A landmark, or cross-sectional, analysis simply gives a snapshot of patients' QL at a specific timepoint. One advantage of landmark analyses is that all the data available at that timepoint can be used. Choosing the most appropriate timepoint is important, as shown by Curran and colleagues  who compared global QL at six different timepoints in an analysis of two regimens for locally advanced breast cancer, and showed major differences in treatments depending on the timepoint. Indeed, as mentioned in Chapter 6, there is always a danger that the landmark timepoint chosen may not be the point at which differences are occurring or may give an unrepresentative picture.
Simple landmark analyses give no indication of how patients have improved or deteriorated since the start of the trial or the start of treatment. An alternative is to calculate the change in score i.e. the score at the timepoint minus the score at baseline. The disadvantage is that it limits the analysis to only those patients with data available at both timepoints.
Was this article helpful?