Clinical trials Flashcards
Define what a clinical trial is
- Form of planned experiment
- Involves patients and is designed to elucidate most appropriate method of treatment for future patients with a given medical condition
Define efficacy in the context of a clinical trial
- Ability of a health care intervention to improve the health of a defined group under specific conditions
Define safety in the context of a clinical trial
- Ability of a health care intervention not to harm a defined group under specific conditions
What are the different phases of a clinical trial?
- Phase I: <100 healthy volunteers - checks for pharmacodynamics, pharmacokinetics, major side effects
- Phase 2: <1000 patients - checks for effects and dosages, common side effects
- Phase 3: <10 000 patients - comparison with standard treatments
What are the most important ethical considerations for any clinical trial to go ahead?
- Trials of new drugs may do harm
- You should only conduct a trial if you are in clinical equipoise
- Patient/participants must understand what participation in the trial involves
Outline what equipoise means
- If there are only a few choices for a drug then it is more ethical to do a trial
- Trial will benefit patients by providing them with a treatment for their condition when few/no treatments currently exist
What do non-randomised clinical trials involve?
- Allocation of patients receiving a new treatment to compare with a group of patients receiving standard treatment
How are patients allocated in a non-randomised clinical trial?
- Allocation may be by:
- Site
- Historical controls
- Simple system devised by investigator
What can allocation lead to in non-randomised clinical trials?
- Allocation bias by patient, clinician, or investigator
- Confounding - known and unknown
What does comparison with historical controls involve?
- Comparison of a group of patients who had standard treatment with a group of patients receiving new treatment
What are the problems with using historical controls in clinical trials?
- Selection often less well defined, less rigorous
- Treated differently from new treatment group
- There may be less information about potential bias/confounders
- Unable to control for confounders
Why is randomisation ideal?
- Eliminates allocation bias, giving each participant an equal chance of being allocated to each of the treatments in the trial
- Minimal confounding, leading to treatment groups that are likely to be similar in size and characteristics by chance
What is a confounder?
- A factor associated with the exposure and is independently a risk factor for the disease
- This association is separate to the relationship between risk factor being investigated
How should randomisation be carried out?
- Should use concealed allocation so there is no possibility of predicting allocation of next patient
- Randomisation sequence should be prepared by someone who is not entering patients and allocation should be at a distance
- E.g. use a third party
How do we do randomisation?
- Normally use a 3rd party, computer generated random allocation, accessed by phone/internet
What does a clinical trial need to be in order to give a fair comparison of effect and safety?
- Reproducible - in experimental conditions
- Controlled - comparison of interventions
- Fair - unbiased without confounding
- Differences observed in small trials are more prone to chance
How might knowledge of which participant is receiving which treatment bias the results of a clinical trial?
- Patient may alter behaviour, other treatment, or even expectation of outcome (behaviour effect)
- Clinician may alter their treatment, care and interest om patient (non-treatment effect)
- Investigator may alter their approach when making measurements and assessing outcomes (measurement bias)
What is a single blind trial?
- Either patient, clinician, or assessor does not know treatment allocation (usually patient)
What is a double blind trial?
- Two of patient, clinician or assessor does not know treatment allocation
- Usually patient + assessor/clinician
Give examples of how blinding/masking might be carried out?
- Aim to make treatments appear identical in every way e.g. appearance, taste, texture, dosage regimen, warnings
- Use a designated pharmacy to label identical containers for the treatments with code numbers to differentiate between them
When might blinding be difficult?
- Surgical procedures
- Psychotherapy vs antidepressants
- Alternative medicine vs Western medicine
- Lifestyle interventions
- Prevention programmes
What does randomisation eliminate in a trial?
- Confounding
What does good randomisation eliminate in a trial?
- Selection bias
What does blinding eliminate in a trial?
- Outcome measurement bias
How do we determine who should or shouldn’t take part in a clinical trial?
- Inclusion criteria
- Exclusion criteria
- E.g. age, sex, health
Why do we pre-define outcomes in clinical trials?
- Prevent data dredging, repeated analyses
- Have a clear protocol for data collection
- Agreed criteria for measurement and assessment of outcomes
What is meant by pre-defining outcomes in clinical trials?
