Title: Multiple imputation in Cox regression when there are time-varying effects of exposures

Abstract: In Cox regression it is sometimes of interest to study time-varying effects (TVE) of exposures and to test the proportional hazards assumption. TVEs can be investigated with log hazard ratios modelled as a function of time. Missing data on exposures are common and multiple imputation (MI) is a popular approach to handling this, to avoid the potential bias and loss of efficiency resulting from a 'complete-case' analysis. Two MI methods have been proposed for when the substantive model is a Cox proportional hazards regression: an approximate method (White and Royston, Statist. Med. 2009;28:1982-98) and a substantive-model-compatible method (Bartlett et al., SMMR 2015;24:462-87). At present, neither method accommodates TVEs of exposures. We extend them to do so for a general form for the TVEs and give specific details for TVEs modelled using restricted cubic splines. Simulation studies assess the performance of the methods under several underlying shapes for TVEs. Our proposed methods give approximately unbiased TVE estimates for binary exposures with missing data, but for continuous exposures the substantive-model-compatible method performs better. The methods also give approximately correct type I errors in the test for proportional hazards when there is no TVE, and gain power to detect TVEs relative to complete-case analysis. Ignoring TVEs at the imputation stage results in biased TVE estimates, incorrect type I errors and substantial loss of power in detecting TVEs. We also propose a multivariable TVE model selection algorithm. The methods are illustrated using data from the Rotterdam Breast Cancer Study. Example R code is provided.