A Crossover Study to Assess the Effect of an Artificial Intelligence (AI)-Based Bedtime Smart Snack Intervention in Preventing Overnight Low Glucose in People With T1D on Multiple Daily Injections.
试验速览
- 阶段
- 不适用
- 状态
- 已完成
- 入组人数
- 21
- 试验地点
- 1
- 主要终点
- Probability of Overnight Hypoglycemia
研究概览
简要总结
This is a single-center, open-label, crossover trial with two arms and two periods (2x2) and one-week washout period. The study is designed to evaluate the efficacy of an AI-based bedtime smart snack intervention in reducing nocturnal low glucose in people living with T1D on MDI therapy compared with traditional CGM-augmented MDI therapy as the control.
详细描述
Participants will be randomized to either first use CGM only to manage glucose for four weeks (control arm) followed by four weeks of DailyDose App + bedtime smart snack intervention (intervention arm), or vice-versa. There will be a one-week washout period between arms.
During the control arm, participants will wear CGM and will manage their glucose as usual. Participants will be asked to wear a smart watch overnight to collect sleep metrics, weigh themselves weekly in the morning before eating, and answer a one-item sleep quality scale survey weekly. We will collect CGM measurements during the control arm for evaluation of effect of intervention and assessment of the accuracy of low glucose prediction.
During the intervention arm, participants will use the DailyDose Smart Snack smart phone app. When they are getting ready for bed, an AI-based model in DailyDose will predict the likelihood of overnight low glucose at bedtime and will recommend a personalized snack to help avoid nocturnal hypoglycemia. The nutritional content of the snack (carbohydrate, protein, fat, etc.) will be dependent on the predicted overnight minimum glucose and the predicted time of the minimum overnight glucose level. During this arm, participants will also be asked to wear a smart watch overnight, weigh themselves weekly, and answer a one-item sleep quality scale survey weekly.
研究设计
- 研究类型
- Interventional
- 分配方式
- Randomized
- 干预模型
- Crossover
- 主要目的
- Treatment
- 盲法
- None
入排标准
- 年龄范围
- 18 Years 至 100 Years（Adult, Older Adult）
- 性别
- All
- 接受健康志愿者
- 否
入选标准
- •Diagnosis of type 1 diabetes mellitus for at least 1 year
- •Male or female participants 18 years of age or older
- •Using multiple daily injections
- •HbA1c <10% at screening
- •Current use of a continuous glucose monitoring system with at least two episodes of overnight hypoglycemia (defined as sensed glucose <70 mg/dL for at least 10 minutes between the hours of 10 PM and 6 AM) within 30 days prior to screening
- •Individuals with history of severe hypoglycemia requiring third party assistance must have a companion in the same dwelling as the study participant who will be linked to the participant's Dexcom app during the control arm of the study, and who is trained in the administration of glucagon.
- •Willingness to follow all study procedures
- •Willingness to sign informed consent and HIPAA documents
排除标准
- •Individual of childbearing potential who is pregnant or intending to become pregnant or breast-feeding or is not using adequate contraceptive methods. Acceptable contraception includes birth control pill / patch / vaginal ring, Depo-Provera, Norplant, an IUD, the double barrier method (the woman uses a diaphragm and spermicide, and the man uses a condom), or abstinence
- •Any active infection
- •Known or suspected abuse of alcohol, narcotics, or illicit drugs (except marijuana use)
- •Seizure disorder
- •Use of non-insulin glucose lowering medications
- •Use of steroids
- •Stage-three or more advanced chronic kidney disease
- •Hypo- or hyper- thyroidism that is not medically optimized and on a stable regimen define as Thyroid-stimulating hormone (TSH) outside of the normal reference range based on screening labs
- •Adrenal insufficiency
- •Cirrhosis
- •Any life-threatening disease, including malignant neoplasms and medical history of malignant neoplasms within the past 5 years prior to screening (except basal and squamous cell skin cancer).
- •Any clinically significant disease or disorder which in the opinion of the Investigator may jeopardize the participant's safety or compliance with the protocol
- •Individual working night shifts
研究组 & 干预措施
Intervention
Participants will use the DailyDose Smart Snack smart phone application which contains an AI-based model that predicts the likelihood of overnight low glucose at bedtime every night and will recommend a personalized snack to help avoid nocturnal hypoglycemia. The nutritional content of the snack (carbohydrate, protein, fat, etc.) will be dependent on the predicted overnight minimum glucose and the predicted time of the minimum overnight glucose level. During this arm, participants will also be asked to wear a smart watch overnight, weigh themselves weekly, and answer a one-item sleep quality scale survey weekly.
干预措施: DailyDose Smart Snack app (Device)
Control
Participants will wear Dexcom G6 CGM and will manage their glucose as usual. Participants will be asked to wear a smart watch overnight to collect sleep metrics, weigh themselves weekly in the morning before eating, and answer a one-item sleep quality scale survey weekly.
干预措施: Dexcom G6 CGM (Device)
结局指标
主要结局
Probability of Overnight Hypoglycemia
时间窗: 8 weeks (4-week control period vs. 4-week intervention period)
An episode of overnight hypoglycemia is counted if sensor glucose is \<70 mg/dL for at least two measurements during an eight-hour period following announced bedtime. This is assessed by number of episodes divided by total number of nights.
次要结局
- Time to the First Overnight Low-glucose Event (<70 mg/dL)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Less Than 54 mg/dL (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Less Than 54 mg/dL (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Less Than 70 mg/dL (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Less Than 70 mg/dL (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Between 70-180 mg/dL (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Between 70-180 mg/dL (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Greater Than 180 mg/dL (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time With Sensed Glucose Greater Than 180 mg/dL (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time Sensed Glucose Greater Than 250 mg/dL (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Percentage of Time Sensed Glucose Greater Than 250 mg/dL (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Mean Sensed Glucose (Overnight)(8 weeks (4-week control period vs. 4-week intervention period))
- Mean Sensed Glucose (24-hour/Day Study Duration)(8 weeks (4-week control period vs. 4-week intervention period))
- Accuracy of Overnight Low Glucose Prediction by Sensitivity(4 weeks of control period)
- Accuracy of Overnight Low Glucose Prediction by Specificity(4 weeks of control period)
- Change in Weight(8 weeks (4-week control period vs. 4-week intervention period))
研究者
Clara Mosquera-Lopez
Principal Investigator
Oregon Health and Science University