Transforming Health Care:A State Purchaser’s Perspective

Leah Hole-Curry, JD

Washington State Health Care Authority, Health Technology Assessment

,May 21, 2010

Conflicts and Overview of Agency

CONFLICT DISCLOSURE: Leah Hole-Curry -WA HCA– Employed by state government– No funding or other resources conflict – no unlabelled/ unapproved use disclosure

Cabinet level agency of approximately 275 employees Administers 2 health care programs:

– Public Employees Benefit Board (State employees & retirees) Self-Insured PPOs (Uniform Medical Plan, Aetna) Fully Insured Plans/MCOs (Group Health, Kaiser)

– Basic Health (Income eligible, state-funded program) Ancillary programs

– Community Health Services (CHS) – Grant Awards– Prescription Drug Program (PDP)– Health Technology Assessment (HTA)– Washington Wellness (state employees & retirees)

3

The State Budget and Health Care

State budget faces a $2.8 billion shortfall The state will spend nearly $7 billion to provide medical coverage to millions of

Washington residents in the 2009-11 budget period … about 33% of the budget.

Medical Assistance Programs,

$3,582,184 ,000

Basic Health $337,757,000

State/Higher Education Employees,

$1,040,600,000

Child Care Worker Health Benefits, $8,700,000

Home Care Worker Health Benefits, $101,505 ,000

K-12 Employees, $1,588,705 ,000

Juvenile/Adult Corrections (inmate health care), $281,686 ,000 Community Clinics, $25,068 ,000

*ProjectedSource: Mercer’s National Survey of Employer-Sponsored Health Plans; Bureau of Labor Statistics, Consumer Price Index, U.S. City Average of Annual Inflation (April to April) 1990-2009; Bureau of Labor Statistics, Seasonally Adjusted Data from the Current Employment Statistics Survey (April to April) 1990-2009.

Total health benefit cost per employee rises 5.5% in 2009, the lowest annual increase in a decade.

17.1%

10.1%

8.0%

-1.1%

2.5%

7.3%

14.7%

10.1%

7.5%

2.1%

6.1% 6.1%5.6%*

5.5%6.3%6.1%

11.2%

8.1%

6.1%

0.2%

12.1%

-2.0%

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Workers' earningsAnnual change in total health benefit cost per employee

Overall inflation

5

How a Large Purchaser Can Impact the Market

Must change the delivery system to impact cost and quality– Driving change through purchasing

Must target manageable changes for the long haul (lesson learned from 1993) Governor targeted key initiatives early and stuck with them

– Emphasis on evidence-based health care– Promote prevention and healthy lifestyles– Better manage chronic care– More transparency in health care system– Better use of health information technology

Focus has endured despite a bad economy and politics pressure This focus has helped other employers, health plans and provider groups to think

differently

7

Dartmouth Atlas Identified Issues:Unwarranted Variation

There are structural problems in the current delivery system– Underuse of Effective care and Quality Variation

Align pay incentives; Chronic care management focus; Facilitate Patient information exchange; Wellness Promotion

– Overuse of Supply Sensitive Care Manage Capacity, Reward quality not volume, Promote Conservative Practice

Patterns– Misuse of Preference Sensitive (Discretionary Treatments)

Outcomes Research and Shared Decision Making The Atlas points to three strategies

– Improving the scientific basis of care delivery– Promoting the growth of organized, accountable care– Shared savings programs- to reduce overuse and improve coordination

Wennberg, et al. http://www.dartmouthatlas.org/atlases/Unwarranted_Variations.pdf

8

Variability:Back Surgery per 1,000 Medicare Enrollees (2002-03)

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

Spokane 6.2Tacoma 5.7Olympia 5.3Yakima 4.5Seattle 4.2Everett 4.1Port Angeles 2.8

1

10

100

1,000

10,000

100,000

1,000,000

Health Care Quality Defects Occur at Alarming Rates

U.S Airline flight fatalities/U.S. Industry Best of Class

Airline baggage handling

Breast cancerScreening (WA)

Detection &treatment ofdepression

Adverse drugevents

Hospital acquired infections

Hospitalized patientsinjured through negligence

1(69%)

2(31%)

3(7%)

4(.6%)

5(.002%)

6(.00003%)

Overall Health Care Quality in U.S.

