Explaining randomization and rationale for trials

A major obstacle to trial participation appears to be explaining the randomization process. Some doctors may find it difficult or awkward to discuss, and participants may find the concept difficult to deal with. One study reported that 63 per cent of patients approached refused entry to a trial because of an aversion to randomization [25], and it is thought that many trial participants do not fully understand the nature of the trials in which they are involved [9]. Other studies, which have paid particular attention to explaining the process of randomization, have suggested that comprehension levels can be good.

The UK population survey described previously, in which interviewers were specifically instructed to ensure that randomization was explained thoroughly, found that most respondents understood the concept of randomization [6]. Likewise, a study, using in-depth semi-structured interviews, of men participating in a randomized trial (CLasP) evaluating the effectiveness and cost-effectiveness of three different types of intervention for the management of urinary symptoms related to benign prostatic disease, found that almost all participants were aware of some aspects of randomization. Most acknowledged the involvement of chance in their treatment, often transformed to lay terms such as a lucky dip or lottery [26].

However, a revealing study that involved interviewing parents who had consented for their critcally ill infants to enter a randomized trial, found that although superficially many appeared to know that the trial was randomized, when questioned further it became clear that few understood either the design of the trial or how or why randomization was done [27]. Agreement to participate was made in one of the most distressing and difficult of situations. Their newly born babies, already perceived by the majority of parents to be dying, were to be randomized to continue with unchanged ventilation or to be transferred to another hospital to receive oxygenation of the blood through an external circuit (ECMO) inserted through a cannula in the neck. The majority of parents found it difficult to accept the idea of clinical uncertainty and thought that the new treatment was obviously better and that they were being offered the chance of taking part because standard care was failing. The trial was viewed as an opportunity to gain access to a new, and implicitly better, treatment, which had only limited availability. Rather than a situation where the two treatments were the focus of randomization, they saw their baby as being at the centre of a decision about access to ECMO (which they were competing with other babies for). Only four parents out of thirty-seven suggested that randomization had been done to obtain a fair comparison between the treatments and only one suggested that the trial was needed to assess associated hazards of ECMO treatment. Interestingly, many of the parents thought that the babies allocated to standard care were not part of the trial and the authors comment that this may have been reinforced by the trial being known as the ECMO trial. Randomization was not fully understood and often focussed on the use of the computer. Some believed that the computer was used to make complex therapeutic decisions, or to check on the availability of an ECMO bed. Some, acknowledging the play of chance, believed that randomization was used as a route to fair rationing rather than treatment allocation - the computer was selecting babies for ECMO rather than the treatment for their particular child. Perceived competition for places on ECMO was described by many parents, and randomization was described as 'unfair' 'tough' and 'heartless.'

Undoubtedly, when discussing trials with potential participants, particular attention needs to be paid to describing the process of treatment allocation, and describing why randomization is necessary. This should perhaps start with a clear statement that new treatments are not necessarily better than existing ones, and that sometimes they may in fact be worse. Information concerning the various treatment options should be accompanied by a clear explanation that treatments are allocated in a way that makes sure that there is a fair test between the treatments and that nobody knows which treatment the patient will be allocated until they are actually entered into the trial or 'randomized.' This may be more important than describing the practicalities of the procedure. A survey that investigated how members of the public rated different explanations of randomization found that the statement that scored best was as follows:

'Once you have agreed to enter the trial, you will be allocated to one of two treatments with equal chances of each treatment being the one you will receive.'

This description was favoured over six other statements that included references to tossing a coin, drawing names from a hat, and a computer making a choice at random. The least liked statement was as follows:

'Once you have agreed to enter the trial, one of two methods of treatment will be chosen by chance and not by a decision made by the patient or doctor.'

It appeared that preferred wordings were less explicit, perhaps not allowing the mind to dwell too long on the random nature of treatment assignment or on the perceived loss of medical control [24]. Thus, although on the face of it phrases such as 'tossing a coin' may appear to be good ways of describing the process of randomization, they may also appear flippant or uncaring. We perhaps need to question whether such descriptions are actually useful. Thus, explaining how the allocation will be done may not be so important, provided that it is made clear that there is no foreknowledge of which treatment will be allocated, and that options are equally likely (or as appropriate for the ratio of allocation).

+1 0

Post a comment