allocate subjects based on pre-defined criteria of “exposure”
- Scientifically and consistently determined
11
Q
Describe how to select an Unexposed study group (harder than exposed)
A
- Make the groups as close as possible (coming from the same cohort/population (yet not exposed))
- If exposure truly has no effect, then risk will be exactly the same for both groups and RR will be 1.0 (no difference)
- This unexposed group can come from 3 sources:
1. Internal
2. General
3. Comparison
12
Q
Describe the unexposed study population, Internal
A
- Patients from the same “cohort”, yet who are unexposed (most similar)
- If there are only levels of exposure, you may have to use the lowest exposure group as comparator (if there is no “no” exposure group internally-available)
13
Q
Describe the unexposed study population, General
A
- General Population
- Used as a second choice when the best-possible comparison group (internal) is not realistically possible (e.g., everyone is exposed; or the exposure subjects were drawn from the general population)
14
Q
Describe the unexposed cohort study group source, Comparison Cohorts
A
- This is the least acceptable group (but still can be utilized)
- Simply attempt to match groups as close as possible on numerous personal characteristics (can’t control for other potentially harmful exposures in comparison cohort; also causing disease)
15
Q
Describe how to measure the strength of a cohort study
What are some weaknesses of a cohort study?
A
- Strengths of Cohort Studies (IN GENERAL; compared to other study designs):
- Good for assessing Multiple Outcomes of ONE exposure
- Hard to control for other exposures if more than one plausible for being associated with an outcome
- Useful when EXPOSURES are rare
- Useful in calculating risk and RR’s
- Less expensive than interventional trials
- Good when ethical issues limit use of interventional Good for long Induction/Latent periods (Retrospective)
- Weaknesses: Weaknesses of Cohort studies may be the opposite of these general points listed above
- Able to represent “Temporality” (Prospective)
16
Q
What are some advantages of Prospective Cohort Studies?
A
- Can obtain a greater amount of study-important information from patients
- More control over specific data collection process
- Interviews / Laboratory assessments / Physical examinations
- Follow-up/Tracking of patients may be easier
- IF you plan ahead!
- Better at giving answer to “Temporality”
- Simple Association vs. Causal Inference (Hill’s criteria)
- May look at multiple outcomes from a (supposed) single exposure
- Can calculate Incidence & Incidence rates
17
Q
What are some disadvantages of Prospective Cohort Studies
A
- Time, Expense & Lost-to-follow-up’s
- Not efficient for rare diseases
- use Case-Control study in this situation
- Not suited for long Induction/Latency conditions
- Exposure (or its “amount”) may change over time
18
Q
Discuss his rule on Loss to Follow-Ups
A
- Possible with Prospective Cohorts
- Lowers Sample Size (Power) (just as it does with Interventional studies!)
- Increased risk of Type 2 error (more to come in future lecture)
- Loss of study participation (follow-up) may not be equal between groups
- Authors MUST list LTFU’s by group (exposed/un-exposed)
- Point Being: Do ALL you can (and think of) to limit LTFU’s! [lots of time, energy, and resources]
19
Q
What are advantages of Retrospective Cohort Studies
A
- Best for long Induction/Latency conditions
- Able to study rare exposures
- Useful if the data already exists
- Saves time and money compared to Prospective studies
20
Q
What are some Disadvantages of Retrospective Studies
A
- Requires access to charts, databases, employment records (may not be Complete/Thorough Enough for study)
- “Information” may not factor in or control for other exposures to harmful elements
- Patients may not be available for interview if contact necessary for missing or incomplete data
- Exposure (or its “amount”) may have changed over time
21
Q
Describe Matching in a Cohort Study Design
A
- A way to strive to make groups as equal as possible on known/potential confounders
- Can match on a 1:1 or even higher (1:5) ratio [exposed to unexposed]
22
Q
What are two key biases associated with Cohort Study Designs
A
- Healthy-worker effect
- If healthy, you work (even if exposed). If too ill to work (due to exposure?) you may be unemployed (now part of non-working general population)
- Selection bias
- How exposure status is defined/determined (less of an issue with exposure status)