(Rand Study 2003)

IRS Phone-in Tax Advice

U.S. birth defects

Recommendedwell-child visits (WA)

Treatment ofBronchitis (WA)

NBA Free-throws

Sources: modified from C. Buck, GE; Dr. Sam Nussbaum, WellPoint; Premera 2004 Quality Score Card; March of Dimes

level (% Defects)

Def

ects

per

mil

lio

n

∑

10

Evidence Based Health Care

Cutting edge programs that have become part of our offerings:– Prescription Drug Program

Preferred drug list used by PEBB, Medicaid, and workers compensation programs

– Drug Purchasing Consortium Pooling of state and private purchasing power, used by PEBB & workers compensation

– Health Technology Assessment State pays for procedures and medicine that show evidence of efficacy, cost-effectiveness,

and safety Estimated savings of $27 million since 2007

– Patient decision aid pilot Focus on high-variation, preference-sensitive areas that involve multiple options and

tradeoffs, e.g. cardiac disease; breast & prostate cancer

– Advanced imaging management Using evidence based guidelines, identify highest cost/utilization advanced diagnostic

imaging services for state programs

11

Why Health Technology?

Part of an overall strategy

Medical technology is a primary driver of cost– The development and diffusion of medical technology are primary

factors in explaining the persistent difference between health spending and overall economic growth.

– Some health experts arguing that new medical technology may account for about one-half or more of real long-term spending growth.  Kaiser Family Foundation, March 2007: How Changes in Medical Technology Affect Health Care Costs 

Medical Technology has quality gaps– Medical technology diffusing without evidence of improving quality Highly

correlated with misues, overutilization, underutilization. Cathy Schoen, Karen Davis, Sabrina K.H. How, and Stephen C. Schoenbaum, “U.S. Health System

Performance: A National Scorecard,” Health Affairs, Web Exclusive (September 20, 2006): w459

12

1. HCA Administrator Selects TechnologyNominate, Review, Public Input, Prioritize

2. Vendor Produce Technology Assessment ReportKey Questions and Work Plan, Draft, Comments, Finalize

3. Clinical Committee makes Coverage DeterminationReview report, Public hearing

4. Agencies Implement DecisionImplements within current process unless statutory conflict

Meet Quarterly

2-8 Months

Semi-annual

HTA Program Elements

2. Evidence Report Key Questions and Work Plan, Draft, Comments, Finalize

Fifteen reports to date• Comprehensive, unbiased, peer-reviewed technology assessment

reports that summarize and rate the available clinical literature. • Highlight that many technologies have widespread use and a lot of

evidence, but the data is unreliable, low quality, or absent on health benefit and value.

• Out of the hundreds of thousands or articles, more than 6,000 potentially relevant articles were identified, and 1073 were reviewed, and 383 articles were critically appraised:

Av Days Av MonthPublic Comment 83 2.8Report 155 5.2

Overall 435 14.5

14

KEY HTA Products

Transparency: Publish topics, criteria, reports, open meeting

Technology Assessment Report: Formal, systematic process to review appropriate healthcare technologies.

Independent Coverage decision: Committee of practicing clinicians make decisions that are scientifically based, transparent, and consistent across state health care purchasing agencies.

Key focus questions:• Is it safe?• Is it effective?• Does it provide value (improve health

outcomes)?

