Methods The different approaches considered in this investigation can be grouped into simple methods, and more sophisticated methods which make adjustments to the hazard ratio or use accelerated failure time models. selleck chemicals llc Simple methods Various methods have been used in existing literature in situations Inhibitors,Modulators,Libraries where patients depart from their randomised treatment. We refer to them here as simple methods which tend to involve only small adjustments to stan dard survival techniques. This section will focus on four of these, intention to treat, excluding or censoring patients if they switch treatments and modelling treat ment as a time varying covariate. Intention to treat Many authors take the pragmatic approach and use an intention to treat analysis. Patients are analysed Inhibitors,Modulators,Libraries depending on which treatment arm they were rando Inhibitors,Modulators,Libraries mised to.

An important feature of an ITT analysis is that data from all randomised patients is utilised, with censoring used in time to event analysis for patients who are lost to follow up to ensure this. Inhibitors,Modulators,Libraries The results from an ITT analysis should always be given regardless of whether the effectiveness of the treatment is of interest as it reflects the design and con duct of the study. While analysis of this type is perfectly valid, it may underestimate the appropriate policy effec tiveness of a treatment. For example, if the experi mental treatment truly is superior to the control treatment, and some patients have switched from con trol to experimental, and are therefore receiving the benefits of this, using an ITT analysis will make the treatments appear more similar than they really are.

The benefit of this type of analysis is that randomisation balance between groups is maintained, reducing the pos sibility of bias affecting the results. Per protocol A per protocol or as treated approach involves ana lysing patients according to the treatment they Inhibitors,Modulators,Libraries actually received rather than that to which they were rando mised. This is commonly used to supplement an inten tion to treat analysis. Here we define a per protocol approach as an attempt to estimate the policy effectiveness by censoring patients at the point at which they switch, or completely exclud ing any switching patients from the analysis. Therefore unlike the ITT analysis www.selleckchem.com/products/CP-690550.html described previously, not all available patient data is utilised. Whereas ITT uses ran domisation to ensure treatment arms are balanced in all aspects other than that of treatment, PP analysis may be subject to selection bias as groups may no longer be balanced after a patient is censored or excluded. This type of bias is particularly likely if a patients prob ability of switching treatments is strongly related to their underlying prognosis.