Abstract
In many clinical studies that involve follow-up, it is common to observe one or more sequences of longitudinal measurements, as well as one or more time to event outcomes. A competing risks situation arises when the probability of occurrence of one event is altered/hindered by another time to event. Recently, there has been much attention paid to the joint analysis of a single longitudinal response and a single time to event outcome, when the missing data mechanism in the longitudinal process is non-ignorable. We, in this paper, propose an extension where multiple longitudinal responses are jointly modeled with competing risks (multiple time to events). Our shared parameter joint model consists of a system of multiphase non-linear mixed effects sub-models for the multiple longitudinal responses, and a system of cause-specific non-proportional hazards frailty sub-models for competing risks, with associations among multiple longitudinal responses and competing risks modeled using latent parameters. The joint model is applied to a data set of patients who are on mechanical circulatory support and are awaiting heart transplant, using readily available software. While on the mechanical circulatory support, patient liver and renal functions may worsen and these in turn may influence one of the two possible competing outcomes: (i) death before transplant; (ii) transplant. In one application, we propose a system of multiphase cause-specific non-proportional hazard sub-model where frailty can be time varying. Performance under different scenarios was assessed using simulation studies. By using the proposed joint modeling of the multiphase sub-models, one can identify: (i) non-linear trends in multiple longitudinal outcomes; (ii) time-varying hazards and cumulative incidence functions of the competing risks; (iii) identify risk factors for the both types of outcomes, where the effect may or may not change with time; and (iv) assess the association between multiple longitudinal and competing risks outcomes, where the association may or may not change with time.
Similar content being viewed by others
References
Aalen O, Johansen S (1978) An empirical transition matrix for nonhomogeneous Markov chains based on censored observations. Scand J Stat 5:141–150
Aksoy O, Cam A, Goel SS, Houghtaling PL, Williams S, Ruiz-Rodriguez E, Menon V, Kapadia SR, Tuzcu EM, Blackstone EH, Griffin BP (2012) Do bisphosphonates slow the progression of aortic stenosis? J Am Coll Cardiol 59:1452–1459
Andrinopoulou E, Rizopoulos D, Takkenberg JJB, Lesaffrea E (2014) Joint modeling of two longitudinal outcomes and competing risk data. Stat Med 33:3167–78
Asar O, Ritchie J, Kalra PA, Diggle P (2015) Joint modelling of repeated measurement and time-to-event data: an introductory tutorial. Int J Epidemiol 44:334344
Banga A, Gildea T, Rajeswaran J, Rokadia H, Blackstone EH, Stoller JK (2014) The natural history of lung function after lung transplantation for a1-antitrypsin deficiency. Am J Respir Crit Care Med 190:274–281
Beach J, Mihaljevic T, Rajeswaran J, Marwick T, Edwards ST, Nowicki ER (2014) Ventricular hypertrophy and left atrial dilatation persist and are associated with reduced survival after valve replacement for aortic stenosis. J Thorac Cardiovasc Surg 147:362–9
Blackstone EH, Naftel DC, Turner ME Jr (1986) The decomposition of time-varying hazard into phases, each incorporating a separate stream of concomitant information. J Am Stat Assoc 81:615–624
Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical information-theoretic approach, 2nd edn. Springer, New York
Bycott P, Taylor J (1998) A comparison of smoothing techniques for CD4 data measured with error in a time-dependent Cox-proportional hazards model. Stat Med 17:2061–2077
Chena LM, Ibrahima IG, Chub H (2011) Sample size and power determination in joint modeling of longitudinal and survival data. Stat Med 30:22952309
Crowder MJ (2001) Classical competing risks. Chapman & Hall/CRC, Boca Raton
Davidian M, Gilitinan DM (1995) Nonlinear models for repeated measurement data. Chapman & Hall, New York
Diggle PJ, Sousa I, Chetwynd AG (2008) Joint modeling of repeated measurements and time-to-event outcome: the fourth Armitage lecture. Stat Med 27:2981–2998
Elashoff RM, Li G, Li N (2007) An approach to joint analysis of longitudinal measurements and competing risks failure time data. Stat Med 26:2813–2835
Elashoff RM, Li G, Li N (2008) A joint model for longitudinal measurements and survival data in the presence of multiple failure types. Biometrics 64:762–771
Gillinov AM, Bhavani S, Blackstone EH, Rajeswaran J, Svensson LG, Navia JL (2006) Surgery for permanent atrial fibrillation: impact of patient factors and lesion set. Ann Thorac Surg 82:502–14
