Pith. sign in

REVIEW 3 minor 29 references

A Bias-Corrected Two-Stage Approach for Joint Modelling of Multidimensional Longitudinal HRQoL and Survival Data

T0 review · 0 major / 3 minor · reviewed 2026-06-26 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A slope-corrected two-stage method reduces bias in joint models of multidimensional quality-of-life trajectories and survival times.

desk verdict The SC2S procedure gives a workable bias fix for multidimensional HRQoL-survival joint models that tracks full Bayesian estimation at lower cost. read the letter →

arxiv 2606.23146 v1 pith:2JIYPC2B submitted 2026-06-22 stat.ME stat.AP

Hortense Doms , Philippe Lambert , Catherine Legrand This is my paper · ORCID
classification stat.MEstat.AP
keywords jointmodelslongitudinaldatasurvivalanalysisHRQoLbiascorrectiontwo-stageestimationordinallatenttraitmodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Joint models link repeated health-related quality-of-life scores to survival risk so that dropout from death is properly accounted for, yet full joint fitting grows too slow once multiple questionnaire dimensions and random effects enter the picture. The paper introduces a slope-corrected two-stage procedure that first fits the longitudinal model, then feeds its information into the survival model through informative priors on the random effects and re-estimates the slope parameters. This keeps most of the speed of separate two-stage fitting while cutting the bias that arises when the dependence between the processes is ignored. Simulation studies show the resulting estimates stay close to those obtained from fully joint Bayesian estimation. An application to glioblastoma patients confirms that the method delivers usable accuracy gains without the full computational load.

What carries the argument

The slope-corrected two-stage (SC2S) procedure, which transfers longitudinal information via informative priors on random effects and re-estimates slope parameters to correct bias from ignored dependence.

What would settle it

A simulation study in which the SC2S estimates show no reduction in bias relative to a standard two-stage fit when the true dependence between longitudinal trajectories and survival is known and nonzero.

Watch

Extended reading notes

Core claim

The slope-corrected two-stage (SC2S) approach propagates longitudinal information to the survival model through informative priors on the random effects while additionally re-estimating longitudinal slope parameters; this substantially reduces bias in both the longitudinal and survival submodels and closely approximates fully joint Bayesian estimation while retaining much of the computational efficiency of two-stage procedures.

Load-bearing premise

That passing longitudinal information through informative priors on random effects and then re-estimating slopes is sufficient to remove bias caused by ignoring dependence between the processes.

Editorial extensions

If this is right

  • Bias drops in both the longitudinal HRQoL submodel and the survival submodel.
  • Computational cost stays far below that of full joint estimation even with multiple latent dimensions.
  • The method handles multivariate ordinal longitudinal data within a latent trait framework.
  • Results in the glioblastoma application match full joint estimates closely enough for practical use.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same information-passing step could be tested in other joint modeling settings where full likelihoods become intractable.
  • Researchers facing large questionnaires might adopt the procedure instead of ignoring dependence or accepting long run times.
  • The approach suggests that targeted re-estimation of a few key parameters can substitute for complete joint likelihood evaluation in bias correction.
  • If the method scales to higher-dimensional data, it could open routine joint analysis for trials that currently use only separate models.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 3 minor

Summary. The manuscript proposes a slope-corrected two-stage (SC2S) approach for jointly modeling multidimensional longitudinal ordinal HRQoL data and survival outcomes in a latent trait framework. The SC2S method obtains random-effect estimates in stage 1, feeds them as informative priors into the survival submodel, and additionally re-estimates longitudinal slope parameters to mitigate bias induced by ignoring the dependence between the processes. Simulation studies under the data-generating model and a real-data application to progressive glioblastoma patients are presented to show that SC2S substantially reduces bias relative to standard two-stage procedures while closely approximating fully joint Bayesian estimation at lower computational cost.

Significance. If the performance claims hold, the SC2S procedure supplies a practical, scalable alternative for joint modeling when the number of latent dimensions and random effects renders full joint likelihood maximization or MCMC prohibitive. By preserving most of the efficiency of two-stage methods while recovering much of the bias reduction of joint estimation, the approach could enable routine incorporation of multidimensional HRQoL trajectories into survival analyses in oncology settings where computational resources are limited.

minor comments (3)
  1. [Abstract] Abstract: the statement that SC2S 'substantially reduces bias' would be strengthened by a parenthetical reference to the specific bias or MSE reductions reported in the simulation tables.
  2. [Section 4] Section 4 (Simulation studies): clarify whether the re-estimation of slope parameters is performed inside the second-stage survival model or as a separate post-processing step, and state the convergence criteria used for the MCMC chains.
  3. [Simulation results] Table 2 or equivalent results table: include the effective sample size or computation time per replication so readers can directly compare the efficiency gain against the joint Bayesian benchmark.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of our manuscript, the accurate summary of the SC2S approach, and the recommendation for minor revision. No major comments were raised in the report.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; SC2S is an independent algorithmic correction validated externally

full rationale

The manuscript introduces the SC2S procedure as a distinct two-stage algorithm that feeds stage-1 random-effect posteriors as informative priors into the survival submodel and then re-estimates longitudinal slopes. Performance claims rest on Monte-Carlo simulations under the data-generating model and a glioblastoma application, both of which compare the method to a fully joint Bayesian benchmark. No equations, uniqueness theorems, or self-citations are shown that would make the bias-reduction result equivalent to its own inputs by construction. The derivation chain is therefore self-contained against external benchmarks.