Pay for What Works: Better Information is Better health

15

Technologies selected– 17 technologies selected since 2007

3 first year; 5 second year; 8 third year

Analysis completed– Over 6,000 articles/trials reviewed– 15 comprehensive technology assessment reports

Coverage Decisions– 9 public meetings and 13 decisions, where reliable evidence:

7 show benefit and support coverage for certain situations 5 do not yet show benefit and are not covered 1 shown unsafe or ineffective

– Estimated $27 million cost avoided

– Projected Utilization impact: 3 increased; 3 same; 7 decrease

HTA Outcomes

Health Technology Assessment Program Outcome

Utilization

Topic Date Safe EffectiveCost-

EffectiveHealth Benefit Coverage

Impact (annual figure)

Upright MRI May-07 Equal Insufficient Less N/A No $2,990,000

Ped Bariatric Surgery <18 Aug-07 Insufficient More Insufficient No No $0Yes/

ConditionsLumbar Fusion Nov-07 Less Equal/More Less Yes Yes/

Conditions $5,240,639

Discography Feb-08 Insufficient Insufficient Insufficient No No $324,000

Virtual Colonoscopy (CTC) Feb-08 Equal Equal/More Less No No $11,100,000

Intrathecal Pump for chronic noncancer painFeb-08 Insufficient Insufficient Equal No No $691,326

Arthroscopic Knee Surgery Aug-08 Less Equal Less No No $400,000

Artificial Disc Replacement Nov-08 Equal Equal/More Insufficient YesYes/

Conditions0*

Computed Tomographic Angiography (cardiac)

Nov-08 Equal Equal Equal/More YesYes/

Conditions$5,063,928

Cardiac Stents May-09 Equal Equal/More Less YesYes/

Conditions$966,760

$27,366,138

*Insufficient current data to calculate conservative estimate.

Less $589,485

Health Technology Assessment ProgramEvidence Decision

More Insufficient YesPed Bariatric Surgery 18-21 Aug-07

17

ConsumerReports.org 10 overused tests and treatments November 2007

1 BACK SURGERY. … surgery, which can cost $20,000 plus physician's fees …..

2 HEARTBURN SURGERY. operation, costs $14,600 or more

3 PROSTATE TREATMENTS. . over treated with surgery that costs $17,000, or by radiation therapy for $20,700

4 IMPLANTED DEFIBRILLATORS. … cost some $90,000 over a lifetime.

5 CORONARY STENTS. Billions are spent each year….

6 CESAREAN SECTIONS. ..cost almost $7,000, about 55 percent more than natural delivery...

7 WHOLE-BODY SCREENS. CT scans, which can cost $1,000 … no proven benefits for healthy people. A few CT scans a year can increase your lifetime risk of cancer.

8 HIGH-TECH ANGIOGRAPHY. Using a CT …costs an average of $450...standard angiography is sometimes still needed.

9 HIGH-TECH MAMMOGRAPHY. Using software to flag suspicious breast X-rays would add $550 million a year to national costs if used for all mammograms. But a 2007 study found that this technique failed to improve the cancer-detection rate significantly, yet resulted in more needless biopsies.

10 VIRTUAL COLONOSCOPY. …Though less costly than a standard colonoscopy, the virtual test isn't cost-effective because any suspicious finding requires retesting with the real thing.

Copyright © http://www.consumerreports.org/cro/health-fitness/index.htm 2000-2006 Consumers Union of U.S., Inc.

18

New Health Purchasing Focus: Hierarchy of Evidence

Best: Meta-analysis of large randomized head-to-head trials.

Large, well-designed head-to head randomized controlled clinical trials (RCT):

Long-term studies, real clinical endpoints

Well accepted intermediates

Poorly accepted intermediates

Smaller RCTs, or separate, placebo-controlled trials

Well-designed observational studies, e.g., cohort studies, case-control studies

Safety data without efficacy studies

Case series, anecdotes

Least: Expert opinion, non-evidence-based expert panel reports, and other documents with no direct clinical evidence

19

Level 3: “What would I recommend to the state or nation?”– Must be based on rigorous assessment of the scientific

evidence. – Affects hundreds of thousands, even millions of people.

Level 2: “What would I recommend to my patient/client?”– Influenced by prior experience, but the scientific evidence may

play a greater role. – Affects possibly hundreds of people.