Hazelrig JB, Turner ME Jr, Blackstone EH (1982) Parametric survival analysis combining longitudinal and cross-sectional censored and interval-censored data with concomitant information. Biometrics 39:1–5
Hu W, Li G, Li N (2009) A Bayesian approach to joint analysis of longitudinal measurements and competing risks failure time data. Stat Med 28:1601–1619
Kalbfleisch JD, Prentice RL (2002) Statistical analysis of failure time data, 2nd edn. Wiley, Hoboken
Krol A, Mauguen A, Mazroui Y, Laurent A, Michiels S, Rondeau V (2017) Tutorial in joint modeling and prediction: a statistical software for correlated longitudinal outcomes. Recurrent events and a terminal event. J Stat Softw 81(3):1–52
Leeaphorn N, Sampaio MS, Natal N, Mehrnia A, Kamgar M, Huang E, Kalantar-Zadeh K, Kaplan B, Bunnapradist S (2014) Renal transplant outcomes in waitlist candidates with a previous inactive status due to being temporarily too sick. Clin Transpl 117–124
Li N, Elashoff RM, Li G (2009) Robust joint modeling of longitudinal measurements and competing risks failure time data. Biom J 51:19–30
Li N, Elashoff RM, Li G, Saver J (2010) Joint modeling of longitudinal ordinal data and competing risks survival times and analysis of the NINDS rt-PA stroke trial. Stat Med 29:546–557
Lindstrom MJ, Bates DM (1990) Nonlinear mixed effects models for repeated measures data. Biometrics 46:673–687
Mason DP, Rajeswaran J, Murthy SC, McNeill AM, Budev MM, Mehta AC, Pettersson GB, Blackstone EH (2008) Spirometry after transplantation: how much better are two lungs than one? Ann Thorac Surg 85:1193–201
Mbogning C, Bleakley K, Lavielle M (2015) Joint modeling of longitudinal and repeated time-to-event data using nonlinear mixed-effects models and the SAEM algorithm. J Stat Comput Simul 85:15121528
Oz MC, Goldstein DJ, Pepino P, Weinberg AD, Thompson SM, Catanese KA, Vargo RL, McCarthy PM, Rose EA, Levin HR (1995) Screening scale predicts patients successfully receiving long-term implantable left ventricular assist devices. Circulation 92:169–173
Pande A, Li L, Rajeswaran J, Ehrlinger J, Kogalur UB, Blackstone EH, Ishwaran H (2017) Boosted multivariate trees for longitudinal data. Mach Learn 106:277–305
Pinheiro JC, Bates DM (1995) Approximations to the log-likelihood function in a nonlinear mixed-effects model. J Comput Graph Stat 4:12–35
Pintilie M (2006) Competing risks: a practical perspective. Wiley, Chichester
Prentice RL, Kalbfleisch JD, Peterson AV Jr, Flournoy N, Farewell VT, Breslow NE (1978) The analysis of failure time in the presence of competing risks. Biometrics 34:541–554
Putter H, Fiocco M, Geskus RB (2007) Competing risks and multi-state models. Stat Med 26:2389–2430
R Core Team (2016) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org/
Raboud J, Reid N, Coates RA, Farewell VT (1993) Estimating risks of progression to AIDS when covariates are measured with error. J R Stat Soc Ser A 156:393–406
Rajeswaran J, Blackstone EH (2017) A multiphase non-linear mixed effects model: an application to spirometry after lung transplantation. Stat Methods Med Res 26:2142
Reinhartz O, Farrar DJ, Hershon JH, Avery GJ Jr, Haeusslein EA, Hill JD (1998) Importance of liver function as a predictor of survival in patients supported with thoratec ventricular assist devices as a bridge to transplantation. J Thorac Cardiovasc Surg 116:633–40
Rizopoulos D (2016) The R Package JMbayes for fitting joint models for longitudinal and time-to-event data using MCMC. J Stat Softw 72:145
Rizopoulos D (2017) JM: joint modeling of longitudinal and survival data. R package version 1.4-7. https://CRAN.R-project.org/package=JM
Rubin DB (1976) Inference and missing data. Biometrika 63:581–592
Schluchter MD (1992) Methods for analysis of informatively censored longitudinal data. Stat Med 11:1861–1870
Shreenivas SS, Rame JE, Jessup M (2010) Mechanical circulatory support as a bridge to transplant or for destination therapy. Curr Heart Fail Rep 7:159–166
Song X, Davidian M, Tsiatis AA (2002) A semiparametric likelihood approach to joint modeling of longitudinal and time-to-event data. Biometrics 58:742–753
Tsiatis AA, Davidian M (1995) Joint modeling of longitudinal and time-to-event data: an overview. Stat Sin 14:809–834
Tsiatis AA, DeGruttola V, Wulfsohn MS (1995) Modeling relationship of survival to longitudinal data measured with error. Applications to survival and CD4 counts in patients with AIDS. J Am Stat Assoc 429:27–37
Turner ME, Hazelrig JB, Blackstone EH (1982) Bounded survival. Math Biosci 59:33–46
Verbeke G, Fieuws S, Molenbergs G, Davidian M (2014) The analysis of multivariate longitudinal data: a review. Stat Methods Med Res 23:42–59
