Payment innovations in healthcare and how they affect hospitals and physicians Christian Wernz, Ph.D. Assistant Professor Dept. Industrial and Systems Engineering Virginia Tech Abridged version of the presentation given at: Work-In-Progress Session Office for Clinical Practice Innovation, George Washington University December 19, 2014 1
The U.S. healthcare system is expensive Per capita health care spending, 2011 USD at purchasing power parity (PPP) United States R 2 =0.72 29% Portugal Spain France Germany Canada Denmark Austria UK Finland Iceland Australia 71% Poland Czech Republic South Korea Data Source: OECD Health Statistics (2013) Per capita GDP, 2011 2
Despite high expenditures, overall quality is low Ranking 11 10 9 8 7 6 5 4 3 2 1 Overall quality US CAN FRA NZ NOR GER NETH AUS SWE SWIZ UK Technology availability Quality care Access Efficiency Equity Healthy lives Source: The Commonwealth Fund (2014), SelectUSA.commerce.gov (2014) 3
U.S. is No.1 in technology availability Ranking 11 10 9 8 7 6 5 4 3 2 1 Overall quality US CAN FRA NZ NOR GER NETH AUS SWE SWIZ UK Technology availability Quality care Access Efficiency Equity Healthy lives Source: The Commonwealth Fund (2014), SelectUSA.commerce.gov (2014) 4
172.1 129.3 129.3 Treatment of U.S. patients is technology intensive and expensive Estimated spending on computed tomography (CT) procedures, 2012 CT scans per 1000 population 256.8 United States Quantity x Price France South Korea Canada $566 $183 $125 $124 Reimbursement price per procedure Source: OECD ilibrary (2014), the Commonwealth Fund (2013) 5
Hospitals rank imaging as top spending priority Year 2011 Currently under construction Planned in the next three years Imaging 17% 15% Emergency department 12% 18% Surgery 10% 15% Ambulatory care 7% 14% Cancer center 10% 11% Interventional suite 9% 11% Laboratory 9% 11% Cardiology 8% 10% Source: HFM/ASHE construction survey (2011) 6
Hospitals are major cost producers U.S. healthcare spending breakdown, 2010 Direct administrative costs 13% Dental services 3% Retail products/services 4% Hospitals 25% Long-term care 7% Other services 8% Physician and clinical services 16% Source: Deloitte (2012) Prescription drugs 8% Supervisory care 15% 7
Expensive, low quality USD at purchasing power parity (PPP) United States + Technology intensive United States + Hospital based U.S. health care spending breakdown, 2010 R 2 =0.72 29% Poland Portugal Spain South Korea Czech Republic France Germany Canada UK Iceland Finland Denmark Austria Australia 71% 256.8 Quantity x Price 172.1 France 129.3 South Korea 129.3 Canada $566 $183 $125 $124 Opportunity: Reduce cost and increase quality for technology usage and investments in hospitals 8
How? 9
Positively shaped health care incentives increases both efficiency and quality of care. The Center for Medicare & Medicaid Services (CMS) is working on payment innovations and is piloting a variety of reimbursement methods that pay for quality, and not quantity. President's Council of Advisors on Science and Technology (PCAST), May 2014: The predominant fee-for-service payment system is the primary barrier to great use of systems methods and tools in health care, as it serves as a major disincentive to more efficient care 10
U.S. healthcare is complex, and a better understanding of the system is needed Environment Regulation, policy, market Regulators Medicare, Medicaid Insurance Companies Research Funders Health Care Purchaser Organization Infrastructure, resource Hospitals Outpatient Clinics Nursing homes Rehabilitation Centers Care Team Frontline care providers Patient Physicians Nurses Patients Family Members 11
I will present two research projects D D O Medicare O O 6 7 D O Hospital 1. Designing multi-level incentives for healthcare systems O O D,O Physician D O O D 4 5 D D O 1 2 3 O Patient D D 2. Helping hospitals make better technology investment decisions. 12
How to get from reality to math? 1 Incentive from CMS Hospital Buy new CT scanner CT av Hig Incentive 2 Status quo Low O Max a C C * 2 0 1 R * H R1 * R2 * ( 2 h ) ( 1 ) ( 2 )... 3 Physicians D D CT scan Alternative Medicare High Low 13
