New Jersey Delaware Valley HIMSS Conference Atlantic City NJ October 29 2015 1 Speakers Dev Culver Executive Director HealthInfoNet Eric Widen CEO HBI Solutions 2 Agenda Background ID: 638615
Download Presentation The PPT/PDF document "Improving Care Coordination and Readmiss..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Improving Care Coordination and Readmissions Using Real Time Predictive Analytics from an HIE
New Jersey / Delaware Valley HIMSS ConferenceAtlantic City, NJOctober 29, 2015
1Slide2
Speakers
Dev Culver, Executive Director, HealthInfoNetEric Widen, CEO, HBI Solutions
2Slide3
AgendaBackgroundHealthInfoNet
HBI SolutionsCase study: St. Joseph’s HealthcareSummary and Q&A 3Slide4
Background:
HealthInfoNetQuick FactsFounded 2004
Independent, nonprofit organization
Operates the Maine HIEProvides a single, state-wide electronic patient health record
Real time data from provider electronic health records
Data is standardized
and
aggregated
Provides reporting and alerts
Disease
reporting to CDC
Real
time clinical rules and
alerts
Predictive risk scoresSlide5
HL7 messages
ADTLaboratory orders and resultsOutpatient prescriptions
Clinical notes and documents
CodingBackground: HealthInfoNet
Key Data
35
of 37 hospitals (all to connect in 2014)
38
FQHC sites
400+ ambulatory sites
Connections
>1.4 million patients
>600,000 annual encounters
>3500
users
Key StatisticsSlide6
Healthcare analytics company founded
in 2011Based in Silicon ValleyLeader in real time patient risk applications
Applications are used by providers, payers, ACOs and HIEs
HBI team collective experience
Mission:
Improve
patient and member health using data science to predict health risks and reduce practice variation
Unique
talent and experience:
Stanford researchers, data scientists
Frontline physicians
Performance improvement practitioners
Healthcare IT executives
6
Background:
HBI SolutionsSlide7
Background: HIE Analytic OfferingModules
Population Health30-Day Readmission / Return RiskVariation ManagementHospital PerformanceVolume
and Market Share
Identify populations and individuals most at risk for future high costs, inpatient admissions, and emergency room visits.
Identify inpatient encounters most at risk for 30-day readmissions or 30 day ED revisits.
Understand resource variation by disease and cost category (length of stay, laboratory, radiology, etc...) to reduce unnecessary practice variation.
Compare actual to target performance for key performance indicators (KPI) using case mix and severity adjusted targets, including statewide norms.
Track and trend volumes and market share by service area, disease, payer and patient demographics.Slide8
Patient History
Patient Risk of Event or Outcome
Risk Model Development
1000s of Patient Features
Age
Gender
Geography
Income
Education
Race
Diagnoses
Procedures
Chronic conditions
Visit and admission history
Outpatient medications
Vital signs
Lab orders and results
Radiology orders
Social characteristics
Behavioral characteristics
Multivariate Statistical Modeling – Decision Tree Analysis
Machine Learning
Available Risk Models
Population Risk Models
(predicts future 12 months)
Predicted future cost
Risk of inpatient admission
Risk of emergency
dept
(ED) visit
Risk of diabetes
Risk of stroke
Risk of AMI
Risk of
h
ypertension
Risk of mortality
Event Based Risk Models
(predicts future 30 days)
Risk
of 30 day
readmissionRisk of 30 day ED re-visit
Background: HIE Analytic OfferingPredictive Risk ModelsSlide9
Background: HIE Analytic OfferingPredictive Risk Use Cases
9Slide10
Background: HIE Analytic OfferingAdoption
Health SystemsFee for Service Community HospitalsACOsMedical Group with Insurance ProductState Medicaid ProgramFederally Qualified Health Centers
The following types of healthcare organizations are using the HIE analytic applications for predictive risk management, population health management, budget forecasting, market share intelligence, and throughput management.Slide11
Case study: St. Joseph Healthcare
11Slide12
Background: Saint Joseph HealthcareHealthcare system in Bangor ME
112 bed acute care community hospital Primary care and specialty physician practices20,000 covered livesPartner with FQHCParticipates in several ACOsMedicare shared savingsMedicaid
Commercial: CIGNA, Anthem, Harvard Pilgrim
Using real time predictive risk scores daily to manage patients
12Slide13
Workflow:Ambulatory Patient Risk Management
St. Joe’s Ambulatory Patient Population Management20,000 Patients
Assigned to St. Joe’s PCPs
Low Risk
Medium Risk
High Risk
Ambulatory based care managers assess real time population risk scores, including patient risks for costs, admission, ED visit, disease, and mortality.
The practice sets thresholds for each risk category to flag “high” risk patients.
Care managers proactively reach out to high risk patients to provide education and manage care gaps.
13Slide14
Workflow:Ambulatory Patient Risk Management
Population Health Dashboard / Patient List
– Understand patients at risk for ED visits, IP admissions, disease and costSlide15
Workflow:Ambulatory to Acute Patient Risk Management
St. Joe’s Ambulatory Patient Population 20,000 Patients Assigned to St. Joe’s PCPs
Low Risk
Medium Risk
High Risk
St. Joe’s Acute Inpatient and Emergency Patient Risk Management
6,000 Annual Inpatient Discharges
20,000 Annual Emergency Visits
% Visit St. Joe’s Hospital
% Visit St. Joe’s Hospital
Low Risk
Medium Risk
High Risk
Upon admission, hospital based care managers assess real time risk scores for 30 day return to the hospital, and develop appropriate discharge plans
Low Risk
Medium Risk
High Risk
Community At Large Population
15Slide16
Workflow:Ambulatory to Acute Patient Risk Management
Inpatient Encounter List– Understand patients at risk for 30 day readmissions
16Slide17
Workflow:Acute to Ambulatory Patient Risk Management
St. Joe’s Ambulatory Patient Population 20,000 ACO Patients Assigned to St. Joe’s PCPsLow Risk
Medium Risk
High Risk
St. Joe’s Acute Inpatient and Emergency Event Management
6,000 Annual Inpatient Discharges
20,000 Annual Emergency Visits
Low Risk
Medium Risk
High Risk
Post discharge, patients assigned to St. Joe’s PCPs are handed off to the ambulatory
c
are
m
anager for follow up. Patient’s risk drives the post-discharge care plan.
Low Risk
Medium Risk
High Risk
Community At Large Population
17
Patients discharge back to homeSlide18
Reduced readmission rate overall and below target
Results: St. Joe’s Readmission Rate Trend
18
Actual Performance
Target (State Adjusted Norm) PerformanceSlide19
Saint Joseph Summary FindingsReal time risk scores using clinical data from EHR
Time savings and productivity improvementAlgorithms have identified at-risk patients that would have been missedAlgorithms provide better prediction at the higher risk levelsHIE provides longitudinal patient record across MaineRisk scores are helpful - a robust care team and processes, however, are required to impact patient outcomes
We use analytics across the care continuum
19