Impact of model misspecification in shared frailty survival models.
Monte Carlo simulation study
correlated survival data
frailty
misspecification
shared frailty
Journal
Statistics in medicine
ISSN: 1097-0258
Titre abrégé: Stat Med
Pays: England
ID NLM: 8215016
Informations de publication
Date de publication:
15 10 2019
15 10 2019
Historique:
received:
18
10
2018
revised:
11
06
2019
accepted:
11
06
2019
pubmed:
23
7
2019
medline:
5
11
2020
entrez:
23
7
2019
Statut:
ppublish
Résumé
Survival models incorporating random effects to account for unmeasured heterogeneity are being increasingly used in biostatistical and applied research. Specifically, unmeasured covariates whose lack of inclusion in the model would lead to biased, inefficient results are commonly modeled by including a subject-specific (or cluster-specific) frailty term that follows a given distribution (eg, gamma or lognormal). Despite that, in the context of parametric frailty models, little is known about the impact of misspecifying the baseline hazard or the frailty distribution or both. Therefore, our aim is to quantify the impact of such misspecification in a wide variety of clinically plausible scenarios via Monte Carlo simulation, using open-source software readily available to applied researchers. We generate clustered survival data assuming various baseline hazard functions, including mixture distributions with turning points, and assess the impact of sample size, variance of the frailty, baseline hazard function, and frailty distribution. Models compared include standard parametric distributions and more flexible spline-based approaches; we also included semiparametric Cox models. The resulting bias can be clinically relevant. In conclusion, we highlight the importance of fitting models that are flexible enough and the importance of assessing model fit. We illustrate our conclusions with two applications using data on diabetic retinopathy and bladder cancer. Our results show the importance of assessing model fit with respect to the baseline hazard function and the distribution of the frailty: misspecifying the former leads to biased relative and absolute risk estimates, whereas misspecifying the latter affects absolute risk estimates and measures of heterogeneity.
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
4477-4502Subventions
Organisme : Medical Research Council
ID : MR/P015433/1
Pays : United Kingdom
Informations de copyright
© 2019 John Wiley & Sons, Ltd.
Références
Royston P, Parmar M. Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Statist Med. 2002;21(15):2175-2197. https://doi.org/10.1002/sim.1203
Rutherford M, Crowther M, Lambert P. The use of restricted cubic splines to approximate complex hazard functions in the analysis of time-to-event data: a simulation study. J Stat Comput Simul. 2015;85(4):777-793. https://doi.org/10.1080/00949655.2013.845890
Vaupel J, Manton K, Stallard E. The impact of heterogeneity in individual frailty on the dynamics of mortality. Demography. 1979;16(3):439-454. https://doi.org/10.2307/2061224
Lancaster T. Econometric methods for the duration of unemployment. Econometrica. 1979;47(4):939-956. https://doi.org/10.2307/1914140
Hougaard P. Life table methods for heterogeneous populations: distributions describing the heterogeneity. Biometrika. 1984;71(1):75-83. https://doi.org/10.2307/2336399
Hougaard P. Survival models for heterogeneous populations derived from stable distributions. Biometrika. 1986;73(2):387-396. https://doi.org/10.1093/biomet/73.2.387
Hougaard P. A class of multivariate failure time distributions. Biometrika. 1986;73(3):671-678. https://doi.org/10.1093/biomet/73.3.671
Gutierrez R. Parametric frailty and shared frailty survival models. Stata J. 2002;2(1):22-44.
Hougaard P. Frailty models for survival data. Lifetime Data Anal. 1995;1(3):255-273. https://doi.org/10.1007/BF00985760
Rondeau V, Filleul L, Joly P. Nested frailty models using maximum penalized likelihood estimation. Statist Med. 2006;25(23):4036-4052. https://doi.org/10.1002/sim.2510
Rondeau V, Michiels S, Liquet B, Pignon J. Investigating trial and treatment heterogeneity in an individual patient data meta-analysis of survival data by means of the penalized maximum likelihood approach. Statist Med. 2008;27(11):1894-1910. https://doi.org/10.1002/sim.3161
Rondeau V, Mathoulin-Pelissier S, Jacqmin-Gadda H, Brouste V, Soubeyran P. Joint frailty models for recurring events and death using maximum penalized likelihood estimation: application on cancer events. Biostatistics. 2006;8(4):708-721. https://doi.org/10.1093/biostatistics/kxl043
Rondeau V, Pignon J, Michiels S. A joint model for the dependence between clustered times to tumour progression and deaths: a meta-analysis of chemotherapy in head and neck cancer. Stat Methods Med Res. 2015;24(6):711-729. https://doi.org/10.1177/0962280211425578
