A simple method to estimate the time-dependent receiver operating characteristic curve and the area under the curve with right censored data

Statistical Methods in Medical Research - Tập 27 Số 8 - Trang 2264-2278 - 2018
Liang Li1, Tom Greene2, Bo Hu3
1Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
2Department of Population Health Sciences, University of Utah, Salt Lake City, UT, USA
3Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA

Tóm tắt

The time-dependent receiver operating characteristic curve is often used to study the diagnostic accuracy of a single continuous biomarker, measured at baseline, on the onset of a disease condition when the disease onset may occur at different times during the follow-up and hence may be right censored. Due to right censoring, the true disease onset status prior to the pre-specified time horizon may be unknown for some patients, which causes difficulty in calculating the time-dependent sensitivity and specificity. We propose to estimate the time-dependent sensitivity and specificity by weighting the censored data by the conditional probability of disease onset prior to the time horizon given the biomarker, the observed time to event, and the censoring indicator, with the weights calculated nonparametrically through a kernel regression on time to event. With this nonparametric weighting adjustment, we derive a novel, closed-form formula to calculate the area under the time-dependent receiver operating characteristic curve. We demonstrate through numerical study and theoretical arguments that the proposed method is insensitive to misspecification of the kernel bandwidth, produces unbiased and efficient estimators of time-dependent sensitivity and specificity, the area under the curve, and other estimands from the receiver operating characteristic curve, and outperforms several other published methods currently implemented in R packages.

Từ khóa


Tài liệu tham khảo

10.1093/oso/9780198509844.001.0001

10.1002/9780470906514

10.1111/j.0006-341X.2000.00337.x

R Core Team. R: A language and environment for statistical computing. 2013; R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/ (accessed 1 July 2016).

10.1002/sim.2299

Song X, 2008, Statis Sinica, 18, 947

Hung H, 2010, Can J Stat, 38, 826, 10.1002/cjs.10046

10.1002/bimj.201200045

10.1198/016214507000000149

10.1111/j.1467-9469.2009.00683.x

Blanche P, Latouche A and Viallon V. Time-dependent AUC with right-censored data: a survey study. In: Risk assessment and evaluation of predictions, Springer, http://arxiv.org/abs/1210.6805 (accessed 25 October 2012).

10.1111/j.1541-0420.2009.01375.x

10.1002/sim.4178

Jacqmin-Gadda H, 2014, Stat Meth Med Res, 1

10.1007/978-1-4899-4541-9

10.1148/radiology.143.1.7063747

10.1093/jnci/djn326

10.1158/1055-9965.EPI-14-1227

Dabrowska DM, 1987, Scand J Statist, 14, 181

10.1002/hep.1840200120

Fleming T, 1991, Counting processes and survival analysis

10.1007/978-1-4757-3294-8

Thông tin
Thông tin xuất bản

Statistical Methods in Medical Research

Tập 27 Số 8

2264-2278

Thông tin tác giả