value based care analytics - florida chapter of hfma€¦ · value based care analytics an overview...
TRANSCRIPT
![Page 1: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/1.jpg)
Value Based Care AnalyticsAn Overview of Measures, Metrics, and Enabling TechnologiesMay 2019
![Page 2: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/2.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 2
BiosThe road to action from data begins with data collection and relies on the integration of analytics into management review and staff workflows
Pam PriceSpecialist LeaderAtlanta, GA
Pam has more than 25 years of experience spanning Health Plans, Provider groups, and pharmaceuticals. Pam’s expertise includes risk adjustment, product development, strategic planning and execution, financial leadership and mergers and acquisitions
Brendan NolanSenior ConsultantNew York, NY
Brendan has over 9 years of experience working with both plans and providers to integrate operations surrounding value based care. His focus is on creating common analytics that allow collaborative programs
Angie DiazManaging DirectorKansas City, MO
Angie is a managing director in the health care provider sector with over 20 years of experience in the industry. Her primary focus is revenue cycle process improvement and revenue management operational re-design
![Page 3: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/3.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 3
Section Topic
1 Background/Context
2 Common Analytics & Dashboard Samples
3 Advanced Technologies
4 Analytics Roadmap
Agenda
![Page 4: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/4.jpg)
Background & ContextPerspective on the future of Value Based Care
![Page 5: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/5.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 5
1. US Census Bureau https://www.census.gov/newsroom/stories/2019/older-americans.html2. CMS: https://www.cms.gov/research-statistics-data-and-systems/statistics-trends-and-reports/nationalhealthexpenddata/nationalhealthaccountshistorical.html
Unsustainable Healthcare Spend
2026 US spend:
(projected)
$5.7 trillion
19.7% of GDP2
2016 US spend:
$3.3 trillion
17.9% of GDP2
Healthcare Market ForcesPopulation aging coupled with healthcare spending trends have been shifting health plan and provider focus to value over volume since the 2010 Affordable Care Act
Value Focused Legislation
2010’s Affordable Care Act remains in
place and is pushing health plans and
providers towards value based
reimbursement
Aging US
Demographic
2023 US population
> 65 (projected):
62 Million1
2019 US population
> 65:
48 Million1
![Page 6: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/6.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 6
Varied or misaligned
incentives by
stakeholder
Lack of consensus
on standards & priorities
Uncontrollable factors (e.g.
patient adherence to
therapy)
Uncertain regulatory
environment expectations
Challenges in
collecting and
analyzing necessary
data
Lack of enabling
systems to
administer program
The Conundrum: Current barriers are slowing widespread adoptionCreating successful value based programs is one of the biggest challenges in health care. There are many challenges that must be addressed in order to effectively drive adoption of value based care programs
![Page 7: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/7.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 7
Market Forces are Pushing Providers Towards Accepting RiskBut each organization is moving at their own pace as they: a) evaluate their ability managing more risk, and b) develop the necessary tools and skills needed
Fee for Service Full CapitationBundled Payments
Upside/Downside Contracts
Upside Only Contracts
As provider organizations move towards accepting incremental risk, many will increasingly
rely on clear insights from analytics to guide their operations
Low Risk High Risk
Leaders need insight on performance against ‘bonus’ metrics that exist in upside only contracts (e.g. quality measures, annual wellness visit completion)
Leaders need a clear picture of their patient panel’s risk and how their
organization is managing the overall costs of that population
Leaders need visibility into how costs are managed by episode of care, by risk stratification, and
by chronic condition
Leaders need to have a complete understanding of how to manage costs and practice patterns within their provider network, and begin to have analytics proactively suggest
areas for tighter focus/control
![Page 8: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/8.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 8
Value Based Care Analytics – Not One-Size-Fit-AllEach health system might approach acquiring needed analytics differently based on size and organizational structure
Large Single EMR Health System ACO/ MSO/ MCO IPA Small Practice
Traits:• Well resourced• Single EMR system• Aligned goals/operations• Low ratio of analytics overhead cost/
provider
Traits:• Well resourced• Potential single EMR vendor• Multiple larger EMR instances• Potentially competing
goals/operations• Medium ratio of analytics overhead
cost/provider
Traits:• Resourcing varies• Potentially aligned EMR vendors• Multiple smaller EMR instances• Varied access to data• Potentially competing
goals/operations• High ratio of analytics overhead
cost/provider
Traits:• Resource constrained• Smaller EMR instance• Varied access to data• Aligned goals/operations• Medium-High ratio of analytics
overhead cost/provider
Typical Method for Analytics:• Build out reporting data warehouse• Evaluate purchasing functionality
from EMR vendor• Advocate for raw data from health
plan
Typical Method for Analytics:• Build out reporting data warehouse
using custom data transformations• Evaluate EMR consolidation
incentives• Advocate for raw data from health
plan
Typical Method for Analytics:• Build out reporting data warehouse
using data exchange standards• Evaluate EMR consolidation
incentives• Advocate for raw data from health
plan
Typical Method for Analytics:• Advocate for raw data from health
plan• Advocate for packaged reports from
Health Plans• Determine what features may be
available through EMR vendor
Key Take-Away• What to build
Key Take-away:• What to build
Key Take-away:• What to build
Key Take-away:• What to request from health plan
and EMR vendor
![Page 9: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/9.jpg)
AnalyticsThe Connective Tissue Between Data and Efficient Decision Making
![Page 10: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/10.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 10
Dashboard DomainsValue based care analytics typically align with the foundations of risk bearing contracts. Metrics included in each dashboardwill vary based on the population managed and the specifics of your contracts.
