Joint modelling with competing risks of dropout for longitudinal analysis of health-related quality of life in cancer clinical trials

Abstract

Health-related quality of life (HRQoL) is an important endpoint in cancer clinical trials. Analysis of HRQoL longitudinal data is plagued by missing data, notably due to dropout. Joint models are increasingly receiving attention for modelling longitudinal outcomes and the time-to-dropout. However, dropout can be informative or non-informative depending on the cause. MethodsWe propose using a joint model that includes a competing risks sub-model for the cause-specific time-to-dropout.We compared a competing risks joint model (CR JM) that distinguishes between two causes of dropout with a standard joint model (SJM) that treats all the dropouts equally. First, we applied the CR JM and SJM to data from 267 patients with advanced oesophageal cancer from the randomized clinical trial PRODIGE 5/ACCORD 17 to analyse HRQoL data in the presence of dropouts unrelated and related to a clinical event. Then, we compared the models using a simulation study. ResultsWe showed that the CR JM performed as well as the SJM in situations where the risk of dropout was the same whatever the cause. In the presence of both informative and non-informative dropouts, only the SJM estimations were biased, impacting the HRQoL estimated parameters. ConclusionThe systematic collection of the reasons for dropout in clinical trials would facilitate the use of CR JMs, which could be a satisfactory approach to analysing HRQoL data in presence of both informative and non-informative dropout.Trial registration: This study is registered with ClinicalTrials.gov, number NCT00861094.

Bayesian Joint Model with Latent Time Shifts for Multivariate Longitudinal Data with Informative Dropout

“…Our findings were consistent with prior studies. First, most of the previous studies on joint models, both linear and nonlinear, showed that joint models outperformed separate models in the presence of informative dropout [24–28, 30, 32], which was consistent with our key findings. Second, the mix-effects model in the separate models that ignored informative dropout showed over-optimistic estimates for the disease progression trend, corresponding to conclusions drawn in previous studies [26–28].…”

“…Second, we considered time to dementia or death as a composite survival outcome instead of modeling them as competing or semi-competing risks. Previous studies showed the competing risks joint model that distinguishes between two causes of dropout may outperform the standard joint model that treats all the dropouts due to different causes equally in general [26, 30, 32]. However, this improvement is most pronounced only when two types of dropouts have substantially different relationships with the longitudinal measures, e.g., one type of dropout is informative and the other one is non-informative, or two competing risks are both informative but oppositely associated with the longitudinal measures.…”

“…Second, our model can be extended to distinguish dropouts due to dementia, death, and other causes by treating them as competing risks or semi-competing risks. Joint models with distinct reasons for dropout not only correct bias in the longitudinal process but also assess the impact of longitudinal process on each dropout type [26, 30, 32]. Third, in aging research, dementia onset is usually treated as interval-censored [12, 53].…”

Monitoring multidimensional aspects of quality of life after cancer immunotherapy: protocol for the international multicentre, observational QUALITOP cohort study

“…We will use mixed effect models and item response models to analyse the repeated measurements, 59 while simultaneously considering joint modelling to account for death as a competing event. 60 …”

“…The data collected in the QUALITOP project will benefit from repeated assessments of HRQoL over 18 months, facilitating the study of both individual trajectories over time and the causes and timing of changes in HRQoL. We will use mixed effect models and item response models to analyse the repeated measurements,59 while simultaneously considering joint modelling to account for death as a competing event 60…”