Lecture 6, part 3 Flashcards
Reasons for non-compliance?
Toxic reactions to the tx
Waning interest, desire to seek other therapies
Inability to meet the demands of the study
Dz progression, co-morbidities, death
Methods of enhancing compliance
Keep tx regimens simple and easy to follow
Enroll motivated and knowledgeable participants with fairly organized lives
Provide realistic picture of tasks involved during the informed consent process
Obtain detailed medical hx and reject ppl who seem unlikely to comply
Mask subjects to tx assignment
Maintain frequent contact
Conduct run-in or lead-in period
Offer incentives
Primary outcomes should be selected to be what?
Easy to diagnose or observe and clinically relevant
Free of measurement or ascertainment error
Capable of being observed independent of tx assignment
What aspects of the outcome can be measured?
Outcomes can be surrogates for actual phenomena of interest
Measure potential adverse effects
Measurement of outcomes should have what characteristics?
Accurate and precise
Objective rather than subjective
Masking definition
A tool used to create comparability with respect to post-randomization observations
What is the goal of masking?
Prevent biased ascertainment of the outcome measure
Types of masking
Single
Double
Triple
Single masking
Participant doesn’t know tx assignment
Double masking
Participant and investigators making the tx assignments are unaware of tx assignment
Triple masking
Participant, investigators making the assignment, and investigator measuring the outcome are all unaware of tx assignment
Aspects on ongoing monitoring
Need to monitor continuously
Evaluate safety/unanticipated effects at baseline and pre-specified periods throughout study
Findings (positive or negative) may necessitate early termination of the trial
Aspects of intent-to-treat analysis
All individuals randomly allocated to a tx are analyzed as receiving that tx, regardless of whether they actually complete or receive the tx
Preserves baseline comparability of the groups for known and unknown confounders
Maintains the statistical power of original study pop
Unbiased results
Gives info on effectiveness of a tx under nl practice conditions
Aspects of efficacy analysis
Determines the tx effects under ideal conditions if adherence to tx is perfect
Analyze the outcomes of those who complete tx in each group (ignore outcomes of those who failed to complete assigned tx)
Potential for bias as compared groups may no longer be similar (lose the benefits of randomization in terms of comparability)
How to create comparability
With the respect to the tx experience: use of placebos, similar tx regimens
With respect to the participants: heterogeneity, randomization
With respect to post-randomization observations: masking or blinding of participants and researchers
Issues with experimental studies
Exclusive study pop criteria limits generalizability of results
Focus on testing outcomes under ideal situations measures efficacy of the tx/intervention, not actual effectiveness
Reasons to do randomized control trials
Arguably the only study design that can “prove” causation
Most influential to clinical practice
Provide the strongest evidence in “evidence-based” medicine
Required by FDA and other agencies for new drugs and some new devices
Often required by third party payers
Best for studying therapeutic interventions
Reasons not to do randomized control trials
Expensive and difficult to fund - typically cost $100s of millions and take yrs to perform
Limited in scope- typically only answer a single question
Limited in generalizability - may not apply to most pts in practice
Logistically complex- usually involve multiple different clinical centers
Ethical considerations/constraints
Practical issues to consider with experimental studies
Randomization to reduce bias
Recruitment of willing/eligible participants
Generalizability of participants/results
Participant noncompliance
Participant attrition
High costs
How many clinical trial phases are there?
4
Lab studies- definition and timing
Pre-clinical studies conducted in laboratories or with animal models
Usually take several yrs
Details about phase I trials
Clinical pharmacological studies
Small sample of 20-80 participants (usually healthy)
Used to assess safety issues related to new drug or tx
-Safety and safe human dosage ranges (tolerable limits), toxicity and side effects
Details about phase II trials
Clinical investigations of 100-300 pts
Evaluates the efficacy of the new drug or tx
Further assess relative safety
Details about phase III trials
Large-scale randomized controlled trials of 1,000-3,000+ pts
Assess effectiveness and relative safety
Usually multi-center recruitment
Once a tx/drug passes this phase, it can be approved and licensed for marketing
Details about phase IV studies
Monitor long-term (over several yrs) safety and effectiveness of new drugs/txs as they come into general use by the public
-AKA post-marketing surveillance
Not randomized studies, but can examine the effects of the drug in pops or situations not studied before
Rationale of phase IV studies
Certain adverse effects may not become manifest for many yrs, or may be so infrequent that they may not be detectable even in relatively long RCTs