Medical Cost Management
• Utilization by Service Category
• Utilization by Site of Service
• Cost by Risk Stratification
• Cost by Chronic Condition
• Cost by Episode of Care
• Network Referral Patterns
• Admission/Readmission Patterns
• Generic Rx Rate
Risk Adjustment
• Annual Wellness Visit Completion
• Demographics, Status, Panel Shifts
• Predictive Analytics
• Chart Review/Coding Quality
• Financial Projections
• Integration at the point of care
Quality Measures
• Overall Stars Performance
• Gap Closure Lists (By Measure)
• Gap Closure Lists (By Patient)
• “Propensity to Act”
• Integration at the point of care
Measures of provider or patient behavior tied to clinical quality, safety, or contractual incentives
Measures of the illness burden of your population that should ultimately be reflected in any capitated payments
Measures of utilization, price, and network referral patterns that drive medical spend
![Page 11: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/11.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 11
Example - Quality Dashboard #1 – Overall Measure Dashboard
Reports as of:09/30/2018
QualityGlobal Filters
Chronic Conditions
(All)
Ethnicities & Races
(All)
PPPM
(All)
Risk Score
(All)
Attribution Type
(All)
Benchmark
(All)
Year
Year Type (Calendar/Trailing)
(All)
Genders
(All)
Languages
(All)
Age Ranges
MA
Payers
(All)
Member Products
(All)
Provider
(All)
Overall Performance
(All)
79%
Overall Performance
Cancer Screening
Disease Prevention
Disease Management
Annual Breast Cancer Screen
Colorectal Cancer Screening
Reducing Risk of Falling
Annual Flu Vaccine
Monitoring Physical Activity
Adult BMI Assessment
Diabetes Care – Eye Exam
Diabetes Care – Kidneys…
Diabetes Care – Blood Sugar
Controlling Blood Pressure
MeasureDomain NumPerformance Denom # to Target
%
24,231
42,344
16,734
59,012
59,012
78,023
12,986
12,986
12,986
16,054
20,596
34,298
12,048
25,375
44,259
67,880
9,350
8,830
12,077
13,485
85%
81%
72%
43%
75%
87%
72%
68%
93%
84%
485
3,811
3,849
23,605
1,770
0
3,376
2,727
0
803
Compared to Calendar Year
74% at this point in time
last year
87%Compared to Trailing Year 83% at this point in time last year
4.5/5
4.0/5
CY
PY
Display “Trailing Year” to reflect how well Providers current behaviors would be scored
Display “Calendar Year” to reflect how much work remains to achieve the desired outcome
For most dashboards, allow the fields to be displayed to be configurable. For quality metrics, consider including thresholds, prior year scores, exclusions, % to target.
Create click-throughs that drill down to the non-compliant patients for an individual measure or across all measures
1 2
3
4
1
2
3
4
![Page 12: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/12.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 12
Example - Quality Dashboard #2 – Patient/Measure Level Detail
“Propensity to Act” algorithms allow users to create outreach programs tailored towards patients based on their likelihood to engage and take action
Patient lists should include all relevant information for outreach. Strong patient lists will include preferred contact methods and outreach history information.
Patient list should enable a user to see all gaps for a patient. One outreach should be done to address and discuss all gaps.
