1. This randomized clinical trial demonstrated that compared with the standard of care, voice-based conversational artificial intelligence (VBAI) improved time to optimal insulin dosing, insulin adherence, glycemic control, and diabetes-related distress among adults with type 2 diabetes.
Evidence Rating Level: 1 (Excellent)
Study Rundown: Type 2 diabetes affects 33 million US adults, with nearly a quarter having poor glycemic control with a hemoglobin A1c (HbA1c) level above 8%. In these individuals, insulin therapy is essential but requires frequent titrations to achieve the optimal dose. Traditional clinic-based titration faces obstacles, leading to suboptimal doses and insufficient glycemic control. This randomized control trial aims to assess the efficacy of VBAI in enhancing medication adherence, glycemic control, and time to achieve optimal insulin dosage. From March 2021 to December 2022, English-speaking adults with type 2 diabetes who required initiation or adjustment of once-daily basal insulin were recruited and randomized to receive VBAI or standard of care. Compared with the standard care group, those in the VBAI group achieved a significantly higher mean insulin adherence, achieved optimal insulin dosing significantly faster, were more likely to achieve glycemic control, and had significantly decreased diabetes-related emotional distress. Overall, these results demonstrate the utility of employing voice-based digital health solutions for medication titration.
Click to read the study on JAMA Network Open
In-Depth [randomized controlled trial]: Many adults with Type 2 diabetes have poor glycemic control (HbA1c > 8%) and therefore require insulin therapy with frequent titrations for optimal dosing. Traditional clinic-based titration faces challenges, leading to suboptimal dosing. This randomized clinical trial conducted at 4 primary care clinics at Stanford University assesses VBAI’s efficacy in enhancing medication adherence, glycemic control, and time to optimal insulin dosage. 39 participants were randomized in a 1:1 ratio to receive basal insulin adjustments via VBAI, an AI software developed for this trial powered by Alexa, or via their clinician per usual, the standard of care. Ultimately 32 participants (mean [SD] age 55.1 [12.7] years, 59.4% women, mean [SD] HbA1c level was 9.6% [1.5%]), with 16 participants in each group completed the enrollment process. Participants in the VBAI group achieved a higher mean insulin difference compared with those in the standard of care group (difference, 32.7% [95% CI, 8.0%-57.4%]; P = .01). The VBAI group achieved optimal insulin dosage in a medium time of 15 days (IQR, 6-27 days), while the standard of care group exceeded 56 days (IQR, >29.5 to >56 days). There was a significant difference in time-to-event curves (P = .006), with fewer than half of the standard-of-care group participants achieving optimal insulin dosing at 8 weeks. Compared with participants in the standard of care group, those in the VBAI group had decreased 5-item Problem Areas in Diabetes Scale (PAID-5), a survey assessing diabetes-related distress, with a mean difference of −3.6 points (95% CI, −6.8 to −0.4 points; P = .03). More participants in the VBAI group achieved glycemic control (difference, 56.3% [95% CI, 21.4%-91.1%]; P = .005) and had greater improvements in fasting blood glucose (difference, −68.9 mg/dL [95% CI, −107.1 to −30.7 mg/dL]; P = .001). There were no adverse events reported in the study. Overall, this trial found that participants in the VBAI group had higher insulin adherence, faster insulin dose optimization, improved FBG, and decreased diabetes-related emotional distress compared with those in the standard-of-care group. These findings demonstrate the utility of digital health tools in medication titration and highlight the effectiveness of voice user interfaces for patient-facing digital technologies.
Image: PD
©2024 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.
