Multistate models are useful for studying exposures that affect transitions among a set of health states. However, they can be challenging to apply when exposures are time-varying. We develop a multistate model and a method of likelihood construction that allows application of the model to data in which interventions or other exposures can be time-varying and an individual may to be exposed to multiple intervention conditions while progressing through states. The model includes cure proportions, reflecting the possibility that some individuals will never leave certain states. We apply the approach to analyze patient vaccination data from a stepped wedge design trial evaluating two interventions to increase uptake of human papillomavirus vaccination. The states are defined as the number of vaccine doses the patient has received. We model state transitions as a semi-Markov process and include cure proportions to account for individuals who will never leave a given state (e.g. never receive their next dose). Multistate models typically quantify intervention effects as hazard ratios contrasting the intensities of transitions between states in intervention versus control conditions. For multistate processes, another clinically meaningful outcome is the change in the percentage of the study population that has achieved a specific state (e.g. completion of all required doses) by a specific point in time due to an intervention. We present a method for quantifying intervention effects in this manner. We apply the model to both simulated and real-world data and also explore some conditions under which such models may give biased results.
Von CubeMSchumacherMWolkewitzM. Basic parametric analysis for a multi-state model in hospital epidemiology. BMC Med Res Methodol2017; 17: 111.
2.
WuGChangSChenT. A Bayesian random-effects Markov model for tumor progression in women with a family history of breast cancer. Biometrics2008; 64: 1231–1237.
3.
O’KeeffeATomBFarewellV. Mixture distributions in multi-state modelling: some considerations in a study of psoriatic arthritis. Stat Med2017; 32: 600–619.
4.
HusseyMHughesJ. Design and analysis of stepped wedge cluster randomized trials. Contemp Clin Trials2007; 28: 182–191.
5.
HemmingKTaljaardMForbesA. Analysis of cluster randomised stepped wedge trials with repeated cross-sectional samples. Trials2017; 18: 101.
6.
TaljaardMTeerenstraSIversN, et al. Substantial risks associated with few clusters in cluster randomized and stepped wedge designs. Clinical Trials2016; 13: 459–463.
7.
SundinPCrespiC. Power analysis for stepped wedge trials with multiple interventions. Stat Med2022; 41: 1498–1512.
8.
DilleyaSMillerKHuhaW. Human papillomavirus vaccination: ongoing challenges and future directions. Gynecol Oncol2020; 156: 498–502.
9.
DubèELabergeCGuayM, et al. Vaccine hesitancy: an overview. Hum Vaccines Immunotherapeutics2013; 9: 1763–1773.
10.
YiuSFarewellVTomB. Clustered multistate models with observation level random effects, mover–stayer effects and dynamic covariates: modelling transition intensities and sojourn times in a study of psoriatic arthritis. Appl Stat2017; 67: 481–500.
11.
BeesleyLTaylorJ. EM algorithms for fitting multistate cure models. Biostatistics2019; 20: 416–432.
12.
CassarlyCMartinRChimowitzM, et al. Assessing type I error and power of multistate Markov models for panel data—a simulation study. Commun Stat-Simu Comput2017; 46: 7040–7061.
13.
AralisHBrookmeyerR. A stochastic estimation procedure for intermittently-observed semi-Markov multistate models with back transitions. Stat Methods Med Res2019; 2: 770–787.
14.
Jackson CHetal. A comparison of two frameworks for multi-state modelling, applied to outcomes after hospital admissions with COVID-19. Stat Methods Med Res2022; 31: 1656–1674.
WalkerTElam-EvansLYankeyD, et al. National, regional, state, and selected local area vaccination coverage among adolescents aged 13–17 years—United States, 2018. Morbidity Mortality Weekly Report2019; 68: 718–723.
HsiehHChenTChangS. Assessing chronic disease progression using non-homogeneous exponential regression Markov models: an illustration using a selective breast cancer screening in Taiwan. Stat Med2002; 21: 3369–3382.
19.
ChenBGraceYCookR. Progressive multi-state models for informatively incomplete longitudinal data. J Stat Plan Inference2011; 141: 80–93.
20.