Mazroui Y, Mathoulin-Pelissier S, Soubeyran P, Rondeau V. General joint frailty model for recurrent event data with a dependent terminal event: application to follicular lymphoma data. Statist Med. 2012;31(11-12):1162-1176. https://doi.org/10.1002/sim.4479
Ha I, Sylvester R, Legrand C, MacKenzie G. Frailty modelling for survival data from multi-centre clinical trials. Statist Med. 2011;30(17):2144-2159. https://doi.org/10.1002/sim.4250
Crowther M, Look M, Riley R. Multilevel mixed effects parametric survival models using adaptive Gauss-Hermite quadrature with application to recurrent events and individual participant data meta-analysis. Statist Med. 2014;33(22):3844-3858. https://doi.org/10.1002/sim.6191
Austin P. A tutorial on multilevel survival analysis: methods, models and applications. Int Stat Rev. 2017;85(2):185-203. https://doi.org/10.1111/insr.12214
Pickles A, Crouchley R. A comparison of frailty models for multivariate survival data. Statist Med. 1995;14(13):1447-1461. https://doi.org/10.1002/sim.4780141305
Glidden D, Vittinghoff E. Modelling clustered survival data from multicentre clinical trials. Statist Med. 2004;23(3):369-388. https://doi.org/10.1002/sim.1599
Lee K, Thompson S. Flexible parametric models for random-effects distributions. Statist Med. 2008;27(3):418-434. https://doi.org/10.1002/sim.2897
Liu X, Pawitan Y, Clements M. Generalized survival models for correlated time-to-event data. Statist Med. 2017;36(29):4743-4762. https://doi.org/10.1002/sim.7451
Ha I, Lee Y. Estimating frailty models via Poisson hierarchical generalized linear models. J Comput Graph Stat. 2003;12(3):663-681. https://doi.org/10.1198/1061860032256
Elbers C, Ridder G. True and spurious duration dependence: the identifiability of the proportional hazard model. Rev Econ Stud. 1982;49(3):403-409.
Bălan T. Advances in Frailty Models [PhD thesis]. Leiden, The Netherlands: Faculty of Medicine, Leiden University Medical Center (LUMC); 2018.
Andersson T, Dickman P, Eloranta S, Lambe M, Lambert P. Estimating the loss in expectation of life due to cancer using flexible parametric survival models. Statist Med. 2013;32(30):5286-5300. https://doi.org/10.1002/sim.5943
Duchateau L, Janssen P, Lindsey P, Legrand C, Nguti R, Sylvester R. The shared frailty model and the power for heterogeneity tests in multicenter trials. Comput Stat Data Anal. 2002;40(3):603-620. https://doi.org/10.1016/s0167-9473(02)00057-9
Diabetic Retinopathy Study Research Group. Preliminary report on effects of photocoagulation therapy. Am J Ophthalmol. 1976;81(4):383-396. https://doi.org/10.1016/0002-9394(76)90292-0
Huster W, Brookmeyer R, Self S. Modelling paired survival data with covariates. Biometrics. 1989;45(1):145-156. https://doi.org/10.2307/2532041
Sylvester R, van der Meijden A, Oosterlinck W, et al. Predicting recurrence and progression in individual patients with stage Ta-T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol. 2006;49(3):466-477.
Bălan T, Boonk S, Vermeer M, Putter H. Score test for association between recurrent events and a terminal event. Statist Med. 2016;35(18):3037-3048. https://doi.org/10.1002/sim.6913
Crowther M, Lambert P. A general framework for parametric survival analysis. Statist Med. 2014;33(30):5280-5297. https://doi.org/10.1002/sim.6300
Bender R, Augustin T, Blettner M. Generating survival times to simulate Cox proportional hazards models. Statist Med. 2005;24(11):1713-1723. https://doi.org/10.1002/sim.2059
Crowther M, Lambert P. Simulating biologically plausible complex survival data. Statist Med. 2013;32(23):4118-4134. https://doi.org/10.1002/sim.5823
Takahasi H, Mori M. Double Exponential Formulas for Numerical Integration. Vol. 9. Kyoto, Japan: Publication of RIMS, Kyoto University; 1974.
Bailey D. Tanh-Sinh High-Precision Quadrature. Technical Report. Berkeley, CA; 2006.
Liu Q, Pierce D. A note on Gauss-Hermite quadrature. Biometrika. 1994;81(3):624-629. https://doi.org/10.2307/2337136
Liu X, Pawitan Y, Clements M. Parametric and penalized generalized survival models. Stat Methods Med Res. 2018;27(5):1531-1546. https://doi.org/10.1177/0962280216664760
Rondeau V, Mazroui Y, Gonzalez J. frailtypack: an R package for the analysis of correlated survival data with frailty models using penalized likelihood estimation or parametrical estimation. J Stat Softw. 2012;47(4):1-28.
Burton A, Altman D, Royston P, Holder R. The design of simulation studies in medical statistics. Statist Med. 2006;25(24):4279-4292. https://doi.org/10.1002/sim.2673
Morris T, White I, Crowther M. Using simulation studies to evaluate statistical methods. Statist Med. 2019:1-29. https://doi.org/10.1002/sim.8086
White I. Simsum: analyses of simulation studies including Monte Carlo error. Stata J. 2010;10(3):369-385.