Reports as of:09/30/2018
QualityGlobal Filters
Chronic Conditions
(All)
Ethnicities & Races
(All)
PPPM
(All)
Risk Score
(All)
Attribution Type
(All)
Benchmark
(All)
Year
Year Type (Calendar/Trailing)
(All)
Genders
(All)
Languages
(All)
Age Ranges
MA
Payers
(All)
Member Products
(All)
Provider
(All)
(All)
1
2
3
1
2 3
Care System D
Care System D
Care System D
Practice A
![Page 13: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/13.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 13
Risk AdjustmentIn the Value Based Care world – Risk Adjustment Analytics most strongly correlate with analytics on your organization’s top line revenue (leaving cost and quality bonus measurements aside)
Domain Categories
Demographics, Status, and Patient Panel Shifts – General demographic information and population shifts (separates signal from noise)
Predictive Analytics – Identification of potential HCC Conditions in the population needing validation, results in adds and deletes
Financial Forecasting – Estimation and projection of expected revenue based on population risk
Chart Review/Coding Quality/Productivity – Productivity and efficiency metrics for coding staff
Initiative Reporting – Reporting of operational initiatives for risk capture (e.g. Wellness visits)
Compliance – Turnaround time and chart reviews
Note: While there are many topics that a report consumer may want to know about that are not readily apparent in the domain categories, those topics can be addressed through further filtering the data
Patient Encounters – For the encounter year, how well is the health system able to hold wellness visits with their population
![Page 14: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/14.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 14
Example - Risk Adjustment Dashboard #1 – Patient Encounters
A health system may have focused operational initiatives aimed at engaging patients for their Annual Wellness Visits and/or In-Home Assessments.
Dashboards should be used to measure the success and/or failure points of the engagement initiatives
An entry level Patient Encounter dashboard will track Annual Wellness visit completions
More advanced dashboards will rely on developing predictive analytics that will suggest areas where conditions are likely to be incomplete or inaccurate
20
,00
01
0,0
00
Fully Documented
HCC
HCC Documented Still Predicted
Had Encounter but no HCC
Interaction but no
EncounterNo Interaction
80,000 (+0.9%) 55,000 (+1.0%) 65,000 (+1.2%) Population 60,000 (+1.4%) 40,000 (+1.3%)
26.7% (-.01%) 18.3% (-.02%) 21.7% (+0%) % Population 20.0% (+.01%) 13.3% (+.01%)
N/A 3% (-1.1%) 4% (-0.9%)Non Recaptured HCC Distribution
5% (-0.8%) 3% (+1.3%)
Patients w/ Encounter
+1.2% (YOY)
67%200,000 patients
Refresh Rate
+1.2% (YOY)
85%
Risk Adjusting Encounters as of 05/01/2019
Patient Encounter Programs
Need for Encounters
% Reached% Scheduled Appointment
% with Encounter
% Reached% Scheduled
Home Assessment
% with Encounter
80%16,000 patients
65%6,500 patients
60%12,000 patients
30%3,000 patients
25%2,500 patients
50%10,000 patients
75% of reached
46% of reached
83% of scheduled
83% of scheduledJ F M A M J J A S O N D
Pat
ien
ts w
/ En
cou
nte
rs
Month
20k
10k
15k
5k
25kAnnual Wellness Goal
Annual Wellness Actual In-Home Assessment Actual
In-Home Assessment Goal
IHA
Tar
gete
