Jonathon Hopkins Course Module 3 Flashcards
Gold standard for analysis of clinical trials?
including all the participants in the analysis regardless of their actual treatment.
synonimus for outcome?
endpoint but we encourage people to use term outcome couse patients can be followed after the endpoint.
definition of outcome
Usually, we try to use the definition of the outcome to make what may be a qualitative thing a quantitative thing
what outcome reflects?
objectives of trial:
- efficacy
- safety
- process
- costs
primary outcomes?
reflects objectives and primary hypothesis
design variable
related to stage/type of research
So, what I mean by that is in, in preliminary phase one studies, the outcome may be a safety Related at any adverse events. In later stages of research the outcome may be a blood level, such as cholesterol, and then in another stage it may actually be a clinical outcome.
secondary outcomes?
other important potential treatments effects:
- defined safety outcomes
- mechanism of effect
other outcomes?
patient compliance
exploratory
criteria for primary outcome?
- It should be set before looking at trial data
- relevant and likely to be influenced by a tretment
- reliable measurment (proven outcome)
- power considerations (variability, frequency,..)
possible outcomes for asthma treatment
- exhaled NO
- spirometry
- asthma symptoms
- composite measures (exacebaration, symptom index)
outcomes in perioperative procedure
- time window
- specific events to be considered as an outcome
- procedures to establish outcomes (So are you going to be interviewing patients?)
Metrics for events as a outcomes
Dichotomous:
clinical state or cut-off value (hemoglobin value)
at a specific time point
Time to event:
adds dimension of time to dichotomous
allows for censoring
more powerful than dichotomous
Metrics for events as a outcomes part 2
Rates:
- 1/0 but allows for repeats
- need follow up time
- Analize count or rate (events within a person is usually not independent)
Composite measures:
- two or more events related to disease process
- couldbe considered dichotomous or time-to event or a rate
Metrics for events as a outcomespart 3
Continuous variables (cholesterol example):
- outcome is a value or chane from a baseline
- standard units, lab values, score
- need to define a important diference
- repeated measurments possible
- typically more powerful than discrete outcomes
- dichotomos require more patients
Ordinal scale (subcategory from continuous)
- adverse event grading for example
- ranked categories 1-5 A-D
- difference between categories is usually qualitative
Objective vs Subjective outcome
Objective:
- Rigorous interpetation to limit interpretation
- may include test results for confirmation
Subjective: karnofsky score(1-10), histological evaluations, need for medications -masking is more important
How to enhance accuracy and objectivity
defined criteria for evaluation
training evaluators
Hawthorne effect? and how to quantify subjective outcomes
The Hawthorne effect is a type of reactivity in which individuals modify an aspect of their behavior in response to their awareness of being observed.
To quantify we use scales
Difference between efficacy and effectiveness?
efficacy trials are usually those that are looking to be able to evaluate the condition under the best of all circumstances. Does this intervention change outcomes given sort of every benefit, versus effectiveness, which is looking how the intervention will work in a more real world setting.
efficacy and effectiveness- 2 example?
vaccine trial:
effectiveness-clinical case of influenza
efficacy-clinical case with laboratory conformation
asthma:
effectiveness-hospitralizations/steroid courses
efficacy- FEV1
factors that could determine choice of an outcome?
prior research key factor of design (eligibility criteria, sample size, duration and frequency of follow up, need for masking) personell and resources required QA procedures Costs Generezibility of results
Three Bs to determine outcome
Biology , Budzet and Biostatistics
Example for outcomes on antiretroviral rherapy
1 survival 2 immunological respose 3 virologic response 4-QoL 5 toxicity 6-side effects ... If we want to look fro a Phase 4 we would look at points 1,4,5,6
How to protect a outcome?
defining it before the trial masking randomization standard methods for measurments standard follow up scheadules
Analysis issues to be covered?
A: Analysis by asigned treatment (intention to treat)
B: Subgroup analysis
Randomized to treatment but not take that, example?
NETT, surgical and medication group, but some participants refused
What should we do when we have those unplaned crossovers?
analize date base purely on randomization and ignore: ineligebility, complete nonadherence, partial adherence, tretment termination, treatment switches.
Violation of ITT (intention to treat)
Any exclusio nafter randomization or adjustment for anything occuring after baseline has potential to introduce bias and make observational instead of experimental trial
Nonadherence in not random
ITT properties
ITT is unbiased
ITT measures effect in global sense
ITT are not post hoc (adherence is something that’s hard to quantify before you’ve seen the data. So to include adherence in our primary analysis plan would be difficult to specify ahead of time.)
ITT data are collected regardles of adherence
Why adjusting for adherence is not goood idea?
since randomization has already given us comparable groups, by adjusting to adherence, we may then introduce non-comparability.
what is subgroup analysis?
In the context of clinical trials subgroup analyses are analyses where we’re looking to see if there are varying treatment effects in different subsets of patients.
why do we want subgruop analysis?
we’d like to check the consistency of the treatment effect across subgroups. For instance, we’d like to know if the effect is the same in women and men or in children and adults. Or we might think that perhaps the treatment effect would vary for people with less severe disease versus those with more severe disease.
2 most common methods of doing subgroup analysis
stratified analysis and test for interaction
stratification subgroups?
To do a subgroup analysis by stratification, you estimate the treatment effect separately in each of the subgroups. So you estimate an effect in men and you estimate an effect in women. And using this method, you can test whether there is a significant effect in men, and you can test if there’s a significant effect in women.
But what you cannot do is test whether the effect in men differs from the effect in women.
test for interaction - subgroups
So, to do this test we build a statistical model that includes main effects for our treatment group and for our subgroups and we use interaction terms to test for interactions between treatments in subgroup. And in this way, we can see if the treatment effects vary by subgroup. We can also use that model to estimate the treatment effects in the various subgroups like we could do with the stratification method.
significance test in subgroup analysis?
the more subgroup analyses that we do, the more likely it is by chance alone that we will find a statistically significant difference.
guidlines for subgroup analysis?
prespecification of subgroupo analysis (inflate a sample size to plan for subgroup analysis)
reporting a number of subgroup analysis peformed
adjustment for multiple comparisons
reporting confidence intervasl instead of or in addition to p-values
