Research Article

A New Method for Imputing Censored Values in Crossover Designs with Time-to-Event Outcomes Using Median Residual Life

Table 4

Event times (min) for a 10 min treadmill test in a crossover clinical trial.

Placebo-drug sequenceDrug-placebo sequence
Period 1 (placebo)Period 2 (drug)Period 1 (drug)Period 2 (placebo)
Subject (age)X1Y1X2Y2Subject (age)X1Y1X2Y2

1 (52.91)1.5111.52 (65)1112.5
3 (45.79)643.5>104 (40)6>102.52.5
5 (64.37)111.54.56 (59.08)321.5
7 (54.13)3.51.5.538 (63.46)2.52.51.52
9 (61.14).513.5810 (51.78)22.52.53
11 (47.59)6106>1012 (58.59)1.54.52.51
13 (70).5.51>1014 (55.08)3.55.54.59.5
15 (57.28)1112.516 (65.16)122>10
17 (59.75)1.51.5.518 (41.34)6>1053.5
19 (67.77)11.52420 (59.44)231.51.5
21 (42.91)55.531.522 (65.54)1.52.51.5.5
23 (50.72)2.5564.524 (70)1.53.52.53
25 (47.01)55.54.5626 (55.63)3.5966
27 (62.26)122.58.528 (62.93)25.53.58
29 (40)55.53.5230 (56.67)2.52.51.5
31 (66.27).5127.532 (63.13)2.53.52.54
33 (48.16)542234 (41.62)5.531.5
35 (65.04).5.511.536 (55.70)35.55.5
37 (66.49)1.523338 (66.62).5115.4
39 (43.19)641.5.540 (53.10)2.552.5.5
Median1.51.7523.5Median2.53.252.52.5

X1: baseline response in period 1; Y1: posttreatment response in period 1; X2: baseline response in period 2; Y2: posttreatment response in period 2.