Longitudinal data are often available in cohort studies and clinical settings, such as covariates collected at cohort follow-up visits or prescriptions captured in electronic health records. Such longitudinal information, if correlates with the health event of interest, may be incorporated to dynamically predict the probability of a health event with better precision. Landmarking is a popular approach to dynamic prediction. There are well-established methods for landmarking using full cohort data, but collecting data on all cohort members may not be feasible when resource is limited. Instead, one may select a subset of the cohort using subsampling designs, and only collect data on this subset. In this work, we present conditional likelihood and inverse-probability weighted methods for landmarking using data from cohort subsampling designs, and discuss considerations for choosing a particular method. Simulations are conducted to evaluate the applicability of the methods and their predictive performance in different scenarios. Results show that our methods have similar predictive performance to the full cohort analysis but only use small fractions of the full cohort data. We use real nested case-control data from the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial to illustrate the methods.
PhungMTTin TinSElwoodJM. Prognostic models for breast cancer: a systematic review. BMC Cancer2019; 19: 1–18.
2.
KlatteTRossiSHStewartGD. Prognostic factors and prognostic models for renal cell carcinoma: a literature review. World J Urol2018; 36: 1943–1952.
3.
ViganòADorganMBuckinghamJ, et al.Survival prediction in terminal cancer patients: a systematic review of the medical literature. Palliat Med2000; 14: 363–374.
4.
KimWRMannalitharaAHeimbachJK, et al.MELD 3.0: The model for end-stage liver disease updated for the modern era. Gastroenterology2021; 161: 1887–1895.
5.
KamathPSKimWR. The model for end-stage liver disease (MELD). Hepatology2007; 45: 797–805.
6.
KamathPSWiesnerRHMalinchocM, et al.A model to predict survival in patients with end–stage liver disease. Hepatology2001; 33: 464–470.
7.
FisherLDLinDY. Time-dependent covariates in the Cox proportional-hazards regression model. Annu Rev Public Health1999; 20: 145–157.
8.
SureshKTaylorJMSprattDE, et al.Comparison of joint modeling and landmarking for dynamic prediction under an illness-death model. Biom J2017; 59: 1277–1300.
9.
PutterH, et al.Landmarking. In: KleinJPVan HouwelingenHCIbrahimJG (eds) Handbook of survival analysis. New York: Chapman and Hall/CRC, 2013, pp.441–456.
10.
TsiatisAADavidianM. Joint modeling of longitudinal and time-to-event data: an overview. Stat Sin2004; 14: 809–834.
11.
IbrahimJGChuHChenLM. Basic concepts and methods for joint models of longitudinal and survival data. J Clin Oncol2010; 28: 2796–2801.
12.
Lawrence GouldABoyeMECrowtherMJ, et al.Joint modeling of survival and longitudinal non-survival data: current methods and issues. Report of the DIA Bayesian joint modeling working group. Stat Med2015; 34: 2181–2195.
13.
AndersonJRCainKCGelberRD. Analysis of survival by tumor response. J Clin Oncol1983; 1: 710–719.
14.
LiWLiLAstorBC. A comparison of two approaches to dynamic prediction: joint modeling and landmark modeling. Stat Med2023; 42: 2101–2115.
15.
RizopoulosDMolenberghsGLesaffreEM. Dynamic predictions with time-dependent covariates in survival analysis using joint modeling and landmarking. Biom J2017; 59: 1261–1276.
16.
WacholderS. Practical considerations in choosing between the case-cohort and nested case-control designs. Epidemiology1991; 2: 155–158.
17.
FuRGilbertPB. Joint modeling of longitudinal and survival data with the Cox model and two-phase sampling. Lifetime Data Anal2017; 23: 136–159.
18.
YoonSHVandalARivera-RodriguezC. Weight calibration in the joint modelling of medical cost and mortality. Stat Methods Med Res2024; 33: 728–742.
19.
Van HouwelingenHC. Dynamic prediction by landmarking in event history analysis. Scandinavian J Statistics2007; 34: 70–85.
ZhuYLiLHuangX. On the landmark survival model for dynamic prediction of event occurrence using longitudinal data. In: ZhaoYChenDG (eds) New Front Biostat Bioinform. Switzerland: Springer, 2018, pp.387–401.
22.
HossainZKhondokerMInitiativeADN. Comparison of joint modelling and landmarking approaches for dynamic prediction using bootstrap simulation. Bull Malays Math Sci Soc2022; 45: 301–314.
23.
BorganØSamuelsenSO. Cohort sampling for time-to-event data: an overview. In: BorganØBreslowNChatterjeeN, et al. (eds) Handb Statist Methods Case-Control Stud. New York: Chapman & Hall/CRC, 2018, pp.285–301.
24.
ThomasDC. Appendum to “methods of cohort analysis: appraisal by application to asbestos mining,” by Liddell, FDK, McDonald, JC and Thomas, DC. J R Statist Soc A (General)1977; 140: 469–491.
25.
