Study Designs and Methodology Flashcards
What are the 2 kinds of studies?
- interventional studies
- Observational studies
Quantitative?
-Numbers used to represent data
Interventional
- forced group-allocation
- the researchers are intervening!
Observational
- No forced group allocation
- we just observe what’s going on
What will the right answers be on interventional?
-Phase 0, 1, 2, 3, 4
What will the right answers on observational?
- Cross sectional
- Case control
- Cohort
What is that pyramid for increasing strength of evidence?
- the more we control things, the stronger it is
- so animal and in vitro research are at the bottom of the pyramid
- systematic reviews, meta analyses, and interventional trials are at the top
If a study is prospective, what does that mean?
-outcome is NOT yet know at the start of the study
Retrospective
-outcome IS already known at the start of the study
Ambidirectional (both perspectives in the same study
-first looking retrospectively, then looking prospectively for additional outcome occurrences
Which studies can be prospective?
- all interventional phases
- The cohort one for observational
What is the null hypothesis
-there will be no difference between the groups being compared
What are the 2 key questions to selecting a correct study design?
- Is researcher forcing group allocation
- For observational studies, how were groups ORGANIZED
If the research is forcing group allocation, what kind of study will we have?
- an interventional kind
- if the answer is no, then it’s observational
If we see the word “randomized”, what do we think of?
-interventional studies
in observational studies, how can the groups be organized?
- by disease status
- by exposure status… cohort
- together due to a common factor… cohort
- data collected across large pop…. cross sectional
What does a case control do?
- breaks up ppl into groups of ppl who have the disease and ppl who ain’t got the disease
- this is screaming case control study
What is retrospective?
-when the outcome of a disease is already known
In case control studies, what are we looking for?
- exposure to something else and relate that to the disease
- especially when the disease is really rare… makes the study easy because they will all be at a center somewhere
What do we know at the start of a caes control study?
-how many people have the disease (the veritcal columns)
What is a nested case-control
-a case control study derived from within or out of a cohort or interventional study
In a cohort study, how do you separate everybody?
-put people in a group who were on the drug, not on the drug in another group…. but NOT FORCED, they just happen to be on the drug
When is cohort design useful?
- when studying a rare exposure
- commonly generates the risk of disease/outcome for each and a Risk Ratio/Relative risk as measure of association
What is a cohort?
- a group with “something” in common
- or it could be based on exposure status
What is a cross-sectional study?
- examines relationships between disease AND exposure among individuals ina defined study population at a point in time
- a “snap-shot” in time (across the entire study population
When should we think of a cross sectional study?
-when the information gathered represents what is going on with disease AND exposures
Are all national studies cross sectional?
- yes
- *so, if the company or whatever entity is that is doing the study starts with an “N”, then it is cross sectional!!!
If we are studying healthy people, what are the options for an interventional study?
-phase 0 or 1
Phase 0
- assess drug target actions and PK’s in non-therapeutic
- “does the drug mechanistically do what we think it does”?
Phase 1
- Assess the safety/tolerance and PK’s of one or more dosages
- healthy ppl
- “does the drug work?”
What 4 things are looking for in the question stem to figure out which phase we are in?
- Purpose
- population studie
- sample size
- Duration
Phase 2
- assess effectiveness
- diseased volunteers
- larger N
- Short to medium duration
Phase 3
- efficacy is paramount
- assess effectiveness
- diseased volunteers
- larger N
- Longer duration
- last phase the drug is at to try to get it approved
Phase 4
- after they are approved and on the market
- assess long term safety, effectiveness, optimal use
- diseased volunteers
- population N
- Longer duration
What is the deal with simple studies?
- there is only 1 randomization step
- can have as many groups as it needs
- but there will be only 1 randomization process!!!
what are factorial studies
- randomized into an initial group, then further randomly divided into a sub group
- multiple randomizations
What does parallel mean?
- subjects exclusively managed in a single treatment group
- NO switching groups after initial randomization
- no crossing over on the picture
Cross-over, what does it mean?
- the subjects forcibly switched to other treatment group after initial treatment assignment
- they will always have a picture where the lines cross
What do we need to do before we make the groups switch in a cross over study?
-a 2 week washout period…
Which studies are better at proving causation?
-prospective studies
What is the most common kind of bias?
-selection bias: “who can be in this study”?
In group allocation procedures, what does random mean?
- subjects have equal probability of being assigned to each of the pre-defined intervention groups
- random number generator program/software/application
What does non random mean?
-some investigator-developed way to select patients
What is the purpose of randomization?
- to make all groups as equal as possible; based on known confounders
- *table 1 needs to have high P values
What is simple randomization
-equal probability for allocation into one of the study groups
What is blocked randomization?
- Ensures balance within each intervention group
- when researchers want to assure that all groups are equal in size
What is stratified randomization?
-ensures balance within KNOWN CONFOUNDING VARIABLES
What does single blind mean?
-study subjects are not informed which intervention group they are in, yet investigators are permitted to know
what is double blind?
- neither investigators nor study subjects are informed which intervention group subjects are in
- post study survey’s can be used to assess adequacy of blinding
What is open-label (unmasked/unblinded)?
-study subjects and researchers know what intervention is being receeived
What can we do when they quit or die or drop out of the study for whatever reason?
- include them: intent to treat
- Exclude them: per protocol
What does intent to treat result in?
- preserves the randomization process
- preserves baseline characteristics and group balance at baseline which controls for known and unknown confounders
- maintains statistical power
What does per protocol rsult in?
- biases estimates of effect
- can limit generalizability!!!! (ultrra adherent)
What is the IRB?
-protects the study subjects before the study even gets started
What does the DSMB do?
- protects the subjects after the study starts
- they will shut it down if bad things are happening
