Evaluating the Value of Healthcare Information Technology: New Studies on Return on Investment from

1
Evaluating the Value of Healthcare Information
Technology New Studies on Return on Investment
from HIT Adoption from the CITL

• Blackford Middleton, MD, MPH, MSc, FACMI
• Chairman, Center for IT Leadership
• Corporate Director Clinical Informatics RD
• Partners Healthcare
• Harvard Medical School

2
Overview

• Highlights from CITL studies of HIT value
• Value of Ambulatory CPOE
• Value of Healthcare Information Exchange and
  Interoperability
• Value of IT-Enabled Chronic Diabetes Management
• Value of Physician-Physician Telehealth
• Value of PHRs
• Insights and Lessons learned
• Contributors to value, detractors/barriers,
  implications
• Conclusions/QA

3
CITL Mission

• Produce timely, rigorous market-driven technology
  assessments which
• Help providers invest wisely
• Help IT firms understand value proposition
• Provide leadership for IT in healthcare
• Established at Partners Healthcare in partnership
  with HIMSS
• C!TL Improving Healthcare Value

www.citl.org
4
The Value Proposition for HIT

• Headlines
• Value of ACPOE suggest
• 28K savings per provider
• 44B savings potential nationally
• Value of Healthcare Information Exchange
• 78B year nationally
• Value of IT in Chronic Diabetes Management
• On average in excess of 5000/diabetic, or 20B
  for diabetes related care
• Value of Telehealth
• In press
• Value of PHRs
• TBD

5
US Healthcare System Will Benefit with ACPOE

• National adoption of Advanced ACPOE systems would
  prevent
• 2 million ADE/yr
• 190,000 ADE admission/yr
• 130,000 life-threatening ADE/yr
• Nationwide implementation of advanced ACPOE
  could
• Save the US 44 billion annually

www.citl.org
6
Value of HIEI Key Findings

• Standardized, encoded, electronic healthcare
  information exchange would
• Save the US healthcare system 337B over a
  10-year implementation period
• Save 78B in each year thereafter
• Total provider net benefit from all connections
  is 34B
• Net benefits to other stakeholders
• - Payers 22B - Pharmacies 1B
• - Laboratories 13B - Public Health 0.1B
• - Radiology centers 8B
• Dramatically reduce the administrative burden
  associated with manual data exchange
• Decrease unnecessary utilization of duplicative
  laboratory and radiology tests

Walker, J et al Health Aff 2005 Jan 19
7
Value of ITDM Key Findings

• Many diverse approaches to chronic disease
  management in DM
• Considerable expenditure on both payer and
  provider side
• Evidence base weak
• Potential savings not as impressive as ACPOE or
  HIEI
• On average in excess of 5000/diabetic, or 20B
  for diabetes related care

8
Diabetes in the US

• US prevalence 14.6 million diagnosed, 6.2
  million undiagnosed
• Our research focuses on Type-2 diabetes
• 90-95 of all diabetics
• Adult-onset
• Impaired response to insulin to regulate blood
  sugar
• Tremendous annual incidence of co-morbidity
• Every day diabetics across the country suffer
  harm
• 33 people become blind
• 118 begin treatment for kidney failure
• 225 undergo lower extremity amputations
• Estimated direct cost of 92 Billion in 2002

American Diabetes Association
9
Potential to Manage Diabetes

• Improved understanding of diabetes risk factors
• Improved weight control lowers the risk of
  diabetes
• Earlier diagnosis and treatment
• Diabetes is diagnosed years earlier, providing
  opportunity to treat before complications occur
• Tight management of diabetes prevents disability
• Clinical trials show that tight blood sugar and
  blood pressure control prevent many
  complications, including stroke, heart attack and
  death
• Advances in treatment
• New drugs emerge to control sugar and prevent
  complications

Therefore, there are more opportunities for
management and prevention of complications
10
Clinical Care Failures
McGlynn, et al The Quality of Health Care
Delivered to Adults in the United States
Appendix. RAND, July 2004
11
CITL ITDM Study Goals

