Population science Flashcards
define a clinical trial
any form of planned experiment which involves patients and is designed to find most appropriate treatment for future patients with a given medical condition
pruprose of a clinical trail
provide evidence of a treatment efficacy and safety
what is efficacy
measuring the effect of the treatment under perfect conditions
what is ‘safety’ in terms of the purpose of a clinical trial
not to harm the group a=under specific conditions
layout of a clinical trial
new treatment measured against the standard treatment which could be a placebo, or the standard treatment or nothing
what factors need to be considered to make the trial fair
needs to be:
reproducible
controlled (comparisons)
fair (unbiased without confounding)
what is the difference between effectiveness and efficacy
efficacy is the ideal clinical conditions
effectiveness is real world clinical experience. without the perfect clinical conditions
what are the different phases of clinical trials
phase 1- volunteer studies (pharmacodynamics and kinetics + side effects)
phase 2- treatment studies (effects, dosages and common side effects)
phase 3 - clinical trials (comparison with other treatments)
phase 4- post marketing surveilllence (monitoring with standard treatments)
what is a non-randomised trial
allocation of patients receiving a new treatment to compare with group receiving standard treatment.
problems with non-randomised clinical trials
- allocation bias
- confounding
what is comparison with historical controls
comparison of group of patients who had standard treatment with those receiving new treatment
problems with comparison with historical controls
selection is less well defined, may have been treated differently from new treatment group , less info about bias and confounders, can’t control for all confounders
explain the conduct of a randomised clinical trial (RCT)
- identify the eligible patients
- invite patients to trial
- consent patients
- allocate the treatments fairly
- follow up identically
- minimise losses to follow up
- maximise adherence to treatments
definitions of the key study variables
- the disease of interest
- the treatments to be compared
- the outcomes measured
- the possible bias and confounders
how to compare outcomes of an RCT
use the 95% confidence interval.
the observed rate ratio is used to work out the 95% confidence interval.
If p value (1) is within the confidence interval then null hypothesis is correct, if not then it is incorrect.
why we need pre-defining outcomes for an RCT
need to define how outcomes are measured before the trial to prevent data dredging (using it to uncover patterns), it is also protocol for data collection.
what is a primary outcome
the thing you want to find out from the trial. is used in the sample size calculation
what is secondary outcome
another outcome of interest
different types of outcomes (x3)
Pashto-physiological e.g tumour size
clinically defined e.g death
patient focused e.g quality of life
features of an ideal outcome
appropriate and relevent - to the patient
valid and attributable - can be linked to treatments being compared
sensitive and specific - detects changes accurately
reliable and robust - still measurable in variable settings = similar result
simple and sustainable - method can be carried out and repeated
cheap and timely - not execssively expensive measure and doesn’t have lag time
what is non-random allocation
when you allocate them to groups that aren’t random but are intended to produce similar groups. potential for selection bias and confounding factors
what is random allocation and its advantages and disadvantages
allocate participants to the treatments fairly.
minimal allocation fairly
minimal confoudning
BUT randomising numbers in small groups can result in unequal groups
what is minimal allocation
randomisation gives each patients equal chance of being selected
minimal confounding
in long run leads to groups that are similar in size and characteristics by chance
what is a confounder
factor associated with the exposure and is indenpnedtly a risk factor for the dsiease
the problem with knowing which treatment you’ve been allocated
patient may alter their behaviour (behaviour effect)
clinician may alter treatment or care or interest (non-treatment effect)
investigator may alter approach when making measurements (measurements bias)
how to avoid selection bias
randomisation
how to avoid recall/reporting bias
black and white tools e.g dead or alive or blinding
how to avid observer bias
blinding the observer
3 types of blinding
single blind - patient doesn’t know
double blind - patient, and investigator don’t know
triple blind - no one knows
examples of when blinding is difficult
surgical procedure
lifestyle interventions
prevention programes
psychotherapy vs antidepressant
what is comparing with no treatment
new treatment group compared to group receiving no treatment. see if it was actually the treatment or just the fact they were receiving ‘care’
what is the placebo effect
attitude of patient towards illness may be improved by a feeling that something is being done about it
what is a placebo
inert substance made to appear identical to the one it is being compared with e.g formula, tenxture, smell eat
what is the ethical implication of a placebo
use of placebo is a form of deception therefore participants must be informed they may be receiving one. also not allowed to use it unless there is no standard treatment to compare it to
what is losses to follow up
not all participants stay in trial - condition may mean worsen and they need to withdraw or they chose to withdraw
how to minimise loses to follow up
make follow up practical and minimise inconvenience
be honest about the commitment needed
maintain contact with patients
causes of non-compliance
misunderstood instructions, don’t like taking treatment, Think the treatment isn’t working, prefer to take another treatment, can’t be bothered
How to maximise compliance
Make instructions easy, ask about compliance, ask about effects and side effects and monitor compliance for example with a tablet count
How to consider issues
Not possible to have 100% follow-up therefore analysis outcome should take these issues into account
What are the two different interpretations of whether the new treatment is better than the standard treatment
- The psychological action of the new treatment e.g. does the drug A lower blood pressure more than the other drug B
- Is the new treatment better than the standard treatment in a routine clinical practice e.g. if you prescribed drug A is it better at lowering blood pressure than drug B
What is an explanatory trial?
It analyses only those who complete follow up and comply with treatment.
BUT this loses affect of randomisation because non-compliance are likely to be systematically different from compilers. This is selection bias and confounding
What is a pragmatic trial
Analyses everyone according to their original allocation.
Preserved randomisation, minimal selection bias and confounding
Compares the likely effects of using treatments in routine clinical practice
As treated versus intention-to-treat
As treated give a larger sizes of affect whereas intention-to-treat analyses tend to give smaller and more realistic sizes of the effect. usually you would use intention to treat
What is the declaration of Helsinki
But the health of patients will be the first consideration and only acting in the patient’s interest
What is collective ethic
All patients should have treatments that are properly tested for efficacy and safety
What is individual ethic
The principles of the four pillars of ethics benefice nonmaleficence autonomy and justice
Issues to consider for a clinical trial to be ethical
Clinical equipoise robust scientifically Ethical recruitment valid consent Volunteriness
What is clinical Equipose
uncertainty or genuine ignorance about the better treatment or intervention - why you need to test it
But what constitutes reasonable uncertainty and how is better defined for the patient and is uncertainty or ignorance by the clinician or the community as a whole?
What does scientifically robust mean
It’s addressing an important issue (which can be justified) and asked a valid question, has the potential to reach some conclusions and is it appropriate study design (taking safety into consideration) and protocol
What is ethical recruitment and what are the issues of it
Issues include participants with the inappropriate inclusion from communities that won’t benefi, participants that have a high risk and participants that are likely to be excluded from the analysis (small subgroup of Chinese), People who can’t you can’t get valid consent from people who differ from the ideal homogenous group.
What does valid consent involve
Knowledgeable informant, that the patients are given appropriate information, that the patient is competent to make a decision and someone to legitimately authorise the consent
Signed consent does not equate to valid consent!
What does voluntariness involve
The decision should be free from coercion or manipulation
Volunteering this is part of the consent to being valid
What is coercion
When the patient doesn’t access the best treatment or has a lower quality of care which is why they take part in the trial
What is the role of the research ethics committee
To reserve the dignity rights safety and well-being of participants
What is the role of the NHS trust in the research ethics committee
They are involved with the financial management, resource implications and research governance
What does the research ethics committee focus on
The recruitment of participants, care and protection protection of research participants’ confidentiality, informed consentand community considerations