OthusMBarlogieBLeBlancM, et al. Cure models as a useful statistical tool for analyzing survival. Clin Cancer Res2012; 18: 3731–3736.
21.
YuBPengY. Mixture cure models for multivariate survival data. Comput Stat Data Anal2008; 52: 1524–1532.
22.
AustinP. Generating survival times to simulate Cox proportional hazards models with time-varying covariates. Stat Med2012; 31: 3946–3958.
23.
MaYJenkinsHSebastianiP, et al. Using cure models to estimate the serial interval of tuberculosis with limited follow-up. Pract Epidemiol2020; 189: 1421–1426.
24.
YuBTiwariR. A Bayesian approach to mixture cure models with spatial frailties for population-based cancer relative survival data. The Canadian Journal of Statistics / La Revue Canadienne de Statistique2012; 40: 40–54. http://www.jstor.org/stable/41724514.
25.
WangYTangYZhangJ. Bayesian approach for proportional hazards mixture cure model allowing non-curable competing risk. J Stat Comput Simul2020; 90: 638–656.
26.
Stan Development Team Stan Modeling Language Users Guide and Reference Manual. (2022), http://mc-stan.org/, Version 2.29.
27.
R Core Team R. A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, 2019. https://www.R-project.org/.
28.
Stan Development Team RStan: the R interface to Stan. (2020), http://mc-stan.org/, R package version 2.21.2.
TitmanA. Transition probability estimates for non-Markov multi-state models. Biometrics2015; 71: 1034–1041.
32.
CookRLawlessJ. Multistate Models for the Analysis of Life History Data. New York, NY: Chapman and Hall/CRC, 2018.
33.
SamaniegoF. A Comparison of the Bayesian and Frequentist Approaches to Estimation. New York, NY: Springer, 2010.
34.
DucrosaFPamphileP. Bayesian estimation of Weibull mixture in heavily censored data setting. Reliab Eng Syst Saf2018; 180: 453–462.
35.
LinLHuangL. Connections between cure rates and survival probabilities in proportional hazards models. Stat2019; 8: e255.
36.
KearnsBStevensonMTriantafyllopoulosK, et al. The extrapolation performance of survival models for data with a cure fraction: a simulation study. Value Health2021; 24: 1634–1642.
37.
StedmanMFeuerEMariottoA. Current estimates of the cure fraction: a feasibility study of statistical cure for breast and colorectal cancer. J Nat Cancer Inst Monogr2014; 2014: 244–254.
OthusMBansalAErbaH, et al. Bias in mean survival from fitting cure models with limited follow-up. Value Health2020; 8: 1034–1039.
40.
Escobar-BachMMallerRVan KeilgomI, et al. Estimation of the cure rate for distributions in the Gumbel maximum domain of attraction under insufficient follow-up. Biometrika2021; 109: 243–256.
41.
FelizziFParachaNPÖhlmannJ, et al. Mixture cure models in oncology: a tutorial and practical guidance. PharmacoEconomics-Open2021; 5: 143–155.
42.
BetenskyRMandelM. Recognizing the problem of delayed entry in time-to-event studies: better late than never for clinical neuroscientists. Ann Neurol2015; 78: 839–844.
43.
OmerMEBakarMAAdamM, et al. Utilization of a mixture cure rate model based on the generalized modified Weibull distribution for the analysis of leukemia patients. Asian Pac J Cancer Prev2021; 22: 1045.
DongQPengYLiP. Time to delisted status for listed firms in Chinese stock markets: an analysis using a mixture cure model with time-varying covariates. J Oper Res Soc2021; 73: 2358–2369.
46.
BerettaAHeuchenneC. Variable selection in proportional hazards cure model with time-varying covariates, application to US bank failures. J Appl Stat2019; 46: 1529–1549.
47.
DirickLBellottiTClaeskensG, et al. Macro-economic factors in credit risk calculations: including time-varying covariates in mixture cure models. J Bus Econ Stat2019; 37: 40–53.
48.
RondeauVSchaffnerECorbiereF, et al. Cure frailty models for survival data: application to recurrences for breast cancer and to hospital readmissions for colorectal cancer. Stat Methods Med Res2013; 22: 243–260.
