Revolutionizing Point of Care with Remote Healthcare Solutions Lance Myers, PhD
How can wireless medical devices enable remote medical care and lower healthcare cost?
Confidential and Proprietary, Sentrian Distribution of healthcare spending Expenditure as % of GDP
Confidential and Proprietary, Sentrian Distribution of healthcare spending $0 $2,000 >$3,200 Average Monthly Cost
Confidential and Proprietary, Sentrian Cost breakdown of healthcare expenditure
Confidential and Proprietary, Sentrian Millions of Hospital Admissions are Preventable
Confidential and Proprietary, Sentrian To Combat Costs, Healthcare is Shifting to Fee-for-Value 0% * * * Department of Health and Human Services (HHS) targets Government Alternative Payment Targets
Confidential and Proprietary, Sentrian ACOs Are Being Created to Capture Shared Savings Health Affairs, March 31, 2015.
Confidential and Proprietary, Sentrian Disease Process Starts in Advance of Symptoms Patient Deterioration Time Target Detection Hospitalization Change in Weight Hemodynamic Parameter Changes Change in Fluid Levels Infection, Pollution, Malaise & Diminished Exercise Tolerance Increased Resting Heart Rate Decreased Peak Flow Reduced O2, Shortness of Breath COPD CHF 1 st Gen Detection
Confidential and Proprietary, Sentrian Remote Patient Intelligence Solution
Confidential and Proprietary, Sentrian Wide Array of Available Wireless Monitoring Devices
Confidential and Proprietary, Sentrian Collecting Symptomatic Data
Confidential and Proprietary, Sentrian CHF Hospitalizations by Age 75% of patients hospitalized for CHF are over the age of 65
Confidential and Proprietary, Sentrian Gateway Options
Confidential and Proprietary, Sentrian Traditional Remote Patient Monitoring Care team receives alerts Care team virtually manages condition Patient Hospitalized Patient Improves or Home hub uploads data Binary rules from single sensor Patient uses single sensor at home
Confidential and Proprietary, Sentrian
Confidential and Proprietary, Sentrian The problem of false alarms
Confidential and Proprietary, Sentrian The value of context Systolic BP = 172mmHg SpO2 = 93%
Confidential and Proprietary, Sentrian The value of context Systolic BP = 172mmHg Heart Rate is trending up
Confidential and Proprietary, Sentrian The value of context Systolic BP = 172mmHg PEF is trending down
Repeat measurements Actual Systolic BP distribution of repeat measurements on 6,314 patients First measurement > 180mmHg Difference between 1 st and 2 nd measurement Count
Confidential and Proprietary, Sentrian Machine Learning Actions Remote Patient Intelligence
Confidential and Proprietary, Sentrian Machine learning can identify predictive patterns CHFCOPDDMAFIBHTNBehavioral Heart Rate BP SP02 Weight Stroke Volume Activity Intrathoracic Fluid Peak Flow Glucose Sleep Respiratory Rate Self Report
Confidential and Proprietary, Sentrian Patterns may not easily be communicated or understood CHFCOPDDMAFIBHTNBehavioral Heart Rate BP O2 Sat Weight Stroke Volume Activity Intrathoracic Fluid Peak Flow Glucose Sleep Respiratory Rate Self Report Traditional machine Learning detects abstract patterns clinicians may not easily understand
Confidential and Proprietary, Sentrian Building a validated data driven model Identify Problem Source Datasets Transform Data Build Model Test Model Deploy Model 1 – 2 years
Confidential and Proprietary, Sentrian Models Required Chronic Disease Patients By Disease By Stage By Comorbidity By Exacerbation
Confidential and Proprietary, Sentrian Patterns can be expressed in a natural grammar Long term trend Short term trend Spike Shift High variability Low variability
Confidential and Proprietary, Sentrian Using an advanced rules-based system Care team receives alerts Care team virtually manages condition Patient Hospitalized Patient Improves or Home hub uploads data Binary rules from single sensor Patient uses single sensor at home
Confidential and Proprietary, Sentrian Using an advanced rules-based system Patient uses multiple sensors at home Care team sent notifications Care team virtually manages condition Patient Hospitalized Patient Improves Sentrian RPI evaluates data against rules expressed in a natural grammar and entered by physicians Data uploaded to secure cloud or
Confidential and Proprietary, Sentrian Sentrian system expresses rules in a natural grammar
Confidential and Proprietary, Sentrian Sentrian Remote Patient Intelligence Patient uses multiple sensors at home Care team sent notifications Care team virtually manages condition Patient Hospitalized Patient Improves Sentrian RPI evaluates data against rules Data uploaded to secure cloud Sentrian RPI learns from actual patient outcomes to improve rules
Confidential and Proprietary, Sentrian Sentrian’s Rule Recommendation Engine
Confidential and Proprietary, Sentrian WEAR-HeFT study conducted at Cedars Sinai Medical Center Sentrian results – Heart Failure Clinical Trial Physician rules 5 data streams Machine learning rules 5 data streams Machine learning rules 3 data streams