The Value Of Randomised Clinical Trials

The benefits of randomised trials are outlined below.

Assessment of effectiveness.

Randomised clinical trials promote the assessment and quantification of treatment / intervention effectiveness (McPherson, 1994). McPherson (1994) opines that decisions regarding health care services, efficacy, equity and cost cannot be made on a scientific, intelligently discussed, basis if health care interventions are not adequately evaluated. While other study designs can detect association between interventions and outcomes, only randomised trials can rule out the possibility that the association was caused by another factor linked to the intervention and the outcome (Sibbald and Roland, 1998).

Furthermore trials can provide some information in order to predict the likelihood of an individual patient responding to an intervention in a similar way to the findings of the trial, providing that the trial is designed to provide data on participant characteristics likely to effect outcome (Mant, 1999).

The value of randomisation.

Randomisation is performed to form prognostically comparable groups at baseline (Koes and Hoving, 1998); since researchers can rarely identify and measure every prognostic variable for an intervention randomisation is relied upon to evenly distribute known and unknown prognostic factors across the groups. When patients are randomised to one of two or more treatments the act of randomisation aims to prevent potential bias in the assignment of patients by introducing unpredictability (Kunz and Oxman, 1998). Each patient is assigned a treatment through the play of chance rather than by advance prediction.

The effects of randomisation were explored in a useful systematic review by Kunz and Oxman (1998) who examined randomised and non randomised clinical trials and trials using concealed random allocation versus trials using inadequate concealment to determine the relationship between randomisation and estimates of effect. The review discovered that failure to use random allocation and concealment of randomisation generally resulted in overestimates of treatment effect (up to 150% or more of the estimated effect) although in some studies this bias went in the other direction and reversed the direction of the effect (up to 90% of the estimated effect) or masked an effect. Kunz and Oxman’s (1998) conclusion is a powerful one:

Failure to use adequately concealed random allocation can distort the apparent effects of care in either direction, causing the effects to seem either larger or smaller than they really are. The

size of these distortions can be as large or larger than the size of the effects that are to be detected.

The need to ensure adequate randomisation and adequate concealment in the proposed trial is therefore clear.

Defence of results.

Abel and Koch (1999) also suggest that randomisation assists researchers in the defence of positive results to possible sceptics since the methodology is held is such high repute. Primacy is given to evidence from randomised clinical trials, sometimes summarised as the “if it moves, randomise it” approach (Croft, 1998).

Prestigious Design.

As mentioned the randomised clinical trial is commonly perceived to be the “gold standard”

for comparing and evaluating different treatments (Pringle and Churchill, 1995). The randomised trial is an established methodology (Oakley, 1998), occurring since the 1940s (Swales, 1998); trials are currently deemed highly credible, attractive to funding bodies (Brewin and Bradley, 1989), and publishable, since randomisation can serve as an indication that the research has been carefully considered and thoughtfully carried out (Abel and Koch, 1999).

Evidence based practice.

In addition to the value of a single study, all studies pertaining to a specific treatment or intervention can undergo systematic review or meta-analysis, as in the previous chapter, to provide an overview of current evidence (Mulrow, 1994; Knipschild, 1994). This can then be used in order to promote evidence based practice (Greenhalgh, 1997; Main 2003) and, as in this trial, to assist the development of interventions for future studies (Knipschild, 1994).

Blinding.

Blinding is also frequently considered a beneficial component of randomised clinical trials.

This aspect of trial design is the subject of Chapter five and is described and discussed there in detail.

Control / Placebo groups.

A major design benefit is that all clinical trials must contain a control group against which the intervention group can be compared so that any differences in outcome may reasonably be attributed to the intervention under investigation (Freidman et al., 1998 p. 2). In laboratory experiments it is possible to provide or withhold an intervention, however in clinical studies this is often inappropriate or unethical Placebo⁴controls may be used, primarily to enable patient attitudes to the research study to be as similar as possible in both treatment and control groups to minimise bias (Pocock, 1983, p. 93), but most often a new intervention is compared to the most effective current treatment or usual care (Friedman et al., 1998 p. 2). This may mean the comparison group receives no intervention at all (Friedman et al., 1998 p. 2). This is not considered disadvantageous by pragmatic trialists: pragmatic trials help clinicians to decide between a new treatment and the best current treatment, therefore the introduction of clinician or patient biases due to the lack of a placebo are not considered detrimental but accepted as part of their overall response to treatment and included in the overall evaluation (Roland and Torgerson, 1998). As Roland and Torgerson state:

….the treatment response is the total difference between two treatments, including both treatment and associated placebo affects, as this will best reflect the likely clinical response in practice.

4 A placebo is an inert substance/procedure appearing similar to the trial intervention (Pocock, 1983 p. 9)