Background: Providing patients with cancer who are undergoing systemic therapy with useful information about symptom management is essential to prevent unnecessary deterioration of quality of life. Objective: The aim was to evaluate whether use of an app for symptom management was associated with any change in patient quality of life or use of health resources. Methods: Outpatients with early stage breast cancer receiving systemic therapy were recruited at the Institute of Oncology in Ljubljana, Slovenia. Patients who received systemic therapy between December 2017 and March 2018 (control group) and between April 2018 and September 2018 (intervention group) were eligible. All patients received standard care, but only those in the intervention group were asked to use mPRO Mamma, an Android-based smartphone app, in addition. The app supported daily tracking of 50 symptoms, allowed users to grade their symptom severity (as mild, moderate, or severe), and also provided in-depth descriptions and recommendations based on reported symptom level. Patient-reported outcomes in both groups were assessed through the European Organisation for Research and Treatment of Cancer (EORTC) core (C-30) and breast cancer (BR-23) questionnaires, as well as a questionnaire about health resources use. The primary outcomes were the difference in the global quality of life between groups and the difference in summary score of the EORTC C-30 questionnaire between groups after 3 time periods (the first week of treatment, the first treatment cycle, and the entire treatment). The secondary outcome was the use of health resources (doctor visits and hospitalizations) in each time period. Other scales were used for exploratory analysis. Results: The mean difference between the intervention group (n=46) and the control group (n=45) in global quality of life (adjusted for baseline and type of surgery) after the first week was 10.1 (95% CI 1.8 to 18.5, P=.02). The intervention group summary scores were significantly higher than those of the control group after the first week (adjusted mean difference: 8.9, 95% CI 3.1 to 14.7, P=.003) and at the end of treatment (adjusted mean difference: 10.6, 95% CI 3.9 to 17.3, P=.002). Use of health resources was not statistically significant between the groups in either the first week (P=.12) or the first treatment cycle (P=.13). Exploratory analysis findings demonstrated clinically important improvements (indicated by EORTC C-30 or BR-23 scale scores)—social, physical, role, and cognitive function were improved while pain, appetite loss, and systemic therapy side effects were reduced. Conclusions: Use of the app enabled patients undergoing systemic therapy for early stage breast cancer to better cope with symptoms which was demonstrated by a better global quality of life and summary score after the first week and by a better summary score at the end of treatment in the intervention group compared to those of the control group, but no change in the use of health resources was demonstrated.

Background: Cardiovascular disorders in general are responsible for 30% of deaths worldwide. Among them, hypertrophic cardiomyopathy (HCM) is a genetic cardiac disease that is present in about 1 of 500 young adults and can cause sudden cardiac death (SCD). Objective: Although the current state-of-the-art methods model the risk of SCD for patients, to the best of our knowledge, no methods are available for modeling the patient's clinical status up to 10 years ahead. In this paper, we propose a novel machine learning (ML)-based tool for predicting disease progression for patients diagnosed with HCM in terms of adverse remodeling of the heart during a 10-year period. Methods: The method consisted of 6 predictive regression models that independently predict future values of 6 clinical characteristics: left atrial size, left atrial volume, left ventricular ejection fraction, New York Heart Association functional classification, left ventricular internal diastolic diameter, and left ventricular internal systolic diameter. We supplemented each prediction with the explanation that is generated using the Shapley additive explanation method. Results: The final experiments showed that predictive error is lower on 5 of the 6 constructed models in comparison to experts (on average, by 0.34) or a consortium of experts (on average, by 0.22). The experiments revealed that semisupervised learning and the artificial data from virtual patients help improve predictive accuracies. The best-performing random forest model improved R2 from 0.3 to 0.6. Conclusions: By engaging medical experts to provide interpretation and validation of the results, we determined the models' favorable performance compared to the performance of experts for 5 of 6 targets.