Pool subjects in average time trend plots when treatment crossover is detected

Check this box to pool subjects across treatment periods in the average time trend plots when a treatment crossover is detected. This results in the display of the average time trend plots for each unique treatment value across all treatment periods.

Note the following differences in output depending on whether Overlay visits when treatment crossover is detected is checked:

•

When the Overlay visits when treatment crossover is detected box is not checked , the average time trends view has one connected line representing the average for that treatment across all periods (assuming that the treatment value is coded with the same value across periods). Treatment values that changed across treatment periods still have separated lines at the moment of crossover. Note that in this case, this does not change the composition of the subjects used for calculating the means .

•

When the Overlay visits when treatment crossover is detected box is checked , subjects that have the same value for treatment across different periods are pooled to calculate the average trend across the visit(s). For example, subjects assigned to Placebo on period 1 are pooled with subjects assigned to Placebo on period 2, and only one line is displayed for Placebo . (If this box is not checked , two separate lines would be drawn for Placebo (1) and Placebo (2) .)

To Specify This Option:

Check the box.

Note : Refer to Crossover Analyses for more information about analysis of crossover data.