Methods

Several steps are taken to reach the study aim:

We set-up a unique international network of experts in rehabilitation and supportive care in cancer. We create a comprehensive database of individual patient data from the best internationally available randomized controlled trials (RCTs), and infrastructure for inclusion of future trials in the database. Consequently, we conduct individual patient data (IPD) meta-analyses. We evaluate the effect of exercise and psychosocial interventions on quality of life compared with usual care or wait-list control group (read more).

We investigate what program works best for whom, under what circumstances (moderators), and through which mechanisms (mediators). We conduct a series of moderated multiple regression analyses to investigate which sociodemographic, clinical and personal characteristics and intervention type and circumstances moderate the effect of exercise and psychosocial interventions on quality of life. We conduct state-of-the-art mediation analyses to study whether the intervention effect on quality of life is mediated by fatigue, physical fitness and function, or psychosocial function (read more).

We conduct multiple backwards logistic regression analysis to build a model that predicts which intervention has the highest probability of success, i.e. improve the patient’s quality of life. Subsequently, the prediction model is internally and externally validated and translated into an easily applicable clinical decision rule (read more).