Objective
The purpose of this research is to improve the lives of patients with diabetes through the rapid discovery and proactive reduction of risks associated with insulin pump use.
Background Rationale
Insulin pumps offer life-changing benefits to patients with diabetes, but they are not without risk. Insulin pumps account for the largest number of device-related adverse events (AEs) reported to the Food and Drug Administration (FDA). These AEs can reveal current and emerging risks associated with insulin pump use, but existing methods for uncovering risks in AE reports rely on human analysts reading and categorizing thousands of AE narratives. This manual approach is suboptimal because it is labor intensive and time consuming, delaying patient awareness of insulin pump risks and the creation of risk reduction strategies.
Description of Project
Objective: The purpose of this research is to improve the lives of patients with diabetes through the rapid discovery and proactive reduction of risks associated with insulin pump use.
Background/Rationale: Insulin pumps offer life-changing benefits to patients with diabetes, but they are not without risk. Insulin pumps account for the largest number of device-related adverse events (AEs) reported to the Food and Drug Administration (FDA). These AEs can reveal current and emerging risks associated with insulin pump use, but existing methods for uncovering risks in AE reports rely on human analysts reading and categorizing thousands of AE narratives. This manual approach is suboptimal because it is labor intensive and time consuming, delaying patient awareness of insulin pump risks and the creation of risk reduction strategies.
Anticipated Outcome: We will develop and use an innovative artificial intelligence (AI) system called AI-assisted Device Risk Identification System (AIDRIS) to accelerate the identification and mitigation of insulin pump risks.
Relevance to Type 1 Diabetes: Among approximately 2 million people in the US living with type 1 diabetes, about 67% use an insulin pump. Insulin pump users currently learn about device risks through personal experience with a pump problem, which exposes them to harm and results in fragmented knowledge of risks. Insulin pump users also learn about device risks through manufacturer communications. However, formal safety notices about widespread insulin pump risks can take a long time for device manufacturers to develop and release to patients, which delays risk awareness and results in missed opportunities to mitigate risks. AIDRIS is innovative because it will be the first comprehensive and efficient system for identifying insulin pump risks through AE reports. AIDRIS will increase patient awareness of device risks and inform risk reduction strategies to prevent patient harm.
Anticipated Outcome
We will develop and use an innovative artificial intelligence (AI) system called AI-assisted Device Risk Identification System (AIDRIS) to accelerate the identification and mitigation of insulin pump risks.
Relevance to T1D
Among approximately 2 million people in the US living with type 1 diabetes, about 67% use an insulin pump. Insulin pump users currently learn about device risks through personal experience with a pump problem, which exposes them to harm and results in fragmented knowledge of risks. Insulin pump users also learn about device risks through manufacturer communications. However, formal safety notices about widespread insulin pump risks can take a long time for device manufacturers to develop and release to patients, which delays risk awareness and results in missed opportunities to mitigate risks. AIDRIS is innovative because it will be the first comprehensive and efficient system for identifying insulin pump risks through AE reports. AIDRIS will increase patient awareness of device risks and inform risk reduction strategies to prevent patient harm.