Digital Serious Games to Promote Behavior Change in Children With Chronic Diseases: Scoping Review and Development of a Self-Management Learning Framework.

BACKGROUND: Digital serious games (SGs) have rapidly become a promising strategy for entertainment-based health education; however, developing SGs for children with chronic diseases remains a challenge. OBJECTIVE: In this study, we attempted to provide an updated scope of understanding of the development and evaluation of SG educational tools and develop a framework for SG education development to promote self-management activities and behavior change in children with chronic diseases. METHODS: This study consists of a knowledge base and an analytical base. This study followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines. To build the knowledge base, 5 stages of research were developed, including refining the review question (stage 1), searching for studies (stage 2), selecting relevant studies (stage 3), charting the information (stage 4), and collating the results (stage 5). Eligible studies that developed SG prototypes and evaluated SG education for children with chronic diseases were searched for in PubMed, Embase, Google Scholar, and peer-reviewed journals. In the analytical base, the context-mechanism-output approach and game taxonomy were used to analyze relevant behavioral theories and essential game elements. Game taxonomy included social features, presentation, narrative and identity, rewards and punishment, and manipulation and control. A total of 2 researchers selected the domains for the included behavioral theories and game elements. The intended SG framework was finalized by assembling SG fragments. Those SG fragments were appropriately reintegrated to visualize a new SG framework. RESULTS: This scoping review summarized data from 16 randomized controlled trials that evaluated SG education for children with chronic diseases and 14 studies on SG frameworks. It showed that self-determination theory was the most commonly used behavioral theory (9/30, 30%). Game elements included feedback, visual and audio designs, characters, narratives, rewards, challenges, competitions, goals, levels, rules, and tasks. In total, 3 phases of a digital SG framework are proposed in this review: requirements (phase 1), design and development (phase 2), and evaluation (phase 3). A total of 6 steps are described: exploring SG requirements (step 1), identifying target users (step 2), designing an SG prototype (step 3), building the SG prototype (step 4), evaluating the SG prototype (step 5), and marketing and monitoring the use of the SG prototype (step 6). Safety recommendations to use digital SG-based education for children in the post-COVID-19 era were also made. CONCLUSIONS: This review summarizes the fundamental behavioral theories and game elements of the available literature to establish a new theory-driven step-by-step framework. It can support game designers, clinicians, and educators in designing, developing, and evaluating digital, SG-based educational tools to increase self-management activities and promote behavior change in children with chronic diseases.

Machine learning algorithms to predict colistin-induced nephrotoxicity from electronic health records in patients with multidrug-resistant Gram-negative infection.

OBJECTIVES: Colistin-induced nephrotoxicity prolongs hospitalisation and increases mortality. The study aimed to construct machine learning models to predict colistin-induced nephrotoxicity in patients with multidrug-resistant Gram-negative infection. METHODS: Patients receiving colistin from three hospitals in the Clinical Research Database were included. Data were divided into a derivation cohort (2011-2017) and a temporal validation cohort (2018-2020). Fifteen machine learning models were established by categorical boosting, light gradient boosting machine and random forest. Classifier performances were compared by the sensitivity, F1 score, Matthews correlation coefficient (MCC), area under the receiver operating characteristic (AUROC) curve, and area under the precision-recall curve (AUPRC). SHapley Additive exPlanations plots were drawn to understand feature importance and interactions. RESULTS: The study included 1392 patients, with 360 (36.4%) and 165 (40.9%) experiencing nephrotoxicity in the derivation and temporal validation cohorts, respectively. The categorical boosting with oversampling achieved the highest performance with a sensitivity of 0.860, an F1 score of 0.740, an MCC of 0.533, an AUROC curve of 0.823, and an AUPRC of 0.737. The feature importance demonstrated that the days of colistin use, cumulative dose, daily dose, latest C-reactive protein, and baseline haemoglobin were the most important risk factors, especially for vulnerable patients. A cutoff colistin dose of 4.0 mg/kg body weight/d was identified for patients at higher risk of nephrotoxicity. CONCLUSIONS: Machine learning techniques can be an early identification tool to predict colistin-induced nephrotoxicity. The observed interactions suggest a modification in dose adjustment guidelines. Future geographic and prospective validation studies are warranted to strengthen the real-world applicability.