Clinical trial
any form of PLANNED EXPERIMENT which involves patients and is designed to elucidate the most appropriate method of treatment for FUTURE PATIENTS with a given medical condition
- to provide reliable evidence of treatment efficacy and safety
Efficacy - the ability of health care intervention to IMPROVE THE HEALTH of a defined group under specific conditions
Safety - the ability of health care intervention NOT TO HARM a defined group under specific conditions
MUST BE - reproducible (in experimental conditions)
- controlled (to compare)
- fair (unbiased and without confounding)
Non-randomised clinical trials
Allocation of patients receiving a treatment or those who act as the controls
Allocation bias - by patients, clinician or investigator
Confounding - known and unknown factors
Comparison with historical records
- comparison between a group who have already received the standard treatment compared with those receiving the new treatment
For the standard treatment group - the selection is often less defined, less rigorous
- are treated differently to the new treatment group
- less information about potential bias or confounding factors
- unable to control the confounders
Definition of factors in a RCT
- the disease of interest must be identified
- the types of treatments to be compared
- the outcomes to be measured
- identity possible bias and confounders
- identify patients eligible for the trail
- identify the patients to be excluded from the trail
Conduct of the RCT
- identify and get consent for the patients eligible and willing to be part of the trail
- allocate the patients to a type of treatment fairly - without bias or confounding
- follow up patients from either treatments in identical ways
- minimise loses to follow up
- maximise compliance to treatments
Comparison of RCT outcomes
- is there an observed difference between the groups? if so, could this have arisen by chance? (is it significantly significant?)
- how big is the observed difference - is it clinically important?
- is the observed difference attributable to the treatments compared in the trail?
Pre-defining outcomes
- prevents repeated a always
- agreed criteria for measurements and assessment of outcomes
Outcomes
Primary - used in sample size calculation
Secondary - other outcomes of interest, occurrence of side-effects
TYPES
patho-physiological - e.g. tumour size, thyroxine level, ECG changes
clinically defined - e.g. death, disease or disability
patient focused - e.g. quality of life, psychological and social well-being, satisfaction
Features of an ideal outcome
death satisfies all of these
Appropriate and relevant (to patient, clinician, etc.)
Valid and attributable - observed effect can be reasonably linked to treatments being compared
Sensitive and specific - measurement can detect changed accurately
Reliable and robust - outcomes similar in different people in different settings
Simple and sustainable - easily carried out repeatedly
Cheap and timely - not excessively expensive it has a long lag time
Non random allocation
Allocation of patients to treatments by e.g. geographical location, numerical order, etc. leads to the potential for allocation bias and confounding factors - causes unidentified differences between the treatment groups being compared
Random allocation (Allocate patients to the treatments fairly)
- minimise allocation bias - individuals have an equal chance of being allocated to each of the treatments in the trail
- minimal confounding - treatment groups are likely to be similar in size and characteristics by chance
(applies to both known and unknown confounding factors)
Open label - knowing the treatment allocation
- patient may alter their own behaviour (placebo effect could occur)
- clinician may alter their care and interest in the patient (non-treatment effect)
- investigator may alter their approach when making measurements and assessing outcomes (measurement bias)
Blinding/Masking
Equal chance of being in either treatment group
Single bind - patient/clinician/assessor doesn’t know the treatment allocation
Double blind - two out of the patient/clinician/assessor doesn’t know the treatment allocation
Triple blind - (double blind usually means that none of them know the treatment allocation
Patients are usually asked what they thought they got - 50/50 ensures that the study is masked
Difficult in
- surgical procedures
- psychotherapy vs antidepressants
- alternative medicine vs western medicine
Comparing with no treatment
the placebo effect
- compare a new treatment with a group receiving no treatment
- to determine if the observed difference is due to the new treatment or due to the fact that the group was receiving care
The placebo effect
The patients attitude to their own illness and the actual illness may be improved by a feeling that something is proactively being done about it
Placebo - an inert substance that appears identical to the active formulation
Ethics - placebo use is a form of deception (patients in trial must be informed that they may receive a placebo)
Loses to follow up
Appropriate - clinical conditions require removal from trial
Unfortunate - patient chooses to withdraw themselves
Minimise loses
- follow up practical and continent
- be honest about expected patient commitment
- avoid incentives or coercion
- maintain contact
Scientifically proven to work
It is better than the other treatment
- more people receiving new treatment were cured compared to the old treatment
Is the ‘new treatment’ better than the ‘standard treatment’?
‘AS-TREATED ANALYSIS’ - Physiological action
- non compilers are likely to be systematically different to the compilers (selection bias and confounding)
- larger sizes of effect
‘INTENTION TO TREAT ANALYSIS’ - routine clinical practice (takes into account those who do not take the drug)
- preserves the effects of randomisation
- more realistic sizes of effect
- should be what clinical trials are based on
RCT ethics
RCT aim to properly test treatments for efficacy and safety
- do not guarantee benefit for an individual
- may result in harm of patients in trail
Are a benefit for future patients
RCT must be approved by a research ethics committee
Valid consent
Knowledgable informant
Appropriate information with no manipulation to take place (free from coercion)
A signed consent form does not equate to valid consent