ThomasDC. Design issues in case-control studies. In: BorganØBreslowNChatterjeeN, et al. (eds) Handbook of statistical methods for case-control studies. New York: Chapman and Hall/CRC, 2018, pp.15–37.
26.
SamuelsenSO. A psudolikelihood approach to analysis of nested case-control studies. Biometrika1997; 84: 379–394.
27.
PrenticeRL. A case-cohort design for epidemiologic cohort studies and disease prevention trials. Biometrika1986; 73: 1–11.
28.
BorganOLangholzBSamuelsenSO, et al.Exposure stratified case-cohort designs. Lifetime Data Anal2000; 6: 39–58.
29.
KangSCaiJ. Marginal hazards model for case-cohort studies with multiple disease outcomes. Biometrika2009; 96: 887–901.
30.
KalbfleischJLawlessJ. Likelihood analysis of multi-state models for disease incidence and mortality. Stat Med1988; 7: 149–160.
31.
GoldsteinLLangholzB. Asymptotic theory for nested case-control sampling in the Cox regression model. Ann Statist1992; 20: 1903–1928.
32.
LangholzBBorganØ. Estimation of absolute risk from nested case-control data. Biometrics1997; 53: 767–774.
33.
ChenKLoS-H. Case-cohort and case-control analysis with Cox’s model. Biometrika1999; 86: 755–764.
34.
BorganØSamuelsenSO, et al.Nested case-control and case-cohort studies. In: KleinJPVan HouwelingenHCIbrahimJG (eds) Handbook of survival analysis. Boca Raton: Chapman and Hall/CRC, 2013, pp.343–367.
35.
ShinYEPfeifferRMGraubardBI, et al.Weight calibration to improve the efficiency of pure risk estimates from case-control samples nested in a cohort. Biometrics2020; 76: 1087–1097.
36.
StøerNCSamuelsenSO. multipleNCC: inverse probability weighting of nested case-control data. R J2016; 8: 5.
37.
ChangYIvanovaAAlbanesD, et al.Pooling controls from nested case–control studies with the proportional risks model. Biostatistics2025; 26: kxae032.
38.
MaziarzMHeagertyPCaiT, et al.On longitudinal prediction with time-to-event outcome: comparison of modeling options. Biometrics2017; 73: 83–93.
39.
RobinXTurckNHainardA, et al.pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinform2011; 12: 1–8.
40.
GrafESchmoorCSauerbreiW, et al.Assessment and comparison of prognostic classification schemes for survival data. Stat Med1999; 18: 2529–2545.
AhnJPetersUAlbanesD, et al.Serum vitamin D concentration and prostate cancer risk: a nested case-control study. J National Cancer Inst2008; 100: 796–804.
43.
GaoXHudgensMGZouF. Case-Cohort studies with time-dependent covariates and interval-censored outcomeEmerging topics in modeling interval-censored survival data. Switzerland: Springer, 2022, pp.221–234.
44.
VinogradovaYCouplandCHippisley-CoxJ. Exposure to statins and risk of common cancers: a series of nested case-control studies. BMC Cancer2011; 11: 1–12.
45.
PutterHvan HouwelingenHC. Landmarking 2.0: bridging the gap between joint models and landmarking. Stat Med2022; 41: 1901–1917.
46.
SignorelliMRetifS. Benchmarking multi-step methods for the dynamic prediction of survival with numerous longitudinal predictors. arXiv preprint arXiv:240314336v2.
47.
KeoghRHSeamanSRBarrettJK, et al.Dynamic prediction of survival in cystic fibrosis: a landmarking analysis using UK patient registry data. Epidemiology2019; 30: 29–37.
48.
GomonDPutterHFioccoM, et al.Dynamic prediction of survival using multivariate functional principal component analysis: a strict landmarking approach. Stat Methods Med Res2024; 33: 256–272.
49.
PutterHSpitoniC. Non-parametric estimation of transition probabilities in non-Markov multi-state models: the landmark Aalen–Johansen estimator. Stat Methods Med Res2018; 27: 2081–2092.
50.
HoffRPutterHMehlumIS, et al.Landmark estimation of transition probabilities in non-Markov multi-state models with covariates. Lifetime Data Anal2019; 25: 660–680.
51.
Van HouwelingenHPutterH. Dynamic prediction in clinical survival analysis. Boca Raton: CRC Press, 2011.
52.
TsengCLiuM. Joint modeling of survival data and longitudinal measurements under nested case-control sampling. Stat Biopharm Res2009; 1: 415–423.
53.
BaartSJBoersmaERizopoulosD. Joint models for longitudinal and time-to-event data in a case-cohort design. Stat Med2019; 38: 2269–2281.
54.
PickettKLSureshKCampbellKR, et al.Random survival forests for dynamic predictions of a time-to-event outcome using a longitudinal biomarker. BMC Med Res Methodol2021; 21: 1–14.
55.
WangYXieJZhaoX. Deepsurv landmarking: a deep learning approach for dynamic survival analysis with longitudinal data. J Stat Comput Simul2025; 95: 186–207.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
