Article

AI Could Improve Blood Sugar Management in Type 1 Diabetes

An app using AI to evaluate blood sugar data, concerning trends, and recommend management tools will be a part of a clinical trial in 2024 for further assessment.

Automated insulin delivery systems make life easier for type 1 diabetics, but these devices can cost thousands of dollars depending on a patient’s insurance or the system model. These high costs make using the technology infeasible for many patients, who must rely on manually injecting themselves with daily insulin.

Researchers at Oregon Health & Science University (OSHU) designed an affordable and scalable app, DailyDose, to help patients with type 1 diabetes who manually inject themselves with insulin to manage their blood sugar levels. The app uses artificial intelligence (AI) to evaluate blood sugar data, concerning trends, and recommend management tools to maintain healthy blood glucose levels.

“Decision support, like that provided through the DailyDose app, greatly improves access to quality diabetes care, regardless of where people live,” said Sean Sullivan, PhD, program officer at the Helmsley Charitable Trust, in a recent press release.

This technology also differs from other diabetes apps because it educates patients on how to modify insulin doses or behaviors.

The OHSU researchers have published 2 papers about the app since 2020—1 determined that DailyDose offers recommendations that match those of diabetic specialists, and the other determined that these recommendations improved blood sugar levels.

But patients face barriers with the app, such as not understanding the rationale behind recommendations, or falsely believing that the recommendations would not improve their health, Peter G. Jacobs, an associate professor in the Department of Biomedical Engineering, OHSU School of Medicine, explained in the press release.

To combat malpractice or avoidant behavior, researchers will update the app’s user interface with transparent and digestible recommendations. Additionally, the app can connect patients with a diabetes educator or behavioral health clinician that can propel the patient towards positive change.

The study will enroll participants in 2024. For 38 weeks, 93 people with type 1 diabetes will use the DailyDose app in a multicenter clinical trial. At week 12, patients that show no benefit will be in touch with the diabetes educator or psychologist. Due to cost and time, this “app-based certified diabetes education therapy, or AB-CDE,” is a last resort, Leah M. Wilson, MD, an assistant professor in the Department of Medicine, Division of Endocrinology, Diabetes and Clinical Nutrition, OHSU School of Medicine, said in the press release.

More than 1.45 million people live with type 1 diabetes in the United States. People with type 1 diabetes require insulin to help them regulate blood sugar levels. Uncontrolled, people with type 1 diabetes are at risk of heart disease, kidney disease, nerve damage, vision loss, hearing loss, coma, or death.

The researchers were awarded $4.3 million for their work, explaining in the study, “Ultimately, we hope this project will improve care for all with type 1 diabetes and make managing the disease in their daily lives easier.”

Reference

Artificial intelligence, diabetes experts combine forces for blood sugar management study. News Release. January 10, 2022. Accessed on January 11, 2022. https://www.eurekalert.org/news-releases/976238

Related Videos
Senior Doctor is examining An Asian patient.
Healthcare, pharmacist and woman at counter with medicine or prescription drugs sales at drug store.
Image Credit: © Birdland - stock.adobe.com
pharmacogenetics testing, adverse drug events, personalized medicine, FDA collaboration, USP partnership, health equity, clinical decision support, laboratory challenges, study design, education, precision medicine, stakeholder perspectives, public comment, Texas Medical Center, DNA double helix
Pharmacy, Advocacy, Opioid Awareness Month | Image Credit: pikselstock - stock.adobe.com
pharmacogenetics challenges, inter-organizational collaboration, dpyd genotype, NCCN guidelines, meta census platform, evidence submission, consensus statements, clinical implementation, pharmacotherapy improvement, collaborative research, pharmacist role, pharmacokinetics focus, clinical topics, genotype-guided therapy, critical thought