d
Pat
ien
tsA
WV
Tar
gete
d
Pat
ien
ts
12
,50
0 (
41
.7%
)
Tota
l P
atie
nts
wit
h E
nco
un
ter
( %
Tar
gete
d)
NOTE: Parentheses = YOY change in rate
Reports as of:09/30/2018
Risk Adjustment Dashboard
Global Filters
Demographics:
Stayers/Leavers/Joiners
(All)
Age
(All)
Patient Cohort
(All)
Gender
(All)
MA Risk Model
(All)
Diagnoses:HCC
(All)
Benchmark
(All)
Organizational:Region/Medical Center/Provider
(All)
Internal/External
(All)
Population/Contract:
State County Code (SCC)
(All)
Product
(All)
MA/ ACA
(All)
Attribution Method
(All)
Provider Specialty
(All)
Export Patent Detail
1
2
3
4
1
2
3
4
![Page 15: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/15.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 15
Example - Risk Adjustment Dashboard #2 – Condition Prevalence
0%
5%
10%
15%
20%
25%
0% 5% 10% 15% 20% 25%
HCC HCC DescriptionNational
Benchmark HCC Prevalence Patients
Variance to
National Rate
107 Vascular Disease With Complications 1.82% 1.44% 375 -20.88%
108 Vascular Disease 17.99% 7.90% 2053 -56.09%
017 Diabetes With Acute Complications 0.36% 0.28% 74 -22.22%
018 Diabetes With Chronic Complications 20.24% 12.18% 3166 -39.82%
019 Diabetes Without Complications 9.11% 6.74% 1752 -26.02%
122 Prolif Diabetic Retinopathy & Vitreous Hemorrhage 0.90% 0.41% 107 -54.44%
021 Protein-Calorie Malnutrition 1.80% 0.77% 199 -57.22%
085 Congestive Heart Failure 12.87% 7.75% 2014 -39.78%
096 Specified Heart Arrhythmias 12.91% 10.63% 2763 -17.66%
111 COPD 15.89% 8.84% 2297 -44.37%
134 Dialysis Status 0.22% 0.08% 22 -63.64%
135 Acute Renal Failure 3.99% 2.37% 615 -40.60%
136 CKD, Stage 5 0.23% 0.07% 19 -69.57%
137 CKD, Severe (Stage 4) 0.89% 0.55% 144 -38.20%
058 Major Depressive, Bipolar, and Paranoid Disorder 11.52% 8.94% 2325 -22.40%
Metric Description Value
Current Yr. Avg. HCCs/Mem. 1.45
Prior Year HCCs/Patient 1.50
Present YOY % Difference -3.44%
Forecast HCCs/Patient 1.55
Forecast YOY % Difference +3.33%
Total Predicted HCCs 50,834
Avg. Conversion Rate 9%
Projected Captured HCCs 4,576
Condition Prevalence v. Benchmark
Benchmark Prevalence
Act
ual
Pre
vale
nce
Condition Capture Rate
Export High Variance Patients
Significant VarianceActual = Benchmark(Size) = # of Patients
Note: Codes from seasonal conditions (e.g. pneumonia) are excluded from the graph and chart
HCC: 111Description: COPD
Patients: 2297
Reports as of:09/30/2018
Risk Adjustment Dashboard
Global Filters
Demographics:
Stayers/Leavers/Joiners
(All)
Age
(All)
Patient Cohort
(All)
Gender
(All)
MA Risk Model
(All)
Diagnoses:HCC
(All)
Benchmark
(All)
Organizational:Region/Medical Center/Provider
(All)
Internal/External
(All)
Population/Contract:
State County Code (SCC)
(All)
Product
(All)
MA/ ACA
MA
Attribution Method
(All)
Provider Specialty
(All)
Any “System Level” analytics should have the capability to drill down via filters or export raw details. Exports should be used to allow customers to take action on the missed or inaccurate conditions.
Advanced practices will find a way to incorporate these actions into the EMR, so the actions are present at the point of care.
Any predictive analytics should be balanced to look for missing, as well as potential coding errors. Analytics looking only for missing codes limit the opportunity to identify anomalies or trends in the data that may result in compliance risk.
