We apply a maximal likelihood ratio test for the presence of multiple change-points in the covariate effects based on the Cox regression model. The covariate effect is assumed to change smoothly at one or more unknown change-points. The number of change-points is inferred by a sequential approach. Confidence intervals for the regression and change-point parameters are constructed by a bootstrap method based on Bernstein polynomials conditionally on the number of change-points. The methods are assessed by simulations and are applied to two datasets.
CoxDR.Regression models and life-tables. J R Stat Soc B1972;
34: 187–220.
2.
LiangKYSelfSGLiuX.The Cox proportional hazards model with change point: an epidemiologic application. Biometrics1990;
46: 783–793.
3.
LiuMLuWShaoY.A Monte Carlo approach for change-point detection in the Cox proportional hazards model.Stat Med2008;
27: 3894–3909.
4.
HePFangLSuZ.A sequential testing approach to detecting multiple change points in the proportional hazards model.Stat Med2013;
32: 1239–1245.
5.
JensenULütkebohmertC.A Cox-type regression model with change-points in the covariates.Lifetime Data Anal2008;
14: 267–285.
6.
LuoXBoyettJM.Estimations of a threshold parameter in Cox regression. Commun Stat-Theor Meth1997;
26: 2329–2346.
7.
PonsO.Estimation in a Cox regression model with a change-point according to a threshold in a covariate. Ann Stat2003;
31: 442–463.
8.
KosorokMRSongR.Inference under right censoring for transformation models with a change-point based on a covariate threshold. Ann Stat2007;
35: 957–989.
9.
DengY, et al.
Proportional hazards model with a change point for clustered event data.Biometrics2017;
73: 835–845.
10.
LamKFXuJXueH.Estimation of age effect with change-points on survival of cancer patients.Stat Med2018;
37: 1732–1743.
11.
Pastor-BarriusoRBanegasJRDaminJ, et al.
Systolic blood pressure, diastolic blood pressure, and pulse pressure: an evaluation of their joint effect on mortality.Ann Intern Med2003;
139: 731–739.
12.
ShapiroJSmithHSchaffnerF.Serum bilirubin: a prognostic factor in primary biliary cirrhosis.Gut1979;
20: 137–140.
13.
DaviesRB.Hypothesis testing when a nuisance parameter is present only under the alternative. Biometrika1987;
74: 33–43.
14.
MatthewsDEFarewellVTPykeR.Asymptotic score-statistic processes and tests for constant hazard against a change-point alternative. Ann Stat1985;
13: 583–591.
15.
LeeCYChenX, andLamKF.Testing for change-point in the covariate effects based on the Cox regression model.Stat Med2020;
39: 1473–1488.
16.
BreslowNE.Discussion of Professor Cox’s paper. J R Stat Soc B1972:
34: 216–217.
17.
KimKDemetsDL.Design and analysis of group sequential tests based on the type I error spending rate function. Biometrika1987;
74: 149–154.
ShaoJ.Bootstrap sample size in nonregular cases. Proc Am Math Soc1994;
122: 1251–1262.
20.
XuGSenBYingZ.Bootstrapping a change-point cox model for survival data.Electron J Stat2014;
8: 1345.
21.
LorentzGG.Bernstein polynomials.
New York, NY:
Chelesa Publishing Co., 1986.
22.
BurnhamKPAndersonDR.Model selection and multimodel inference: a practical information-theoretic approach.
New York, NY:
Springer Science & Business Media, 2003.
23.
FlemingTRHarringtonDP.Counting processes and survival analysis.
Hoboken, NJ:
John Wiley & Sons, 1991.
24.
SchmoorCOlschewskiMSchumacherM.Randomized and non-randomized patients in clinical trials: experiences with comprehensive cohort studies. Stat Med1996;
15: 263–271.
25.
SchumacherMBastertGBojarH, et al.
Randomized 2 × 2 trial evaluating hormonal treatment and the duration of chemotherapy in node-positive breast cancer patients. German breast cancer study group.J Clin Oncol1994;
12: 2086–2093.
26.
AdamiHOMalkerBHolmbergL, et al.
The relation between survival and age at diagnosis in breast cancer.New Engl J Med1986;
315: 559–563.
