Patient-reported outcomes and treatment adherence in type 2 diabetes using natural language processing: Wave 8 of the Observational International Diabetes Management Practices Study.

AIMS/INTRODUCTION: We analyzed patient-reported outcomes of people with type 2 diabetes to better understand perceptions and experiences contributing to treatment adherence. MATERIALS AND METHODS: In the ongoing International Diabetes Management Practices Study, we collected patient-reported outcomes data from structured questionnaires (chronic treatment acceptance questionnaire and Diabetes Self-Management Questionnaire) and free-text answers to open-ended questions to assess perceptions of treatment value and side-effects, as well as barriers to, and enablers for, adherence and self-management. Free-text answers were analyzed by natural language processing. RESULTS: In 2018-2020, we recruited 2,475 patients with type 2 diabetes (43.3% insulin-treated, glycated hemoglobin (HbA1c) 8.0 ± 1.8%; 30.9% with HbA1c <7%) from 13 countries across Africa, the Middle East, Europe, Latin America and Asia. Mean ± standard deviation scores of chronic treatment acceptance questionnaire (acceptance of medication, rated out of 100) and Diabetes Self-Management Questionnaire (self-management, rated out of 10) were 87.8 ± 24.5 and 3.3 ± 0.9, respectively. Based on free-text analysis and coded responses, one in three patients reported treatment non-adherence. Overall, although most patients accepted treatment values and side-effects, self-management was suboptimal. Treatment duration, regimen complexity and disruption of daily routines were major barriers to adherence, whereas habit formation was a key enabler. Treatment-adherent patients were older (60 ± 11.6 vs 55 ± 11.7 years, P < 0.001), and more likely to have longer disease duration (12 ± 8.6 vs 10 ± 7.7 years, P < 0.001), exposure to diabetes education (73.1% vs 67.8%, P < 0.05), lower HbA1c (7.9 ± 1.8% vs 8.3 ± 1.9%, P < 0.001) and attainment of HbA1c <7% (29.7% vs 23.3%, P < 0.01). CONCLUSIONS: Patient perceptions/experiences influence treatment adherence and self-management. Patient-centered education and support programs that consider patient-reported outcomes aimed at promoting empowerment and developing new routines might improve glycemic control.

Development of an early alert model for pandemic situations in Germany.

The COVID-19 pandemic has pointed out the need for new technical approaches to increase the preparedness of healthcare systems. One important measure is to develop innovative early warning systems. Along those lines, we first compiled a corpus of relevant COVID-19 related symptoms with the help of a disease ontology, text mining and statistical analysis. Subsequently, we applied statistical and machine learning (ML) techniques to time series data of symptom related Google searches and tweets spanning the time period from March 2020 to June 2022. In conclusion, we found that a long-short-term memory (LSTM) jointly trained on COVID-19 symptoms related Google Trends and Twitter data was able to accurately forecast up-trends in classical surveillance data (confirmed cases and hospitalization rates) 14 days ahead. In both cases, F1 scores were above 98% and 97%, respectively, hence demonstrating the potential of using digital traces for building an early alert system for pandemics in Germany.