Development and implementation of a digital health intervention (DHI) electronic patient-reported outcome measure (ePROM) platform to monitor patients with cancer.

Abstract

163 Background: Cancer is heterogeneous diseases associated with a significant burden of symptoms and treatment-related adverse events. Digital Health Interventions (DHIs) emerge as promising tools for active/continuous monitoring of patients (pts). Evidence suggests that DHIs may enhance healthcare delivery and improve clinical outcomes by facilitating symptom detection and enabling more effective real-time informed decisions. We aim to develop and implement an electronic Patient-reported Outcome Measure (ePROM) platform to monitor pts undergoing cancer treatment. Methods: We developed ThummiOnco, a DHI ePROM platform designed for monitoring of pts during cancer treatment. It allows pts to report their symptoms (reported and graded according to the CTCAE), quality-of-life (QoL) (EORTC EQ-5D questionnaire), emotional state, treatment adherence (Morisky-Green questionnaire), Emergency Room (ER) visits and hospitalizations through a mobile application. Upon registration, pts consent to share their data with their medical team and provide essential demographic information such as age, sex, cancer type, and treatment regimen. A medical algorithm analyzes the collected data to generate instant personalized recommendations and guidance, including indicating the need of seeking emergency care in the case of serious high-grade or worrisome complaints. Patients also have the option to engage in real-time conversations with the medical team through an online chat feature. Treating physicians and the medical team can continuously access pts data real-time, through an online dashboard, enabling them to visualize information and statistics, address symptom alerts, and make informed management decisions. ThummiOnco is available at no cost to pts at htt ps://thummi.global . Results: Between August 2021 and March 2024, ThummiOnco monitored 925 pts across 7 institutions in Brazil, encompassing both private and public health systems. The majority of pts (72%) were female, with 28% being male. The most frequently monitored cancer types were Breast Cancer (39%), Lung Cancer (7%) and Colon Cancer (6%). During this period, a total of 71.373 interactions were recorded, including 36.175 (51%) chat messages, 21.644 (30%) symptom reports, 10.597 (15%) emotional state reflections, 2.819 (4%) QoL assessments, and 138 (0.2%) ER registries. A total of 23 different symptoms were reported, with the most frequent being: fatigue (16%), joint pain (8%), and nausea (7%). 13.187 (61%) grade 1, 6.545 (30%) grade 2, and 1.912 (9%) grade 3 symptoms were registered. In total, the medical algorithm generated 1.921 recommendations for seeking emergency care. Conclusions: ThummiOnco's capacity to detect a wide range of symptoms and enable real-time informed decisions underscores its potential to enhance patient outcomes and healthcare service delivery.