- Need to define what, when and how outcomes are to be measured before the start of the clinical trial
Define primary and secondary outcomes
- Primary outcome: preferably only one primary outcome, used in sample size calculation
- Secondary outcomes: other outcomes of interest, often includes occurrence of side-effects
What are the types of outcomes in a clinical trial?
- Pathophysiological e.g. tumour size, thyroxine level, other biomarkers
- Clinically defined e.g. death, disease, disability
- Patient focused e.g. quality of life, psychological well-being, social well-being, satisfaction
What are the features of an ideal outcome?
- Appropriate and relevant
- Valid and attributable - any observed effect can be reasonably linked to treatments being compared
- Sensitive and specific - changes detected accurately
- Reliable and robust - similar results obtained by different people in various settings
- Simple and sustainable
- Cheap and timely
When should we monitor clinical trials?
- Need to monitor inadvertent differences in groups
- Monitor outcomes during trial
- Possible effect - is one group being disadvantaged
- Adverse effects - are patients being harmed
- Final measurement of outcomes - compare final effect of treatments in trial
How do we show comparability between groups?
- Ensure that groups compared are as equivalent as possible
- Collecting baseline data on characteristics that we think may relate to both condition and outcomes we are investigating
Define the placebo effect
- Even if the therapy is irrelevant to the patient’s condition, the patient’s attitude to their illness and the illness itself may be improved by a feeling that something is being done about it
What is a placebo?
- Inert substance made to appear identical in every way to active formulation to which it is compared
- E.g. appearance, taste, texture, dosage, regimen, warnings etc.
What is the aim of a placebo?
- Cancel out any placebo effect that may exist in active treatment
What are the ethical implications of placebo?
- Placebo should only be used when no standard treatment is available
- Use of a placebo is a form of deception
- Essential that participants in a placebo-controlled trial are informed that they may receive a placebo
Why might we lose participants from a clinical trial?
- Clinical condition may necessitate their removal from trial (appropriate)
- Participant may choose to withdraw from trial (unfortunate)
How do we minimise losses to follow up?
- Make follow-up practical and minimise inconvenience
- Be honest about commitment required from participants
- Avoid coercion or inducements but can recompense participants for their time/trouble
- Maintain contact with participants
Why might participants not adhere to treatment during clinical trials?
- May have misunderstood instructions
- May not like taking their treatment
- May think treatment is not working
- May prefer to take another treatment
- Can’t be bothered to take treatment
How can we maximise adherence with treatments?
- Simplify instructions
- Ask about adherence
- Ask about effects and side effects
- Monitor adherence e.g. tablet count, urine level, blood level
- Never possible to achieve 100% adherence
How can we tell if a new treatment is better than the standard treatment?
- Is the physiological action of the new treatment better than the standard treatment
- Is the new treatment better than the standard treatment in routine clinical practice
What are the pros and cons of ‘as-treated’ analysis?
- Analyses only those who completed follow-up and adhered to treatment
- Compares physiological effects of treatments
- But loses effects of randomisation
- Non-adherers likely to be systematically different from those adhering to treatment
- Leads to selection bias and confounding
Outline intention to treat analyses
- Analyses according to original allocation of treatment groups
- Regardless of whether participants completed follow-up or adhered to treatment
- Compares likely effects of using treatments in routine clinical practice
- Also preserves effects of randomisation - minimal selection bias and confounding
Compare as-treated with intention to treat analyses?
- As treated analyses tend to give larger sizes of effect
- Intention to treat analyses tend to give smaller and more realistic sizes of effect
- Clinical trials should normally be analysed on an intention to treat basis
Outline the steps involved in a RCT
- Disease of interest
- Is it ethical?
- Methods of group allocation and blinding
- Patients eligible for trial
- Patients to be excluded from trial
- Baseline data collection (demonstrate group comparability)
- Outcomes to be measured
Outline what steps need to be taken in order to conduct a RCT
- Identify a source of eligible patients
- Invite patients to participate
- Obtain consent
- Allocate participants to treatments fairly
- Follow up participants in identical ways
- Minimise losses to follow-up
- Maximise adherence with treatments
Outline the steps taken to analyse comparison of outcomes from a RCT
- Is there an observed difference in outcome between treatment groups
- Is it statistically significant (or due to chance)
- Is the observed difference clinically important
- Is the observed difference attributable to the treatments compared in the trial