Level 1: “Would you have this done for yourself or for someone else in your immediate family?”– Influenced by one’s personal experience with the disease and

capacity to deal with risk.– Affects few people.

Used with Permission from Dr. Mark Helfand, OHSU

Evidence in Health Care Decision Making

20

Evidence for use in Policy Decisions

Different Data Sources Efficacy

– How technology functions in “best environments” Randomized trials-distinguish technology from other variables Meta-analysis

Effectiveness– How technology functions in “real world”

Population level analyses Large, multicenter, rigorous observational cohorts (consecutive pts/objective observers)

Safety– Variant of effectiveness

Population level analyses Case reports/series, FDA reports

Cost– Direct and modeled analysis

Administrative/billing data (charge vs cost) Context

– Mix of historic trend, utilization data, beneficiary status, expert opinion

21

Clinical Committee Decision must give greatest weight to most valid and reliable evidence– Objective Factors for evidence consideration

Nature and Source of evidence Empirical characteristics of the studies or trials upon which

evidence is based Consistency of outcomes with comparable studies

– Additional evaluation factors Recency (date of information) Relevance (applicability of the information to the key questions presented or

participating agency programs and clients) Bias (presence of conflict of interest or political considerations)

WAC 182-55-030: Committee coverage determination process

3. HTCC Decision Basis

22

3. Health TechnologyClinical Committee

Chair: Brian R. Budenholzer, MD, FAAFPVice-Chair: C. Craig Blackmore, MD, MPHMembers– Megan Morris, CPO, LPO– Lydia Bartholomew, MD, MHA, CPE, FAAFP– Louise Kaplan, RN, MN, PhD, ARNP– Michael Myint, MD, MS– Carson E. Odegard, DC, MPH– Richard C. Phillips, MD, MS, MPH– Michelle Simon, PhD, ND– Michael Souter, MB, Ch-B, DA, FRCA– Chris Standeart, MD,

A B C D

Con

fiden

ce

BenefitLarge Moderate Unknown/Inferior

High

Limited

Low

What’s the Real Problem?• We want to pay for Highly Effective Benefits!• We have to be careful not to pay for Ineffective and Unsafe Benefits!• We need a process for Everything in between!

Grades the evidence

?

24

Coronary heart disease (CHD) is an important public health concern, very prevalent where patients range from no symptoms to chest pain (angina), to heart attack- myocardial infarction (MI), or death.

Prediction of risk and symptoms is difficult. Treatment includes:

– Manage and reduce risk– Medication therapy– Surgical treatment by mechanically opening the artery

Use of PCI has steadily risen over past decade while bypass remains relatively unchanged and PCI accounts for over 60% of surgical treatment.

Unanswered questions remain about best use of each option, when, and for what patients

Topic Background:Disease/Diagnosis

25

Cardiac stents are small tubes placed in an artery to keep it open. Stents are either not coated (bare metal stents) or coated with a drug (drug eluting stents)

Stent advantages include physically opening the artery and being less invasive than bypass surgery

Stent disadvantages include targeted solution to widespread disease, unclear protocols, clotting and re-operation

Important, unanswered questions remain about whether and when stent placement is appropriate versus other medical management or surgery.– What patient, disease level, and timing are best for this invasive

procedure

Topic Background:Selected Topic

26

Agency Prioritization

Safety concern: High– Primary safety concerns: long term risks, procedure risks, frequency,

FDA panel findings on thrombosis for DES off label. Efficacy concern: High

– Primary concerns: efficacy of stenting to prevent death or major cardiac event and high stent diffusion with low or mixed evidence on appropriateness

– Concerns about high use variation especially 70% non-FDA approved uses in generally sicker or more complicated patients; drug eluting stent use; use instead of optimized medical therapy in lower risk patients and instead of CABG in high risk patients;

Cost Concern: Medium– Cost concerns reflect mainly concern about over or mis-utilization, and

wide cost differences between treatment choices.