1
2
1
2
![Page 16: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/16.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 16
Investigation of Medical Cost Management Opportunities begins with viewing cost trends on aggregate and performing deeper dives to understand specific drivers
Medical Cost Management Analytics
Evaluate utilization of services or service categories against benchmark e.g.Advanced Imaging, ED Days
Utilization by Service Category
STL Best-In
Class
Before After
Provide the data to identify the appropriate level of care within the cost appropriate setting
Utilization by Site of Service
$$$
$$
$
Cost by Condition
Evaluate costs and utilization to manage chronic conditions against regional benchmarks
Types of Opportunities Typically Identified
Cost by Risk Stratification
Evaluate how effectively your providers are managing their highest risk, rising risk, and healthy populations
Network Leakage
Evaluate referral patterns, and procedures that commonly occur out of network
Cost Per Episode of Care
Cost Per Episode of care (commonly bundled services) e.g. Comprehensive Joint Replacement, Cardiac Rehabilitation
IP/ED/SNF Admission Management
Track Admissions, Re-admissions, preventable re-admissions, Average length of stay and discharge to locations for inpatient settings of care
Generic Rx Rate
Monitor Generic Prescribing rates for providers to identify areas where lower cost alternatives are available
![Page 17: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/17.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 17
Example - Medical Cost Management (MCM) Dashboard
Reports as of:09/30/2018
MCMGlobal Filters
Chronic Conditions
(All)
Ethnicities & Races
(All)
PPPM
(All)
Risk Score
(All)
In/Out of Network
(All)
Benchmark
(All)
Date Range
PCPs
(All)
Genders
(All)
Languages
(All)
Age Ranges
MA
Payers
(All)
Member Products
(All)
DRGs
(All)
Service Provider
(All)
POS
Hospital Outpatient
Primary Dx
(All)
Primary Procedure
(All)
MM/DD/YYMM/DD/YY
A robust set of filters enables real-time free-form investigation of cost drivers and allows for flexible investigation
Benchmarks will vary depending on what data is available:
• Purchased benchmarks
• Historical data
• Plan provided data
Utilization per 1000
Utilization per 1000
J F M A M J J A S O N DCurr
ent
Year
Prior
Year
Hospital Outpatient
Chemotherapy
Home Health
Observation Days
Occupational Therapy
Outpatient - Dialysis
Outpatient - ER
Outpatient – Mental Health
Outpatient – Subs Abuse
Outpatient Surgery
Physical Therapy
Speech Therapy
Service Category (Minor)Service
Category
(Major)
Health System
Utilization
Benchmark
Utilization
% DiffCurrent Annualized
Expenses
Savings
Opportunity
An Effective Medical Cost Management Dashboard should answer:
- How are we doing managing costs?
- Where are the opportunities?
- What is the quantified opportunity?
- How can we take action?
Where the opportunities exist may involve exploring filters and views – but each view should make the answers to the other questions apparent
50.0
2.6
3.7
93.4
161.9
53.1
22.0
143.1
230.3
17.2
31.1
7.3
5.4
37.2
165.7
46.1
11.1
125.2
157.1
11.3
37.8%
N/A
N/A
60.2%
N/A
13.2%
49.5%
17.9%
31.8%
34.3%
$23M
$54M
$12M
$65M
$120M
$8M
$10M
$36M
$43M
$12M
$8.7M
N/A
N/A
$39.13M
N/A
$1.1M
$4.95M
$6.4M
$13.7M
$4.1M
18.7 15.0 19.8% $56M $11.1M
1
1
2
2
![Page 18: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/18.jpg)
Advanced TechnologiesAnd Their Corresponding Jargon
![Page 19: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/19.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 19
Predictive Analytics are used throughout the Healthcare industry to estimate the likelihood of future outcomes based upon historical data or related data
Risk Scoring for chronic
diseases and population
health
Avoiding 30-day Hospital
Readmissions
Averting Patient No-
shows
Predicting Patient
Utilization Patterns
Managing the Supply Chain
Ensuring Strong Data
Security
Developing Precision Medicine and New Therapies
Improving Patient
Engagement and
Satisfaction
Identify high risk and rising risk patients to help avoid developing long-term health conditions that are costly and complex to treat
Use behavioral patterns to create meaningful care plans, predict a patient’s willingness to participate, and prioritize those patients who are likely to engage in healthy behaviors to reduce the risks of chronic diseases
Supplement traditional clinical trials and drug discovery techniques using analytics. Providers are able to choose therapies with high success of likelihood by matching patient genetic information with previous patients’ results.
Identify patients when risk factors occur during a hospital stay (infection development, extended length of stay), indicating a high likelihood of readmission within the 30-day window.
Identify patients likely to skip appointments without advance notice.
Providers can send additional reminders to at-risk patients, offer transportation or suggest rescheduling and offer open slots
to other patients with acute needs.
Source: 10 High-Value Use Cases for Predictive Analytics in Healthcare, HealthITAnalytics newsletter
Monitor patterns in data access, sharing and utilization to create early
warnings when unexpected activity occurs
Support decisions around negotiating pricing, reducing supply variation and
optimizing the order process
Anticipate utilization patterns to enable optimal staffing levels, help reduce wait
times and raise patient satisfaction. Visualization tools can be used to model
patient flow patterns and highlight opportunities to make workflow
adjustments or scheduling changes.
Practice
Management
Medical
Management
![Page 20: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/20.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 20
Robotic Process Automation & Cognitive Automation
Many healthcare organizations are starting to jointly apply robotic process automation and cognitive automation to a broad spectrum of processes, including Utilization Management (UM)
Deterministic Outcomes Probabilistic Outcomes
Realm of Robotic Process Automation (RPA) Realm of Cognitive Automation (C.A.)
