Intention to Treat - John Stites, DC, DACBR, FACO
In the critical appraisal of a randomized controlled trial (RCT) one question usually asked is, “Were patients analyzed in the groups to which they were randomized?” Another way of stating this question is, “Was an Intention-to-Treat principle followed in the analysis?”
Why is this important?
In a RCT randomization is essential. The purpose of randomization is to help ensure that there is prognostic balance between groups. If the groups are balanced you can have greater confidence that any difference detected between groups is due to the intervention.
There are some prognostic factors that may be known or predictable. However, it is entirely possible that there are prognostic factors unknown and not predictable. Randomization helps to minimize differences between groups due to unknown prognostic factors.
Intention-to-treat maintains randomization.
Intention-to-treat can be counterintuitive.
Intention-to-treat means that everyone randomized to a treatment group is analyzed whether they receive treatment or not.
Figure 1 illustrates two treatment arms. Let’s assume we know that they are equally effective. Of the 100 individuals in the first arm 60 responded to care and 40 did not. The same is true in the second arm. Suppose that 20 participants randomized to the first treatment group decided that they liked the doctors in the second treatment group better and got treatment ‘B’. If these 20 individuals were prognostically different from others in the group and analysis was done based on which treatment was received, incorrect conclusions could be drawn (Figure 2). This per-protocol analysis violated randomization and made it look as though treatment ‘A’ is more effective than treatment ‘B.’ In this case 75% of patients in treatment ‘A’ improved but only 50% of patients in treatment ‘B’ improved.
Far more common than study participants moving to a different group is participants dropping out of a study. Even if they drop out, you still need to analyze them in the group to which they have been assigned. Not including the drop outs violates randomization and can lead to an incorrect conclusion if the dropouts are prognostically different from the rest of the group.
Check the methods
The methods section should have a statement regarding use of intention-to-treat and should be explicit in explaining how dropouts are accounted for in the analysis.
There is no such thing as a perfect study.
Intention-to-treat preserves randomization.
Intention-to-treat helps maintain prognostic balance.
Intention-to-treat reporting is associated with higher quality studies.
Guyatt G, Montori V, Devereaux PJ, Durieux P. Chapter 9.4. The principle of intention to treat. In: Guyatt G, Rennie D, Meade MO, Cook DJ, eds. Users’ Guides to the Medical Literature: A Manual for Evidence-Based Clinical Practice. 2nd ed. New York, NY: McGraw-Hill; 2008. http://www.jamaevidence.com/content/3348972. Accessed 1/20/2012
Hollis S, Campbell F. What is meant by intention to treat analysis? survey of published randomised controlled trials. BMJ. 1999;319(7211):670-674.
Ruiz-Canela M, Martinez-Gonzalez MA, de Irala-Estevez J. Intention to treat analysis is related to methodological quality. BMJ. 2000;320(7240):1007-1008