WA Cardiac Stent Procedure Utilization: 2004-2007Clinical Outcomes Assessment Program (COAP)*

27

Year 2004 2005 2006 2007 Total PCI Procedures** 15,158 15,330 15,686 14,164

No Prior PCI 10,022 10,146 10,265 9,135

Repeat Procedures 5,136 5,184 5,421 5,029

% Repeat Procedures 34% 34% 35% 36%

PCI Procedures with Stents 13,348 14,104 14,542 13,032 % stented PCIs 88% 92% 93% 92%

Count of All Stents 18,860 19,931 21,048 19,688

Count of Bare Metal Stents 3,224 1,408 2,122 5,214

Count of Drug-Eluting Stents 15,636 18,523 18,926 14,474

% Bare Metal Stents 17% 7% 10% 26%

* A program of the Foundation for Healthcare Quality in WA state ** Inpatient and outpatient procedures

WA Public PurchasingCardiac Stent Procedure Utilization: 2004-2007

28

  2004 2005 2006 2007

Total Costs* $14,263,103 $15,505,519 $17,218,988 $16,544,589

Total Procedures** 988 1010 1040 954

Bare Metal*** 175 80 117 283

Drug-Eluting*** 781 919 904 650

* Inpatient, outpatient, Medicaid and Uniform Medical Plan as primary and secondary payors** Procedure codes 36.06, 36.07, 92980, 92981, G0290 and G0291 performed as primary or secondary procedure*** Excludes patients who received both types in same procedure

Average Per Procedure Costs  BMS/ DES DifferenceInpatient  

Bare Metal $22,360  Drug-Eluting $26,497 $4,137

Outpatient  Bare Metal $13,038  Drug-Eluting $17,345 $4,307

29

Key Questions

Key Questions: Stent Focus– Originally Off Label usage– Change Direction to create overview of stent use to set context and

take advantage of initial research, – Focus review on well defined and studied sub-topic: Bare metal stents

versus drug eluting stents– Removes some controversy of stent question of overall when or whether to

cover, focuses on which type

– Remains significant issue due to high utilization of drug eluting stents (local and agency data about 80%)

– Recent FDA focus on safety concerns of DES

– Agency’s cost of over $3,000 additional for DES

Future Topic– Broader questions remain on when and in whom stents are most

appropriate. May be informed by subsequent topics, reviews, or potential collaborative and other agency efforts.

Evidence Report: Primary data sources• HTAs or similar reports

2 (Hill, ECRI) did own meta-analysis of RCTs1 (KCE) used results from previous meta-analyses1 (Ontario) did meta-analysis on registry studies4 (Hill, KCE, Ontario, FinOHTA) did full economic analyses

• Meta-analyses published after HTAs1 meta-analysis in general populations included 38 RCTs, N = 18,023 (Stettler 2007 Lancet 370(9591): 937-48)

1 meta-analysis with outcomes for diabetic patients separated and length of anti-platelet therapy evaluated from 35 RCTs, N = 14,799 (Stettler 2008 BMJ 337: a1331)

HTA Report interpretation: What we know

Effectiveness There is no statistically significant difference between DES and BMS with

regard to death, cardiac death or myocardial infarction up to 4 years. DES are consistently associated with lower rates of TLR

Safety While no statistically significant differences in stent thrombosis or late stent

thrombosis were seen, analyses may be underpowered; no comparative studies for bleeding

Among diabetic patients, < 6 months of dual anti-platelet therapy was associated with a 2-fold increase in death and cardiac death with DES but there was no difference in MI regardless of therapy duration

Nonrandomized studies show mixed results for death and MI

Cost: Most extensive CEAs concluded DES were not cost-effective in general

populations; ICERs driven by DES cost, #,TLRGuidelines: Professional guidelines do not address use of DES vs. BMS