“Mimics Human Actions”
Robotics Process Automation is the process of automating simple defined tasks (think ‘Macros’)
RPA Realm:• Rules-based tasks• Operational processes
“Augments Human Intelligence”
Cognitive Automation is the process of capturing a significant volume of decisions made on a set of criteria, and using that data to construct/train a model that can probabilistically predict the outcomes of future decisions on similar criteria
Cognitive Realm:• Cognitive analytics• Decision making
Robotics Process Automation and Cognitive Automation, in brief:
Emerging health system core operations use cases include:
Prior Authorizations: Augment prior-auth review using advanced analytics to further automate clinical decision making (vs. manual review) and triage cases to appropriate staff level
Appeals and Grievances: Triage appeals and grievance cases to appropriate staff level for review and/or decisioning
Predicting Health Conditions: Measure correlation between related datasets (e.g. Prescriptions and vitals) to confirmed health conditions (on the problem list)
![Page 21: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/21.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 21
Optical Character Recognition (OCR) and Natural Language Processing (NLP)
OCR and NLP are commonly used jointly to ‘digitize’ paper records (fax or EMR Notes) and parse them for key insights
Optical Character
Recognition
Natural Language Processing
Ocular Character Recognition (OCR) can be used to convert Faxes or PDFs into discrete data elements that can be used by machines
Natural Language Processing (NLP) logic is used to convert ‘free text’ into complex relationships based on key words found in the free text
Applications
Convert fax or paper mail sent in for a prior-authorization request into a digital format that can automatically be filed and stored with other meta data (sender, date of receipt)
Transform paper electronic medical charts into digital text that can be searched and reviewed using automated processes
Comb through text in prior-auth requests to pull out data elements for patient, provider, diagnoses, and requested procedure. Once the data elements have been accessed, rules can be created to triage the request. Combined with OCR, fax or mail requests can be automated.
Comb through the free text in electronic medical records to identify potential candidates for HCC codes that have been missed or miscoded to ensure more accurate representation of risk
Parse and Triage Customer Service Requests
![Page 22: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/22.jpg)
Getting Started
![Page 23: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/23.jpg)
Copyright © 2019 Deloitte Development LLC. All rights reserved. 23
Analytics FrameworkThe road to action from data begins with data collection and relies on the integration of analytics into management review and staff workflows
Data Collection Analytics Operations
Advocate for Raw Data
Request data from ecosystem partners to be used in analytics:
1) Claims data from your payers
2) Lab feeds from lab partners
3) ADT feeds from inpatient partners
4) Cost data
5) Referral network data
Create a Data Lake
Store data in its native, raw format so that it is unconstrained in future uses
Integrate and Harmonize
(Data Warehouse)
Build logic to ensure that data from the various sources ‘speaks the same language’ and can be used for reporting purposes
Measure Actions and Outcomes – Begin by measuring operational tasks (completions of actions) and outcomes – how your organization performed on measures that are important to the business
Define the responsible parties and review cadence of ‘Summary Level’ analytics which define direction. Determine how data can be meaningfully inserted at the point of care.
• Define the list of metrics that leadership will manage by and define the cadence for which they will be created and reviewed
• Across Medical Cost Management, Quality, and Risk Adjustment, define the top metrics that impact overall performance
• Benchmark like staff against population to identify outliers and investigate need for additional training or remediation
Enable Physicians, Nurses, and Care Team Staff to take action by placing the relevant data in their hands at prescient times for enhanced monitoring and clinical decision-making
Test and Validate Hypotheses for how to improve outcomes – Predictive algorithms can be used to model the hypotheses, machine learning can be used to measure effectiveness and suggest better hypotheses
Package and Deliver subsets of data to staff at the appropriate moments with the appropriate insights – Integration with scheduling and customer centric data sets can enable staff with the right data at the right time
![Page 24: Value Based Care Analytics - Florida Chapter of HFMA€¦ · Value Based Care Analytics An Overview of Measures, Metrics, and Enabling Technologies ... • Build out reporting data](https://reader033.vdocuments.site/reader033/viewer/2022060405/5f0f1edd7e708231d442975e/html5/thumbnails/24.jpg)
About DeloitteDeloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.
This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication.
Copyright © 2019 Deloitte Development LLC. All rights reserved.