Clinical trials & medical research

Understanding allocation (randomisation) to a treatment comparison group

In some trials (especially Phase 1 and Phase 2 trials) everyone taking part will get the treatment being evaluated. (See also 'Why do we have clinical trials in children and young people?’) However, many trials compare a new treatment with the standard or usual treatment by allocating people to two groups. People allocated to one group (the experimental group) are given the new treatment. Those in the other group, called the control group, are given a standard or existing treatment. If no standard treatment exists, the control group may not be given any specific treatment or may be given a placebo. Some trials compare more than two groups, perhaps comparing a new treatment with two or more well-established treatments. (Sometimes you may hear the experimental group called the ‘trial group’ or ‘intervention group’. This can be confusing, as all the groups, including the control group, are part of the trial, and people in the control group may also be given an intervention, in the form of the standard treatment). 

Kay took part in her first clinical trial two years ago. The aim of the trial was to test improved medicines for the treatment of cystic fibrosis. 

Soon after Kay completed her first trial she was invited to take part in a similar trial for a different medicine. Because of the success of the first trial, she was keen to take part, but she had to leave a gap between completing one trial and starting another. (See ‘Different types of trials’ and ‘What is involved in a trial: appointments and monitoring’.)

To make sure that each group contains a similar mix of people, many trials are ‘randomised’. This means that people are allocated at random to one of the comparison groups in the trial, often by using a computer programme. When people are randomised they have an equal chance of being in either trial group. Random allocation helps ensure that two very similar groups of patients will be compared, so if one group does better than another, it is likely to be because the treatments being compared have different effects, and not because of differences between the people in the groups. The UK Clinical Research Collaboration booklet on ‘Understanding Clinical Trials’ explains these and other terms.

Hannah, aged 17, took part in a randomised trial comparing different approaches to giving information to young people with diabetes. She had to complete two questionnaires asking about visits to the doctors, what information she received, and questions about diabetes. She says she was allocated to the control group. 

If no standard treatment is available, the control group may not be given any specific treatment, or may be given a placebo. A placebo is a treatment with no active ingredient, which is designed to appear very like the treatment being tested. By comparing people’s responses to the placebo and to the active treatment being tested, researchers can tell whether the treatment is having any effect in addition to the effect of feeling better because ‘something is being done’.

There are several ways in which the results of trials can be made as reliable and accurate as possible. One of these is to make the trial a ‘blind trial’. In a blind trial the participants are not told which group they are in. This is because, if they knew which treatment they were getting, it might influence how they felt or reported their symptoms. Some trials are double-blind, which means that neither participants nor the doctors and others treating them know which people are getting which treatments. This also avoids the doctors’ hopes and expectations influencing the results of the trial.

Some young people were told by doctors that being randomised was like a ‘lottery draw’, or being drawn ‘out of a hat’. Others were unsure about randomisation, and some found it difficult to explain. Often people knew they would prefer to be in one arm or the other. Joe was pleased to be allocated to the new intervention.

Randomised trials are done when we don’t know which treatment is best, in other words when the relative merits and disadvantages of different treatments are uncertain. It is important to realise that, on average, new treatments are as likely to turn out worse as they are to turn out better than existing treatments. This means that, going into a trial, everyone, regardless of which of the treatment groups the computer allocates them to, must have similar chances of a good outcome. If people would strongly prefer one of the treatments being compared, they should not volunteer for the trial. 

Mohini decided not to take part in a trial after a lot of discussion with the consultant and her parents, because she really wanted the standard treatment. (See also ‘Deciding not to take part although eligible to take part in a clinical trial’.) 

There are several ways in which the results of trials can be made as reliable and accurate as possible. One of these is to make the trial a ‘blind trial’. In a blind trial the participants are not told which group they are in. This is because if they knew which treatment they were getting, it might influence how they felt or reported their symptoms. Some trials are double-blind, which means that neither participants nor the doctors and others treating them know which people are getting which treatments. This also avoids the doctors’ hopes and expectations influencing the results of the trial.

If no standard treatment is available, the control group may not be given any specific treatment, or may be given a placebo. A placebo is a treatment with no active ingredient, which is designed to appear very like the treatment being tested. By comparing people’s responses to the placebo and to the treatment being tested, researchers can tell whether the active treatment is having any real effect in addition to the effect of feeling better because ‘something is being done’.

Some young people we talked to took part in randomised double-blinded placebo controlled trials. This means that young people were allocated randomly to one of the trial groups and neither young people nor doctors or others involved in their care knew which group they were in. For some young people this was quite a scary thought. 

Toby, aged 11, took part in a randomised placebo controlled trial of treatment for migraine in children. He explains a double-blinded trial as “Only the people (pharmacy) that shuffled them up (the pills) and gave them to me know, like the doctors and nurses didn’t”.

Audio onlyText only
Read below

Some young people described placebo as a ‘dummy’, ‘fake’ or a ‘blank’ drug and not the actual medicine.

Jenna aged 13 was diagnosed with polyarticular idiopathic arthritis at the age of 11 and is taking part in a double-blinded trial. The trial is a three year trial and Jenna has been taking part for one year so far. She attends the hospital once a month to receive the treatment which involves being on a day ward to receive the ‘drug’ or the ‘placebo’. She explains randomisation as; “It’s where they just sort of randomly put people in different groups to sort of make it fair”.

Ryan, aged 12, was also a little worried that the doctors didn’t know what drug he would be taking in case his arthritis starting hurting again.

Sometimes when people do not experience a change in symptoms they assume they must be taking the placebo. However, this is not necessarily the case, as not all drugs work on all patients in the same way. Equally, people who are taking a placebo may sometimes experience improvements in their health. This is known as the placebo effect or placebo response. We still do not understand exactly why this happens, but it seems that believing a treatment will help can result in real changes. This is why new drugs are often compared against a placebo – if the patients on the drug do significantly better than those taking the placebo, it suggests the new drug has an important additional effect. If people were only given the new drug, we could not be sure if improvements were simply due to a placebo effect.

Robert was diagnosed with cystic fibrosis at birth. He has taken part in various trials since childhood. When he was 14 he took part in a randomised placebo controlled trial to assess whether a steroid used for people with asthma helped people with cystic fibrosis. 

Some of the young people we talked to were pleased they didn’t know which drug they would be taking as Katie explains.

A lot of the terms used in trials are quite complex and closely related to each other. Below are some brief definitions as a reminder of how they fit together. For help understanding these terms see links to further information in our Medical reseach resources.

Control group – a comparison group in which people often get a standard treatment. If no standard treatment exists the control group receives no specific treatment or a Placebo. 

Randomisation – allocating people at random to one group or another, so that each group contains a similar mix of people. Random allocation helps ensure two very similar groups of patients will be compared, so if one group does better than another, it is very likely to be because the treatments being compared have different effects, and not because of differences between the people in the groups.

Placebo – is a treatment with no active ingredient which is designed to appear very similar to a supposedly active treatment with which it is being compared. By comparing people’s responses to the placebo and to the treatment being tested, researchers can tell whether the treatment provides any additional benefit.

Double blind – trying to ensure neither patients nor doctors or others treating people in a clinical trial know which treatment each person is getting, so that this knowledge does not influence how patients feel or how doctors interact with their patients or interpret the results.

For more information, see resources.

Last reviewed March 2017.

Feedback

Please use the form below to tell us what you think of the site. We’d love to hear about how we’ve helped you, how we could improve or if you have found something that’s broken on the site.

Make a Donation to healthtalk.org





Find out more about how you can help us.

Send to a friend

Simply fill out this form and we'll send them an email