Remaining Questions

EFFECTIVENESS Are statistically significant findings also clinically significant? Are the risk

differences of public health importance? How should the relative importance of the various outcomes be weighed, over

the short-term and over the long-term? What are the specific indications for DES vs. BMS in general and special

populations? What are the indications for TLR?SAFETY Is TLR/TVR correlated with decreased rates of death, cardiac death and MI

over the long term? Why or why not? How might newer DES designs or drugs compare with BMS for various

outcomes in the short term and long term? What is the long term safety of prolonged anti-platelet use?COST Will methodologically rigorous US-based CEAs draw different conclusions from

HTA CEAs as ICERs are driven by DES cost, number of stents and TLR? How does comparison of DES vs. BMS fit within the bigger context of

comparative effectiveness with medical therapy, CABG and other treatments?

33

Cardiac Stents (HTCC Decision)

Effective? Majority voted that the comprehensive evidence reviewed shows DES and BMS effective

The committee identified four key health outcomes that impacted effectiveness; with three have high quality evidence available.

– Freedom from Cardiac Mortality: the committee concluded that data from multiple RCTs demonstrated that there is no overall or cardiac related benefit with DES compared to BMS.

– Freedom from Myocardial Infarction (MI): the committee concluded that the data from multiple RCTs demonstrated that there is no benefit from DES compared to BMS in reducing rates of MI.

– Freedom or reduction of revascularization (TVR): the committee concluded that data from multiple RCTs demonstrates a benefit of an 11% reduction in the rate of revascularization with use of DES compared to BMS.

– Quality of Life: the committee believes that quality of life is an important health outcome to demonstrate overall effect of treatment, but concluded that there was not reliable data to conclude whether DES provided a benefit over BMS. The committee discussed the previous revascularization reduction as a component of quality of life

34

Cardiac Stents (Decision)

Safety? Majority voted that the comprehensive evidence reviewed shows DES and BMS are equally safe– Morbidity related to Stent Thrombosis: The committee agreed with the evidence report

conclusions that these are rare events, where even the larger RCT’s and observational data may not be powered to detect. However, the best available meta analysis of RCT data shows difference relied heavily on the most recent meta-analysis with four year follow up: 1.4% SES; 1.7%PES and 1.2%BMS.

– Bleeding: the committee concluded that bleeding is a very serious complication. Due to dual anti-platelet therapy proscribed with DES, this complication could be higher in DES; but not enough information and registry data, though lower quality, showed equivalence with 3.4% BMS vs 3.6% DES rate.

– Stent Fracture: The committee agreed that this issue was not applicable since evidence was not obtainable on this outcome and no other reason to believe rates between the two stent types would be different.

 Value? The committee agreed that overall, DES is not cost-effective, especially considering the state’s $3,600 differential, where lower price premiums produced staggering cost per QALYs.

For certain subpopulations of high risk patients, some HTAs reported, and five committee members agreed that DES is cost-effective.

35

WA HTA Cardiac Stent Coverage Decision

Based on the deliberations of key health outcomes, the committee decided– that the current evidence on Cardiac Stents demonstrates that there is sufficient

evidence of a health benefit to cover the use of cardiac stents, but limit the use of Drug eluting stents to certain circumstances.

– The committee found that drug eluting stents were proven to be equivalent to bare metal stents in safety and efficacy overall. The committee found that drug eluting stents were proven to be more effective in one area: reducing revascularization, and were proven to cost more

Bare Metal Stents are covered without conditions. Drug eluting stents are conditionally covered for:

– Stent diameter of 3 mm or less;– Length of stent(s) of longer than 15 mm placed within a single vessel;– Patients with diabetes mellitus;– Stents placed to treat in stent restenosis; or– Treatment of left main coronary disease

36

Change is a Journey

• Lessons learned• Be Transparent • Engage the provider community• Find Common Values• Make Consistent Coverage Decisions• Make Bias Free zones

• Challenges• Resource intensive• Collaborations involve time and tradeoffs• Cultural change - new decision model (not persuasion, expert or political clout)• Often identifies information gaps