Cause-specific individual predictive cumulative incidences for mpjlcmm

[Yeqian227]

Dear Cécile, dear Viviane,

I am using your lcmm package, and it is wonderful! Thank you very much for providing this work to the public. While using the mpjlcmm function, I have encountered some issues.

I noticed in issue #164 you mentioned the dynpred function is not implemented for a mpjlcmm model. However, in your paper :

Joint modeling of repeated multivariate cognitive measures and competing risks of dementia and death: a latent process and latent class approach.” Statistics in medicine vol. 35,3 (2016): 382-98. doi:10.1002/sim.6731

I observed that in section 4.7 'Individual dynamic predictions,' you calculated and plotted 'cause-specific individual predictive cumulative incidences' for 'an ApoE− man who graduated from primary school and entered the cohort at 68 years old up to a horizon of 5 years.’

Would appreciate any suggestions on how to calculate 'cause-specific individual predictive cumulative incidences' using repeated multivariate measures (such as IST15 and WST) in conjunction with baseline fixed variables (like sex, EL, and E4) for a mpjlcmm using lcmm package or function in other package in R, because I want to compute them, as mentioned in issue #118 , to calculate time-ROC and Brier scores and then I could compare mpjlcmmwith other survival models.

Thank you very much!

Best, Qian Ye

[VivianePhilipps]

Hi,

the analysis of this paper was not done in R. It uses a Fortran program, which is available on request, but is not as user-friendly as the R package.

Viviane

[Yeqian227]

Dear Viviane,

I apologize for my delayed response. Thank you very much for your reply, and I look forward to further updates from your team on this research. Once again, I appreciate your contribution in providing the lcmm program to the public.

Best regards, Qian