Objective To assess the validity and effectiveness of parameter estimation using a time-dependent Weibull proportional hazards model for survival analysis containing partly interval censored data and explore the impact of different covariates on the results of analysis. Methods We established a time-dependent Weibull proportional hazards model using the Weibull distribution as the baseline hazard function of the model which incorporated time-varying covariates. Maximum likelihood estimation was employed to estimate the model parameters, which were obtained by optimization of the likelihood function. Results and Conclusion Numerical simulation results showed that with higher proportions of precise observations across different sample sizes and parameter settings, the proposed model resulted in improved accuracy of parameter estimation with coverage probabilities approximating the theoretical expectation of 95%. As the sample sizes increased, the parameter biases of the model tended to decrease. Experiments with empirical data further validated the effectiveness of the model. Compared with the failure time data for each precisely observed individual, additional interval-censored data helped to obtain more effective estimates of the regression parameters. Comparison with the Cox model that included time-varying covariates further demonstrated the effectiveness of the time-dependent Weibull proportional hazards model for parameter estimation in survival analysis with partly interval censored data.
LiJQ, MaJ. On hazard-based penalized likelihood estimation of accelerated failure time model with partly interval censoring[J]. Stat Methods Med Res, 2020, 29(12): 3804-17.
[2]
FauziNA, ElfakiF, AliMY. Some method on survival analysis via weibull model in the present of partly interval censored: a short review[J]. Int J Comput Sci Net, 2015, 15(4): 48-51.
[3]
Mohsin SaeedN, ElfakiFAM. Parametric weibull model based on imputations techniques for partly interval censored data[J]. Austrian J Stat, 2020, 49(3): 30-7.
[4]
JiangJJ, WangCJ, PanD, et al. Transformation models with informative partly interval-censored data[J]. Stat Comput, 2023, 34(1): 8.
OdellPM, AndersonKM, D'AgostinoRB. Maximum likelihood estimation for interval-censored data using a Weibull-based accelerated failure time model[J]. Biometrics, 1992, 48(3): 951-9.
[7]
Ramlau-HansenH, Chr Bang JespersenN, Kragh AndersenP, et al. Life insurance for insulin-dependent diabetics[J]. Scand Actuar J, 1987, 1987(1/2): 19-36.
[8]
KimJS. Maximum likelihood estimation for the proportional hazards model with partly interval-censored data[J]. J R Stat Soc Ser B Stat Methodol, 2003, 65(2): 489-502.
[9]
ElfakiFAM, AzramM, UsmanM. Parametric cox's model for partly interval-censored data with application to AIDS studies[J]. Int J Appl Phys Math, 2012: 352-4.
HarrellFE. Regression modeling strategies: with applications to linear models, logistic and ordinal regression, and survival analysis [M]. 2nd ed. Cham, Switzerland: Springer International AG, 2015.
[12]
PanC, CaiB. A Bayesian semiparametric mixed effects proportional hazards model for clustered partly interval-censored data[J]. Stat Model, 2024, 24(5): 459-79.
[13]
FisherLD, LinDY. Time-dependent covariates in the Cox proportional-hazards regression model[J]. Annu Rev Public Health, 1999, 20: 145-57.
[14]
AustinPC. Generating survival times to simulate Cox proportional hazards models with time-varying covariates[J]. Stat Med, 2012, 31(29): 3946-58.
[15]
PinskyPF, ZhuCS, KramerBS. Lung cancer risk by years since quitting in 30+ pack year smokers[J]. J Med Screen, 2015, 22(3): 151-7.
[16]
WebbA, MaJ. Cox models with time-varying covariates and partly-interval censoring-a maximum penalised likelihood approach[J]. Stat Med, 2023, 42(6): 815-33.
[17]
CrowleyJ, HuM. Covariance analysis of heart transplant survival data[J]. J Am Stat Assoc, 1977, 72(357): 27.
GuanZ. Maximum approximate Bernstein likelihood estimation in proportional hazard model for interval-censored data[J]. Stat Med, 2021, 40(3): 758-78.
[22]
AustinPC, FangJ, LeeDS. Using fractional polynomials and restricted cubic splines to model non‐proportional hazards or time‐varying covariate effects in the Cox regression model[J]. Stat Med, 2022, 41(3): 612-24.
[23]
XuR, AdakS. Survival analysis with time-varying relative risks: a tree-based approach[J]. Methods Inf Med, 2001, 40(2): 141-7.
[24]
AdelekeKA, AbiodunAA, IpinyomiRA. Semi-parametric non-proportional hazard model with time varying covariate[J]. J Mod App Stat Meth, 2015, 14(2): 68-87.
[25]
TherneauTM, GrambschPM. Modeling Survival Data: Extending the Cox Model[M]. New York, NY: Springer New York, 2000.
[26]
ThackhamM, MaJ. On maximum likelihood estimation of the semi-parametric Cox model with time-varying covariates[J]. J Appl Stat, 2020, 47(9): 1511-28.
[27]
ZhouQN, SunYQ, GilbertPB. Semiparametric regression analysis of partly interval-censored failure time data with application to an AIDS clinical trial[J]. Stat Med, 2021, 40(20): 4376-94.
[28]
GreenA, Borch-JohnsenK, AndersenPK, et al. Relative mortality of type 1 (insulin-dependent) diabetes in Denmark: 1933-1981[J]. Diabetologia, 1985, 28(6): 339-42.
[29]
HammerSM, KatzensteinDA, HughesMD, et al. A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with CD4 cell counts from 200 to 500 per cubic millimeter. AIDS Clinical Trials Group Study 175 Study Team[J]. N Engl J Med, 1996, 335(15): 1081-90.