Predictive modeling for 14-day unplanned hospital readmission risk by using machine learning algorithms.

BACKGROUND: Early unplanned hospital readmissions are associated with increased harm to patients, increased medical costs, and negative hospital reputation. With the identification of at-risk patients, a crucial step toward improving care, appropriate interventions can be adopted to prevent readmission. This study aimed to build machine learning models to predict 14-day unplanned readmissions. METHODS: We conducted a retrospective cohort study on 37,091 consecutive hospitalized adult patients with 55,933 discharges between September 1, 2018, and August 31, 2019, in an 1193-bed university hospital. Patients who were aged < 20 years, were admitted for cancer-related treatment, participated in clinical trial, were discharged against medical advice, died during admission, or lived abroad were excluded. Predictors for analysis included 7 categories of variables extracted from hospital's medical record dataset. In total, four machine learning algorithms, namely logistic regression, random forest, extreme gradient boosting, and categorical boosting, were used to build classifiers for prediction. The performance of prediction models for 14-day unplanned readmission risk was evaluated using precision, recall, F1-score, area under the receiver operating characteristic curve (AUROC), and area under the precision-recall curve (AUPRC). RESULTS: In total, 24,722 patients were included for the analysis. The mean age of the cohort was 57.34 ± 18.13 years. The 14-day unplanned readmission rate was 1.22%. Among the 4 machine learning algorithms selected, Catboost had the best average performance in fivefold cross-validation (precision: 0.9377, recall: 0.5333, F1-score: 0.6780, AUROC: 0.9903, and AUPRC: 0.7515). After incorporating 21 most influential features in the Catboost model, its performance improved (precision: 0.9470, recall: 0.5600, F1-score: 0.7010, AUROC: 0.9909, and AUPRC: 0.7711). CONCLUSIONS: Our models reliably predicted 14-day unplanned readmissions and were explainable. They can be used to identify patients with a high risk of unplanned readmission based on influential features, particularly features related to diagnoses. The operation of the models with physiological indicators also corresponded to clinical experience and literature. Identifying patients at high risk with these models can enable early discharge planning and transitional care to prevent readmissions. Further studies should include additional features that may enable further sensitivity in identifying patients at a risk of early unplanned readmissions.

Mobile App (UPrEPU) to Monitor Adherence to Pre-exposure Prophylaxis in Men Who Have Sex With Men: Protocol for a User-Centered Approach to Mobile App Design and Development.

BACKGROUND: Daily and on-demand pre-exposure prophylaxis (PrEP) has been well demonstrated to effectively prevent HIV acquisition for men who have sex with men (MSM). More than half of the MSM PrEP users in Taiwan prefer on-demand PrEP; however, on-demand PrEP involves a complicated dosing regimen because it requires precoital and postcoital dosing and sex events are hard to anticipate. Although there are a growing number of mobile apps designed to improve access to HIV prevention services and HIV medication adherence, few mobile apps focus on adherence to PrEP or are designed to accommodate a complicated, on-demand PrEP dosing schedule. OBJECTIVE: The aim of this project is to evaluate the usability of a newly developed mobile app (UPrEPU) to assist MSM PrEP users to self-monitor their adherence to either daily or on-demand PrEP using a user-centered scheme. METHODS: This research will be conducted in 2 phases: app development and usability study. In the app development phase, we will first conduct formative research with end users and stakeholders through in-depth interviews; the results will provide PrEP users' and PrEP navigators' personas as material used in the app conceptualization stage. PrEP navigators are individuals in the health care system that help HIV-negative individuals who need assistance in accessing PrEP care. A low-fidelity prototype of the app feature will be formatted by applying a participatory design approach to engage PrEP users, designers, and app developers in the design process of the app. Then, a high-fidelity prototype of the app will be developed for the usability study and refined iteratively by the multidisciplinary team and new internal testers. Internal testers include the research team consisting of experts in public health, infectious disease, and industrial design and a close network of the research team that is taking PrEP. In the usability study phase, we will enroll 70 MSM PrEP users and follow them up for 4 months. Usability, feasibility, and effectiveness of adherence monitoring will be evaluated. RESULTS: Refinement of the UPrEPU app is currently ongoing. The usability study commenced in May 2020. CONCLUSIONS: The UPrEPU app is one of the first apps designed to help MSM PrEP users to self-manage their PrEP schedule better regardless of dosing modes. With a design-thinking approach and adapting to the cultural context in Taiwan's MSM population, this novel app will have substantial potential to be acceptable and feasible and contribute to the reduction of new HIV infections. TRIAL REGISTRATION: ClinicalTrials.gov NCT04248790; https://clinicaltrials.gov/ct2/show/NCT04248790. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/20360.