Bias and Its Remedy in Interval-Censored Time-to-Event Applications - Interval-Censored-Time-to-Event Data: Methods and Applications - page 326

positive bias in the estimated regression coecients, which is acceptable from a
practical point of view although care should be taken in interpretation. They also
observed problems in the algorithm with respect to maximizing the likelihood in
the presence of a high percentage of right-censored data (> 30%). This is consistent
with the nding from our simulation. We therefore recommend using the \IntCox"
approach in analyzing interval-censored time-to-event data. A novel method in Chen
et al. (2012: In press) using Taylor series to approximate the log baseline hazard
function in Cox proportional hazards regression showed further improvement for
bias correction.
Bibliography
Chen, D. G. and Peace, K. E. (2010). Clinical Trial Data Analysis Using R. Boca
Raton: FL: Chapman and Hall/CRC Biostatistics Series.
Chen, D. G., Yu, L., Peace, K. E., Lio, Y. L., and Wang, Y. (2012: In press). Ap-
proximating the baseline hazard function by Taylor series for interval-censored
time-to-event data. Journal of Biopharmaceutical Statistics .
Cox, D. R. (1972). Regression models and life-tables (with discussion). Journal of
the Royal Statistical Society: Series B 34, 187{220.
Henschel, V., Heib, C., and Mansmann, U. (2007). Intcox: Compendium to apply
the iterative convex minorant algorithm to interval-censored event data. Online
Paper with the IntCox package .
Pan, W. (1999). Extending the iterative convex minorant algorithm to the Cox model
for interval-censored data. Journal of Computational and Graphical Statistics 78,
109{120.
Panageas, K. S., Ben-Porat, L., Dickler, M. N., Chapman, P. B., and Schrag, D.
(2007). When you look matters: The effect of assessment schedule on progression-
free survival. Journal of National Cancer Institute 99, 428{432.