Vonesh EF (1996) A note on the use of Laplace’s approximation for nonlinear mixed-effects models. Biometrika 83:447–452
Vonesh EF, Chinchilli VM (1997) Linear and nonlinear models for the analysis of repeated measurement. Marcel Dekker, New York
Williamson PR, Kolamunnage-Done R, Philipson P, Marson AG (2008) Joint modeling of longitudinal and competing risks data (2008). Stat Med 27:6426–6438
Wu L (2002) A joint model for nonlinear mixed-effects models with censoring and covariates measured with error, with application to AIDS studies. J Am Stat Assoc 97:955–964
Wulfsohn MS, Tsiatis AA (1997) A joint model for survival and longitudinal data measured with error. Biometrics 53:330–339
Acknowledgements
The first author would like to thank Dr. Mark Schluchter, Department of Epidemiology and Biostatistics, Case Western Reserve University, Cleveland, for his valuable comments. This research was partially supported by a grant from the National Institutes of Health—NATIONAL HEART, LUNG, AND BLOOD INSTITUTE, Grant Number: 1R01HL103552.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Non-Linear Time Function
We now briefly describe the following non-linear equation to model the non-linear (or linear) temporal trend of the longitudinal process (Sect. 2—Eq. 2). Development of the generic equation (12) started with differential equation [17, 45] that was formed from different dynamics models such as biochemical reaction rate, allometric growth, and population growth. They also found that this generic equation has relationship with some statistical distributions. By setting \(\nu \) to \(\pm \,1\) or limiting case, \(0/ \infty \), special and simpler cases of models are found. All these linear and various non-linear models are nested by dint of the common generic differential equation. This equation is originally used by Hazelrig et al. [17] to model cumulative mortality and then later used by Blackstone et al. [7] in multiphase hazard modeling (see “Data Analysis”—Sect. 4.2). In their modeling, both had used a 3-parameter time function. Since our data analysis experience [2, 5, 6, 16, 25] showed that a reduced 2-parameter equation adequately fits the longitudinal process, we use the following family of 2-parameter equation:
where \(\nu \ne 0\) is a shaping parameter, and \(\kappa >0\) is a scaling parameter. In survival terminology, Blackstone et al. [7] called \(\kappa \) as the half life time. Natural constraints on G are that \(G(0, \varvec{{\Theta }})=0\) and \(G(t, \varvec{{\Theta }}) \rightarrow 1\) as \(t \rightarrow \infty \). Note that, when \(\nu >0, \; G(t, \varvec{{\Theta }})\) simplifies as
and when \(\nu <0\),
Blackstone et al. [7] in their multiphase hazard modeling approach used \(g(t, \varvec{{\Theta }})=\frac{\partial G(t, \varvec{{\Theta }})}{\partial t}\) to model early phase of hazard and \(h(t, \varvec{{\Theta }})=\frac{g(t, \varvec{{\Theta }})}{1-G(t, \varvec{{\Theta }})}\) to fit late phase of the hazard. The motivation to use these equations, in survival terminology, is that if \(G(t, \varvec{{\Theta }})\) is cumulative hazard \(g(t, \varvec{{\Theta }})\) is the hazard. Note that, since \(G(t, \varvec{{\Theta }})\) is bounded above by 1, if one uses \(G(t, \varvec{{\Theta }})\) as cumulative hazard, it forms an incomplete time distribution. Suppose \(G(t, \varvec{{\Theta }})\) is cumulative distribution, and \(h(t, \varvec{{\Theta }})\) is a proper hazard function. We can use \(G(t, \varvec{{\Theta }})\) or any of its transformation or combination of transformations to fit a longitudinal process. In our data analysis experience in fitting longitudinal process, we found that, in most cases, \(g(t, \varvec{{\Theta }})\) can be used as early time function \(T(t, \varvec{{\Theta }})\) and \(h(t, \varvec{{\Theta }})\) can be used as late time function \(T(t, \varvec{{\Theta }})\). Examples of different shapes of \(T(t, \varvec{{\Theta }})=g(t, \varvec{{\Theta }})\) and \(T(t, \varvec{{\Theta }})=h(t, \varvec{{\Theta }})\) are given in Figs. 6 and 7.
Appendix B: Temporal Decomposition
We illustrate the temporal decompositions of the bivariate longitudinal process that resulted in the overall temporal trends depicted in Fig. 3.
For GFR, there is an early peaking phase followed by a late increasing phase (Top—Fig. 8), and for Log(Bilirubin) the analysis yielded an early decreasing phase followed by an increasing phase that plateaued after 4 months (Bottom—Fig. 8).
Rights and permissions
About this article
Cite this article
Rajeswaran, J., Blackstone, E.H. & Barnard, J. Joint Modeling of Multivariate Longitudinal Data and Competing Risks Using Multiphase Sub-models. Stat Biosci 10, 651–685 (2018). https://doi.org/10.1007/s12561-018-9223-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12561-018-9223-6