• Can IT-enabled disease management change the
  course of a chronic disease?
• Can it reduce the rate of complications?
• Can it improve quality of life?
• Can it reduce care costs?
• Identify long term value of IT Enabled Diabetes
  Management that reflects
• Program costs
• Patient turnover
• Changes in care processes
• Changes in disease progression

12
Evidence Gathering

• Literature review
• Over 800 academic and general/trade sources
• Market research
• Phone interviews with 50 organizations
• DMAA partner in surveying vendors, plans, and
  providers on DM program and IT costs
• Expert panel
• Day-long meeting, ongoing consultation and
  feedback

13
Expert Panel

• Madhu Agarwal, MD, Acting Deputy Chief Officer of
  Patient Care Services, Veterans Administration
• Brian Austin, Deputy Director, The Improving
  Chronic Illness Care Program, Group Health
  Cooperative, Seattle
• Stephen J Brown, President and CEO, Health Hero
  Network
• Lawrence P Casalino, MD, PhD, Assistant
  Professor, University of Chicago
• Tim Ferris, MD, M.Phil, MPH, Director of
  Pediatric Quality Improvement, Mass General,
  Partners HealthCare, Boston
• Jeremy M Grimshaw, MBCHB, PhD, FRCGP, Director,
  Centre for Best Practice, University of Ottawa
• Karen Kuntz, ScD, Associate Professor, Harvard
  School of Public Health
• John A Merenich, MD, Regional Director, Kaiser
  Permanente Colorado
• David Wennberg, MD, President and COO, Health
  Dialog Data Services
• Special thanks to the Disease Management
  Association of America (DMAA) and Karen Fitzner,
  Director Research and Program Development

14
DM Program Components May Target Patients and/or
Providers

• Some common DM program interventions
• Personal Health Risk Assessment
• Patient self-management support (education from
  certified diabetes educator, nurse case manager)
• Electronic and paper-based guidelines
• Care teams (MD, CDE, RN, others)
• Physician education
• Feedback to providers on guideline compliance,
  care quality (care audit)
• Orders support (CPOE)

15
ITDM Taxonomy

• Technologies used by payers
• Technologies used by providers
• Disease registries
• Clinical decision-support systems
• Technologies used by patients
• Self-management
• Remote monitoring
• Integrated provider-patient systems

16
ITDM Model Overview
Input
Output
Financial and Clinical Benefit
Diabetic Population
Diabetes Simulation Model
minus
ITDM Impact
ITDM Cost

Yr 0
Yr 10
Yr 5
NET VALUE
17
ITDM Impacts Positive Example
7.60
Evidence suggest payer interventions can improve
control of blood sugar
7.33
18
ITDM Impacts Negative Example
No published evidence to suggest self management
improves foot screening rates
44.9
44.9
19
Disease Burden Engine

• Based on a CDC-RTI diabetes disease model
• Simulates the progression of Type-2 diabetes in a
  population ages 25-94
• Markov model projecting progression in five major
  diabetic complications
• Cardiovascular
• Cerebrovascular
• Renal
• Ocular
• Neuropathic
• Modified to show the impact of improved care
  processes on clinical outcomes

HOERGER TJ, RICHTER A, BETHKE AD, GIBBONS CB A
MARKOV MODEL OF DISEASE PROGRESSION AND
COST-EFFECTIVENESS FOR TYPE 2 DIABETES. 2002. NO.
RTI PROJECT NUMBER 6900.016.
20
Evidence Impacts in the Model
Available Evidence
Complications
Rate of Eye exams
Rate of Microalbuminuria Screening
Rate of Foot Exams
Change in Systolic BP
Change in Cholesterol
Change In HbA1c
21
Projecting Impact on Disease Progression

• We rely on benchmark clinical trials to project
  the impact of process improvements on disease
  progression

Kendrick DC, Bu D, Pan E, Middleton B. Crossing
the Evidence Chasm Building evidence bridges
from process changes to clinical outcomes. J Am
Med Inform Assoc. 2007 May-Jun14(3)329-39. Epub
2007 Feb 28.
22
Cost Model Approach

• Collect cost data for ITDM programs described in
  academic literature
• Challenges
• Not published
• Proprietary
• Varied implementation approaches
• Solution
• With support from the Disease Management
  Association of America, contacted member
  organizations to share cost data anonymously
• Conducted over 50 cost interviews
• Synthesized results into scalable model that
  mirrors payer and provider taxonomies