Dempster A, Laird N, Rubin D. Maximum likelihood from incomplete data via the EM algorithm. J R Stat Soc Series B Stat Methodol. 1977;39(1):1-38.
Ripatti S, Palmgren J. Estimation of multivariate frailty models using penalized partial likelihood. Biometrics. 2000;56(4):1016-1022. https://doi.org/10.1111/j.0006-341x.2000.01016.x
Hirsch K, Wienke A. Software for semiparametric shared gamma and log-normal frailty models: an overview. Comput Methods Programs Biomed. 2012;107(3):582-597. https://doi.org/10.1016/j.cmpb.2011.05.004
Monaco J, Gorfine M, Hsu L. General semiparametric shared frailty model: estimation and simulation with frailtySurv. J Stat Softw. 2018;86(4):1-42. https://doi.org/10.18637/jss.v086.i04
Ha I, Noh M, Kim J, Lee Y. FrailtyHL: Frailty Models via Hierarchical Likelihood. R package version 2.2. https://CRAN.r-project.org/package=frailtyHL. 2018.
Munda M, Rotolo F, Legrand C. parfm: parametric frailty models in R. J Stat Softw. 2012;51(11):1-20. https://doi.org/10.18637/jss.v051.i11
Schemper M, Smith T. A note on quantifying follow-up in studies of failure time. Contemp Clin Trials. 1996;17(4):343-346. https://doi.org/10.1016/0197-2456(96)00075-x
Vaida F, Blanchard S. Conditional Akaike information for mixed-effects models. Biometrika. 2005;92(2):351-370. https://doi.org/10.1093/biomet/92.2.351
Liang H, Wu H, Zou G. A note on conditional AIC for linear mixed-effects models. Biometrika. 2008;95(3):773-778. https://doi.org/10.1093/biomet/asn023
Ha I, Noh M, Lee Y. frailtyHL: a package for fitting frailty models with H-likelihood. R J. 2012;4(2):28-37. https://doi.org/10.32614/RJ-2012-010
Bower H, Crowther M, Rutherford M, et al. Capturing simple and complex time-dependent effects using flexible parametric survival models: a simulation study. Commun Stat Simul Comput. 2017. Submitted.
Hougaard P. Analysis of Multivariate Survival Data. New York, NY: Springer; 2000.
Amorim L, Cai J. Modelling recurrent events: a tutorial for analysis in epidemiology. Int J Epidemiol. 2015;44(1):324-333. https://doi.org/10.1093/ije/dyu222
Kuyken W, Warren F, Taylor R, et al. Efficacy of mindfulness-based cognitive therapy in prevention of depressive relapse: an individual patient data meta-analysis from randomized trials. JAMA Psychiatry. 2016;73(6):565-574. https://doi.org/10.1001/jamapsychiatry.2016.0076
Chrcanovic B, Kisch J, Albrektsson T, Wennerberg A. Bruxism and dental implant failures: a multilevel mixed effects parametric survival analysis approach. J Oral Rehabil. 2016;43(11):813-823. https://doi.org/10.1111/joor.12431
Gargiulo G, Windecker S, Da Costa B, et al. Short term versus long term dual antiplatelet therapy after implantation of drug eluting stent in patients with or without diabetes: systematic review and meta-analysis of individual participant data from randomised trials. BMJ. 2016;355:i5483. https://doi.org/10.1136/bmj.i5483
Devaraj S, Patel P. Attributing responsibility: hospitals account for 20% of variance in acute myocardial infarction patient mortality. J Healthc Qual; 38(1):52-61. https://doi.org/10.1097/JHQ.0000000000000008
Aoudjhane M, Labopin M, Gorin N, et al. Comparative outcome of reduced intensity and myeloablative conditioning regimen in HLA identical sibling allogeneic haematopoietic stem cell transplantation for patients older than 50 years of age with acute myeloblastic leukaemia: a retrospective survey from the Acute Leukemia Working Party (ALWP) of the European group for Blood and Marrow Transplantation (EBMT). Leukemia. 2005;19:2304-2312. https://doi.org/10.1038/sj.leu.2403967
Ha I, Lee Y, MacKenzie G. Model selection for multi-component frailty models. Statist Med. 2007;26:4790-4807. https://doi.org/10.1002/sim.2879
Donohue M, Overholser R, Xu R, Vaida F. Conditional Akaike information under generalized linear and proportional hazards mixed models. Biometrika. 2011;98(3):685-700. https://doi.org/10.1093/biomet/asr023
Reid N. A conversation with Sir David Cox. Stat Sci. 1994;9(3):439-455.
Nelson C, Lambert P, Squire I, Jones D. Flexible parametric models for relative survival, with application in coronary heart disease. Statist Med. 2007;26(30):5486-5498.
Searle S, Casella G, McCulloch C. Variance Components. Hoboken, NJ: John Wiley & Sons; 2008.
Skrondal A. Design and analysis of Monte Carlo experiments: attacking the conventional wisdom. Multivar Behav Res. 2000;35(2):137-167. https://doi.org/10.1207/S15327906MBR3502_1