Dynamic risk predictions based on all available information are useful in timely identification of high‐risk patients. However, in contrast with time to event outcomes, there is still a lack of studies that clearly demonstrate how to obtain and update predictions for a future binary outcome using a repeatedly measured biomarker. The aim of this study is to give an illustrative overview of four approaches to obtain such predictions: likelihood based two‐stage method (2SMLE), likelihood based joint model (JMMLE), Bayesian two‐stage method (2SB), and Bayesian joint model (JMB). We applied the approaches to provide weekly updated predictions of post–molar gestational trophoblastic neoplasia (GTN) based on age and repeated measurements of human chorionic gonadotropin (hCG). Discrimination and calibration measures were used to compare the accuracy of the weekly predictions. Internal validation of the models was conducted using bootstrapping. The four approaches resulted in the same predictive and discriminative performance in predicting GTN. A simulation study showed that the joint models outperform the two‐stage methods when we increase the within‐ and the between‐patients variability of the biomarker. The applicability of these models to produce dynamic predictions has been illustrated through a comprehensive explanation and accompanying syntax (R and SAS®).

With the created nomograms, individualized risk of progression to post-molar GTN or MTX resistance can be predicted.• The nomograms are easy to use in daily clinical practice.• Four weeks after evacuation, 66% of patients who will progress to post-molar GTN can be identified, at 97.5% specificity.

The aim of our study was to investigate adherence to lifestyle recommendations and lifestyle changes after diagnosis in patients with non-muscle invasive bladder cancer (NMIBC). Second, we aimed to identify distinct trajectories of lifestyle change and their correlates. We analysed data of 935 patients with NMIBC from a prospective cohort study at six weeks (evaluating pre-diagnostic lifestyle), three months, and fifteen months after diagnosis. An overall lifestyle score (range 0–7) was calculated based on the 2018 World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) recommendations focusing on diet, body mass index, and physical activity. Linear mixed models were used to analyse absolute lifestyle changes over time. Distinct trajectories of change were identified with latent class trajectory models. We found an overall lifestyle score of 3.3 which remained constant over time. The largest lifestyle changes were observed for the consumption of red and processed meat (−96 g/week) and fruit and vegetables (−38 g/day). Two to four trajectory groups were identified for each single lifestyle behaviour. Correlates differed per trajectory group. In conclusion, adherence to the WCRF/AICR recommendations was low. Small to moderate changes in and different trajectories of single lifestyle behaviours were observed. Effective strategies for lifestyle improvement are warranted.

No abstract

Background Recovery trajectories differ between individual patients and it is hypothesizes that they can be used to predict if an individual patient is likely to recover earlier or later. Primary aim of this study was to determine if it is possible to identify recovery trajectories for physical functioning and pain during the first six weeks in patients after TKA. Secondary aim was to explore the association of these trajectories with one-year outcomes. Methods Prospective cohort study of 218 patients with the following measurement time points: preoperative, and at three days, two weeks, six weeks, and one year post-surgery (no missings). Outcome measures were performance-based physical functioning (Timed Up and Go [TUG]), self-reported physical functioning (Knee injury and Osteoarthritis Outcome Score-Activities of Daily Living [KOOS-ADL]), and pain (Visual Analogue Scale [VAS]). Latent Class Analysis was used to distinguish classes based on recovery trajectories over the first six weeks postoperatively. Multivariable regression analyses were used to identify associations between classes and one year outcomes. Results TUG showed three classes: “gain group” (n = 203), “moderate gain group” (n = 8) and “slow gain group” (n = 7), KOOS showed two classes: “gain group” (n = 86) and “moderate gain group” (n = 132), and VAS-pain three classes: “no/very little pain” (n = 151), “normal decrease of pain” (n = 48) and “sustained pain” (n = 19). The” low gain group” scored 3.31 [95% CI 1.52, 5.09] seconds less on the TUG than the “moderate gain group” and the KOOS “gain group” scored 11.97 [95% CI 8.62, 15.33] points better than the “moderate gain group” after one year. Patients who had an early trajectory of “sustained pain” had less chance to become free of pain at one year than those who reported “no or little pain” (odds ratio 0.11 [95% CI 0.03,0.42]. Conclusion The findings of this study indicate that different recovery trajectories can be detected. These recovery trajectories can distinguish outcome after one year.