Assumptions & free parameters 0 free parameters · 0 assumptions · 0 invented entities

Only abstract available; no explicit free parameters, axioms, or invented entities can be extracted from the provided text.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Bias-Corrected Two-Stage Approach for Joint Modelling of Multidimensional Longitudinal HRQoL and Survival Data." pith.science (2026). https://pith.science/paper/2JIYPC2B

@misc{pith2026260623146,
  author       = {Pith},
  title        = {Pith review of: A Bias-Corrected Two-Stage Approach for Joint Modelling of Multidimensional Longitudinal HRQoL and Survival Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2JIYPC2B}},
  note         = {Machine review of arXiv:2606.23146}
}
read the original abstract

Health-related quality-of-life (HRQoL) outcomes are increasingly incorporated into oncology research to complement traditional survival endpoints by capturing patients' well-being over time. These outcomes are typically collected through multidimensional questionnaires yielding longitudinal ordinal data, and are often subject to dropout due to disease progression or death. In this context, joint models provide a well-established framework to account for the dependence between longitudinal HRQoL trajectories and time-to-event outcomes, but fully joint estimation rapidly becomes computationally prohibitive when multiple latent dimensions and random effects are involved. We propose a novel slope-corrected two-stage (SC2S) approach for the joint analysis of multivariate ordinal HRQoL data and survival outcomes within a multidimensional latent trait framework. The proposed approach propagates longitudinal information to the survival model through informative priors on the random effects, while additionally re-estimating longitudinal slope parameters. This strategy substantially reduces bias in both longitudinal and survival submodels while preserving much of the computational efficiency of two-stage procedures. Through simulation studies and an application to HRQoL data from patients with progressive glioblastoma, we show that the proposed method closely approximates fully joint Bayesian estimation while requiring notably less computation time.

Figures

Figures reproduced from arXiv: 2606.23146 by the authors.

Figure 1
Figure 1. Estimation errors for the survival parameters [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 2
Figure 2. Posterior standard deviations for the survival parameters [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Estimation errors of the fixed time effect parameters [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 1 canonical work pages

  1. [1]

    P. M. Fayers and D. Machin.Quality of life: the assessment, analysis and interpretation of patient- reported outcomes. John Wiley & Sons, 2013

    2013

  2. [2]

    Fayers, N

    P. Fayers, N. K. Aaronson, K. Bjordal, M. Grønvold, D. Curran, and A. Bottomley.EORTC QLQ- C30 scoring manual. European Organisation for Research and Treatment of Cancer, 2001

    2001

  3. [3]

    Gorter, J

    R. Gorter, J. P. Fox, and J. W. Twisk. Why item response theory should be used for longitudinal questionnaire data analysis in medical research.BMC Medical Research Methodology, 15:1–12, 2015

    2015

  4. [4]

    Barbieri, J

    A. Barbieri, J. Peyhardi, T. Conroy, S. Gourgou, C. Lavergne, and C. Mollevi. Item response models for the longitudinal analysis of health-related quality of life in cancer clinical trials.BMC Medical Research Methodology, 17(1):148, 2017

    2017

  5. [5]

    Wang and S

    J. Wang and S. Luo. Multidimensional latent trait linear mixed model: an application in clinical studies with multivariate longitudinal outcomes.Statistics in Medicine, 36(20):3244–3256, 2017

    2017

  6. [6]

    Henderson, P

    R. Henderson, P. Diggle, and A. Dobson. Joint modelling of longitudinal measurements and event time data.Biostatistics, 1(4):465–480, 2000

    2000

  7. [7]

    Touraine, B

    C. Touraine, B. Cuer, T. Conroy, B. Juzyna, S. Gourgou, and C. Mollevi. When a joint model should be preferred over a linear mixed model for analysis of longitudinal health-related quality of life data in cancer clinical trials.BMC Medical Research Methodology, 23(1):36, 2023

    2023

  8. [8]

    Saulnier, V

    T. Saulnier, V. Philipps, W. G. Meissner, O. Rascol, A. Pavy-Le Traon, A. Foubert-Samier, and C. Proust-Lima. Joint models for the longitudinal analysis of measurement scales in the presence of informative dropout.Methods, 203:142–151, 2022

    2022

Show all 29 references
  1. [9]

    H. Doms, P. Lambert, and C. Legrand. Joint modelling of longitudinal HRQoL data accounting for the risk of competing dropouts.Journal of the Royal Statistical Society: Series C, 2026. doi: 10.1093/jrsssc/qlag036. Accepted for publication

