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

Table 5

Estimated survival time of censored observations in treadmill test for the multiple imputation method, the proposed method, and the proposed method by adding age to the model.

Placebo-drug sequence			Drug-placebo sequence
Model	Subject (age)	Period 2 (drug)	Subject	Period 1 (drug)	Period 2 (placebo)
		Y2		Y1	Y2
Multiple imputation	3		4		_
	11		16	_
	13		18		_
MERL	3		4		_
	11		16	_
	13		18		_
MERL (adding age)	3 (45.79)		4 (40)		_
	11 (47.58)		16 (65.16)	_
	13 (70)		18 (41.34)		_

Y₁: posttreatment response in period 1; Y₂: posttreatment response in period 2.