Box 92 Intentiontotreat ITT analyses

♦ include all patients randomized in the analysis according to their allocated treatment, whether or not they actually received it

♦ are the preferred form of analysis of trials which are planned to show a difference between two treatments

♦ minimize bias

♦ are likely to produce 'conservative' estimates of the difference between treatments

♦ may be supplemented (but not replaced) by analyses excluding certain groups of patients; if this is done, then the exclusion of patients should be justified, and the likely bias that these exclusions bring to the analysis should be fully discussed.

intention-to-treat analyses tend to dilute an effect between treatments, an ITT analysis in an equivalence trial may make the treatments appear to be more similar than they actually are. Thus, for equivalence trials a 'per protocol analysis' (PPA) could be regarded as 'conservative' and therefore is often given as much emphasis as an ITT. A PPA is defined as the analysis of the set of data generated by the subset of patients who have complied with the protocol sufficiently that they are likely to show the effects of the treatment. Thus, for example, in an equivalence trial comparing two chemotherapy regimens PPA analysis may concentrate on those patients who received at least two courses of treatment.

0 0

Post a comment