MedPath

Real-time Reminders To Decrease Late or Missed Meal Boluses

Not Applicable
Completed
Conditions
Type1 Diabetes Mellitus
Interventions
Device: Klue
Registration Number
NCT03970889
Lead Sponsor
Stanford University
Brief Summary

This study is examining whether the Klue app is effective in detecting missed or late meal boluses in patients with Type 1 diabetes. The app is programmed onto an Apple Watch and will detect potential missed boluses from hand motion. It will send text alerts to the user asking if they have bolused. This is a pilot study and will assess whether there is a change in the number of missed meal boluses in the two weeks prior to each visit. If the findings are significant, this software can be integrated in future closed-loop algorithms for automatic insulin delivery.

Detailed Description

This will be a randomized, cross-over, unblinded study with subjects randomized upon entry into the study to either begin using the Klue software for 6 weeks or to continue with their usual care for 6 weeks. Subjects are eligible for enrollment if they are using a continuous glucose monitor (CGM) and insulin pumps or an insulin pen with memory and CGM, are missing or late in giving at least 4 food boluses in the previous two weeks, and they have an iPhone. Missing or late meal boluses will be assessed through their pump/pen and sensor downloads. This is a pilot study. The investigators do not have preliminary data to do a true power calculation. The primary outcome will be the change in the number of missed meal boluses in the two weeks prior to each visit. Secondary outcome measures will be the number of missed meal boluses in each month of the study, change in HbA1c levels (measured every 6 weeks), accuracy of the Klue software in detecting meals (true positive and false positive rates), and a user satisfaction survey.

Recruitment & Eligibility

Status
COMPLETED
Sex
All
Target Recruitment
24
Inclusion Criteria
  1. Age 13 and over
  2. On an insulin pump for at least 3 months and using CGM or using an insulin pen with memory and a CGM
  3. Willing to wear a CGM at least 70% of the time while in the study
  4. Using an iPhone
  5. Willing to wear an Apple watch on their dominant hand while awake
  6. Missing or late in giving at least four food boluses in the previous two weeks
  7. Understanding and willingness to follow the protocol and sign informed consent
  8. Ability to speak, read and write in the language of the investigators
Read More
Exclusion Criteria
  1. Pregnant or lactating
  2. A known medical condition that in the judgment of the investigator might interfere with the completion of the protocol
  3. Current treatment for a seizure disorder
  4. Inpatient psychiatric treatment in the past 6 months

Note: Adequately treated thyroid disease and celiac disease do not exclude subjects from enrollment. They may be participating in another trial, if this study is approved by the investigators of the other trial.

Read More

Study & Design

Study Type
INTERVENTIONAL
Study Design
CROSSOVER
Arm && Interventions
GroupInterventionDescription
KlueKlueSubjects will wear an Apple watch on their dominant hand and receive alerts when eating behavior is detected by the Klue software.
Primary Outcome Measures
NameTimeMethod
Change in the number of missed meal bolusesThe two weeks prior to each study visit

A missed meal bolus is defined by no insulin dose within 2 hours from the start of the CGM rise.

Secondary Outcome Measures
NameTimeMethod
Positive predictive value of a Klue meal alertAll data for the 6 weeks on the Klue intervention arm when Klue is active

Percent of time a meal alert signaled a meal

Change in time in rangedata for 1 month prior to each study visit at baseline, 6 weeks and 12 weeks of this 12 week study for each subject

sensor glucose values between 70-180 mg/dl

Trial Locations

Locations (1)

Stanford

🇺🇸

Palo Alto, California, United States

© Copyright 2025. All Rights Reserved by MedPath