    2026 doi

  2. [10]

    Wang and S

    J. Wang and S. Luo. Joint modeling of multiple repeated measures and survival data using mul- tidimensional latent trait linear mixed model.Statistical Methods in Medical Research, 28(10-11): 3392–3403, 2019

    2019

  3. [11]

    Papageorgiou, K

    G. Papageorgiou, K. Mauff, A. Tomer, and D. Rizopoulos. An overview of joint modeling of time- to-event and longitudinal outcomes.Annual Review of Statistics and Its Application, 6(1):223–240, 2019

    2019

  4. [12]

    Self and Y

    S. Self and Y. Pawitan. Modeling a marker of disease progression and onset of disease. InAIDS epidemiology: methodological issues, pages 231–255. Birkh¨ auser Boston, Boston, MA, 1992

    1992

  5. [13]

    A. A. Tsiatis, V. Degruttola, and M. S. Wulfsohn. Modeling the relationship of survival to longitudinal data measured with error. applications to survival and CD4 counts in patients with AIDS.Journal of the American Statistical Association, 90(429):27–37, 1995. 20

    1995

  6. [14]

    A. A. Tsiatis and M. Davidian. Joint modeling of longitudinal and time-to-event data: an overview. Statistica Sinica, 14(3):809–834, 2004

    2004

  7. [15]

    W. Ye, X. Lin, and J. M. Taylor. Semiparametric modeling of longitudinal measurements and time- to-event data–a two-stage regression calibration approach.Biometrics, 64(4):1238–1246, 2008

    2008

  8. [16]

    Mauff, E

    K. Mauff, E. Steyerberg, I. Kardys, E. Boersma, and D. Rizopoulos. Joint models with multiple longitudinal outcomes and a time-to-event outcome: a corrected two-stage approach.Statistics and Computing, 30(4):999–1014, 2020

    2020

  9. [17]

    Alvares and V

    D. Alvares and V. Leiva-Yamaguchi. A two-stage approach for Bayesian joint models: reducing complexity while maintaining accuracy.Statistics and Computing, 33(5):115, 2023

    2023

  10. [18]

    Samejima

    F. Samejima. Estimation of latent ability using a response pattern of graded scores.Psychometrika, 34(S1):1–97, 1969

    1969

  11. [19]

    Rizopoulos.Joint models for longitudinal and time-to-event data: With applications in R

    D. Rizopoulos.Joint models for longitudinal and time-to-event data: With applications in R. CRC Press, 2012

    2012

  12. [20]

    Lambert and P

    P. Lambert and P. H. Eilers. Bayesian proportional hazards model with time–varying regression coefficients: A penalized Poisson regression approach.Statistics in Medicine, 24(24):3977–3989, 2005

    2005

  13. [21]

    Rizopoulos

    D. Rizopoulos. The R package JMbayes for fitting joint models for longitudinal and time-to-event data using MCMC.Journal of Statistical Software, 72:1–46, 2016

    2016

  14. [22]

    Alsefri, M

    M. Alsefri, M. Sudell, M. Garcia-Finana, and R. Kolamunnage-Dona. Bayesian joint modelling of longitudinal and time to event data: a methodological review.BMC Medical Research Methodology, 20:1–17, 2020

    2020

  15. [23]

    Lewandowski, D

    D. Lewandowski, D. Kurowicka, and H. Joe. Generating random correlation matrices based on vines and extended onion method.Journal of Multivariate Analysis, 100(9):1989–2001, 2009

    1989

  16. [24]

    P. H. Eilers and B. D. Marx. Flexible smoothing with B-splines and penalties.Statistical Science, 11(2):89–121, 1996

    1996

  17. [25]

    Lang and A

    S. Lang and A. Brezger. Bayesian P-splines.Journal of Computational and Graphical Statistics, 13 (1):183–212, 2004

    2004

  18. [26]

    Springer, Berlin, 2009

    C´ edric Villani.Optimal Transport: Old and New. Springer, Berlin, 2009

    2009

  19. [27]

    National library of medicine (US)

    ClinicalTrials.gov. National library of medicine (US). identifier NCT01290939: Bevacizumab and lomustine for recurrent GBM. Accessed 27 April 2023

    2023

  20. [28]

    W. Wick, T. Gorlia, M. Bendszus, M. Taphoorn, F. Sahm, I. Harting, et al. Lomustine and beva- cizumab in progressive glioblastoma.New England Journal of Medicine, 377(20):1954–1963, 2017

    1954

  21. [29]

    A Bias-Corrected Two-Stage Approach for Joint Modelling of Multidimensional Longitudinal HRQoL and Survival Data

    A. Gelman and D. B. Rubin. Inference from iterative simulation using multiple sequences.Statistical Science, 7(4):457–472, 1992. 21 Web-based Supporting Materials for “A Bias-Corrected Two-Stage Approach for Joint Modelling of Multidimensional Longitudinal HRQoL and Survival D...

    1992

Pith tools

Reviewed June 26, 2026 · model on record in the stance chip above.