Manan Shukla, Jianjing Lin PhD, Oshani Seneviratne PhD
Rensselaer Polytechnic Institute, Troy, NY, USA
#AMIA2021
BlockIoT: Blockchain-based Health Data Integration using IoT Devices
Oral Presentations - Informatics Data Innovations
S67
Disclosure
There are no financial conflicts of interest to disclose.
2
AMIA 2021 Annual Symposium | amia.org
Learning Objectives
After participating in this session the learner should be better able to:
- Understand the current issues regarding interoperability between various health infrastructures, and specifically consider the potential and accessibility of medical device data.
- Realize potential benefits in terms of quality and efficiency in healthcare delivery from using medical-device sourced physiological data, and understand the challenges that limit its application in current clinical practices.
- Learn about potential benefits of managing and transferring patient information via a decentralized storage mechanism and smart contracts, which may further facilitate the application of emerging technologies like AI/ML in medicine
3
AMIA 2021 Annual Symposium | amia.org
Introduction
4
AMIA 2021 Annual Symposium | amia.org
Linkous, Lauren & Zohrabi, Nasibeh & Abdelwahed, Sherif. (2019). Health Monitoring in Smart Homes Utilizing Internet of Things. 10.1109/CHASE48038.2019.00020.
Problems Addressed
Ex: How can you treat hypertension without knowing the patient’s BP value over time?
5
AMIA 2021 Annual Symposium | amia.org
5
BlockIoT
6
AMIA 2021 Annual Symposium | amia.org
BlockIoT Architecture
7
AMIA 2021 Annual Symposium | amia.org
7
Participatory Design Survey
Goal:
- Gauge the importance of medical device data on a provider’s daily workflow
- Determine whether addition of medical device data can influence a physician's medical decision-making process
- Feedback for improvements to be made with data representation
8
AMIA 2021 Annual Symposium | amia.org
8
Participatory Survey
- 13 physicians (recruited through personal/professional connections)
- Different specialties (cardiologist, GI, dermatologist, geriatrician, emergency physician, surgeon, anesthesiologist, pulmonologist/allergy/sleep medicine, internal medicine, and primary care)
- 1-1.5 hr interviews, with Google Form
9
AMIA 2021 Annual Symposium | amia.org
9
Participatory Survey
- A sample HPI of a patient was given with:
— 5 chronic conditions
— 5 personal health devices whose data reflects these conditions
- A mock EHR was provided that contains the representations of the patient’s device data
- Goal is to make the system comparable to what would be used on a regular basis
10
AMIA 2021 Annual Symposium | amia.org
Mock Patient Data for Participatory Survey
11
Evaluation Methodology
1. Brief Overview of BlockIoT
2. Questions regarding their current practice- ex. use of medical device data
3. Display the patient’s HPI
4. Questions in regards to how they would treat the patient
5. Display medical device data
6. Ask whether data from each graph would potentially change their treatment decision
7. Ask for the physician’s overall opinion of the system
12
AMIA 2021 Annual Symposium | amia.org
12
Survey Results- After BlockIoT Demo
- Access to means for obtaining physiological data from patients is minimal or non-existent.
- Access to device data is also variable, i.e., no standardized method currently exists
- Physician 4 institutes a log for patients with chronic illnesses
— Admitted that the quality of the data in these logs is only “about 50%”
13
AMIA 2021 Annual Symposium | amia.org
Physician 3
“The only way we know whether the patient has high blood pressure is whether the patient takes a one- time blood pressure reading a day at home in the morning, and that is only when we really ask to do it. Even then, they will only do it for some of the time, not all the time. Or when they come to the doctor’s office every 2 to 3 months. A blood pressure reading can change every second, which means that we are missing millions of data points right now!”
14
Evaluation Results- After BlockIoT
- 87% of physicians reporting that they would be willing to use the summary views as well as the patient/physician interventions throughout the patient care
- Other 13% were more inclined to just use the summary views
- 2 physicians cited alert fatigue as a reason to not receive alerts from BlockIoT and preferred to simply use the summary views.
- All of the physicians stated that the given data would make them more efficient in the diagnosis process
15
AMIA 2021 Annual Symposium | amia.org
Future Work
- Physicians expressed the need for UI changes based on preferences and specific diseases
- More research is needed in using BlockIoT for:
- Insurance claims
- Authorizing Tests
- Malpractice Lawsuits
16
AMIA 2021 Annual Symposium | amia.org
Summary
- BlockIoT enables connectivity between personal health devices and EHR systems
- It monitors, analyzes, ands intervenes in case a health metric is abnormal.
- A 13-physician participatory survey was conducted to determine the impact and utility of BlockIoT in a physician’s workflow
- Physicians expressed a need for standard access for medical data, and found BlockIoT to be useful in their practice
For more information, please contact: Manan Shukla (shuklm@rpi.edu), Dr. Oshani Seneviratne (senveo@rpi.edu), Dr. Jianjing Lin (linj17@rpi.edu)
Thank you!