23
ITDM Costs
Adler-Milstein J, Bu D, Pan E, Walker J, Kendrick
D, Hook JM, Bates DW, Middleton B. The cost of
information technology-enabled diabetes
management. Dis Manag. 200710(3)115-28.
from over 50 interviews with DMAA member
organizations
24
Detailed ITDM Model Overview
Input
Output
QALY
Morbidity
Diabetic Population
Mortality
Medical Costs
minus
ITDM Impact
ITDM Costs

NET VALUE
Yr 0
Yr 5
Yr 10
25
ITDM Model Architecture
26
Care Processes Results 10th Year Screening
Rates
Bu D, Pan E, Walker J, et al. Benefits of
Information Technology-Enabled Diabetes
Management. Diabetes Care 10.2337/dc06-2101.
27
Physiology Results10th Year Average Value
n/s denotes a statistically non-significant
result at alpha0.05.
Bu D, Pan E, Walker J, et al. Benefits of
Information Technology-Enabled Diabetes
Management. Diabetes Care 10.2337/dc06-2101.
28
Mortality Results Reduction in 10 Year
Cumulative Incidence
Bu D, Pan E, Walker J, et al. Benefits of
Information Technology-Enabled Diabetes
Management. Diabetes Care 10.2337/dc06-2101.
29
Financial Results 10 Year Cumulative Net
Present Value
Cost of care savings results published in Bu
D, Pan E, Walker J, et al. Benefits of
Information Technology-Enabled Diabetes
Management. Diabetes Care 10.2337/dc06-2101.
30
Key Lessons (1)

• All forms of ITDM can improve the lives of
  diabetics
• All forms of ITDM improves care processes
• Improved care process results in improved quality
  of life
• Technologies used by providers have the greatest
  potential for benefit
• Diabetes registries showed the greatest
  improvements in clinical outcomes
• CDSS showed the next greatest improvements in
  clinical outcome
• Diabetic registries are cost beneficial over ten
  years
• National adoption of registries saves money
• Registries are cost beneficial for all size
  organization except single physician practices

31
Key Lessons (2)

• For other technologies, national adoption costs
  more than it saves
• National adoption of all other technologies cost
  money in net
• CDSS achieves positive cost-benefit for practices
  with more than six physicians
• Economies of scale vary widely
• CDSS show strong economies of scale, due to the
  high cost of knowledge management
• Payer technologies show strong economies of
  scale, due to negotiating leverage of large
  organizations
• Patient technologies show weak economies of
  scale, due to high patient associated costs (e.g.
  individual devices)

32
Implications (1)

• Market inefficiencies may foster suboptimal
  solutions
• Payers reap most cost savings
• Many diabetes-management programs are implemented
  by payers
• Our research suggests provider technologies may
  be more effective overall
• Research points to the benefit for public
  clinical knowledge repositories
• Most of CDSS associated costs stem from knowledge
  management (KM)
• Public knowledge repositories may allow small
  practices to benefit by defraying the large fixed
  costs of KM

33
Implications (2)

• Implementation cost control and careful targeting
  of interventions be important
• All technologies provide benefit, but in only
  some cases do the benefits exceed the costs
• Cost containment should not be overlooked when
  looking for high benefit solutions
• Careful targeting, through severity
  stratification or predictive modeling, may play
  an important role

34
Limitations

• Strength of evidence
• Future studies may show more or less potential of
  ITDM to improve care
• Our methodology selected a study that showed a
  negative effect of CDSS on blood pressure
• Only one study showing the effect of foot
  screening on amputation was identified
• Severity stratification
• Severity stratification and predictive modeling,
  techniques often used to increase the efficiency
  of disease management, were not included in the
  analysis
• Scope of benefits
• ADA estimate approximately 2/3 of diabetic costs
  due to utilization that was not modeled (general
  medical utilization and indirect economic costs)
• Cross-applicability of studies
• Diabetes-management programs vary widely in
  salient features (population under study,
  programmatic components)

35
Thank You!

• More information www.citl.org
• Blackford Middleton, MD
• bmiddleton1_at_partners.org