1. The agent interdependence diagram captures decisions, outcomes, interactions 14
2. The detailed graphical representation allows for micro-modeling Incentive from CMS Hospital Buy new CT scanner CT availability High Costs High Status quo Low Low Incentive O O D D CT scan Medicare billings High Patient health High Physicians Alternative Low Low Stakeholders Decisions Uncertainty Outcomes 15
3. The mathematic formulation is based on Multiscale Decision Theory (MSDT) SUP SUP a1 SUP a2 p s s, a Influence on transition probability SUP SUP INF SUP final i j m SUP s1 SUP s2 INF s1 INF s2 p s a INF INF INF j n Influence on reward INF a1 INF a2 INF 0.3 0.25,, p s s a p s a f s s a SUP SUP INF SUP SUP SUP SUP SUP INF SUP final i j m i m i j m, r s s r s b r s INF INF SUP INF INF SUP SUP final j i j i E r a, a r s p s a p s s SUP SUP INF SUP SUP INF INF INF SUP SUP INF final m n final i j n final i j i j,, E r a a r s s p s a p s INF SUP INF INF INF SUP INF INF INF SUP SUP final m n final j i j n final i i j Area 2a Area 2b r s r s SUP SUP SUP SUP final i i Area 5b share coefficient b 0.2 0.15 Area 1 Area 3 0.1 0.05 Area 4 Area 5a 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 change coefficient c 16
Payer Savings Different inventive programs exist High Bundle payment Shared savings Medical home Pay for Performance (P4P) Low Fee for Service (FFS) Low Provider Financial Risk High 17
We modeled the Medicare Shared Saving Program (MSSP) Hospital and physicians form an Accountable Care Organization (ACO) to coordinate care of Medicare patients MSSP Standard Reimbursement (as before): Physicians/radiologists: Fee for service Hospital: Diagnosis-related groups for inpatients Outpatient prospective payment system MSSP Incentive: ACO receives 50%-60% of cost savings when achieving quality goals 18
Physicians decision and outcomes related to CT scans CT scan Medicare billings High Patient health High Physicians Alternative Low Low Stakeholders Decisions Uncertainty Outcomes Agent P P 0a 1 Pr P s P P b,c a p ( ) = a b,c;p { } P S P = s P 1,1, s P P b,c,,s B,C 19
Hospital and physicians affect each other Operating & Maintenance cost Hospital reputation payoff Incentive for H Bundled incentive from CMS Investment decision Buy advanced CT scanner Maintenance costs High Operating costs High Hospital reputation High Hospital Incentive for P Status quo Low Interdependencies Low Low Incentive passed on to physicians CT scan Medicare billings High Patient health High Physicians Alternative Low Low Stakeholders Decisions Uncertainty Outcomes 20
Converting the graph into math Hospital Payoff before incentive for H: Reward Probability Final payoff for H: H H H, P CMS H H, P H P H, P final total h h h E r E r a a g a a g a a Payoff before incentive Bundled incentive Distributed incentive 21
Converting the graph into math Physician Payoff before incentive for P: Reward Probability Final payoff for P: P P H, P H P H, P final total h h E r E r a a g a a Payoff before incentive Distributed incentive 22
Incentives from CMS to ACO Bundled incentive from CMS to Hospital: E ég CMS H ë ( a H h,a P ) ù û = 50% é ë M - E(r H cost a H h,a P ) - E(r P cost a H h,a P ) ù û Benchmark set by CMS Distributed incentive from Hospital to Physician:,, h h g a a m g a a H P H P CMS H H P Sharing percentage set by Hospital 23
Decision problem is a sequential game 24
Results: Incentives can prevent new equipment purchase and lower CT scan rate CT scan rate M=35 m=0.5 a 50-50 split of CMS incentives between H and P 25
Physicians optimal CT scan rate depends on incentive and equipment Physician payoff P s P'sreward payoff 25 With incentive, investment No incentive, investment 20 With incentive, status quo 15 No incentive, status quo 10 0.2 0.4 0.6 0.8 1.0 CT scan rate Scan Test rate 26
How should hospitals and physicians split the incentive? Optimal sharing percentage m* m 1.0 all goes to Physician 0.8 0.6 0.4 0.2 0.0 26 28 30 32 34 Hospital keeps all Benchmark M 27
In summary Physicians - Incentives motivate physicians to reduce CT scan rate. Hospitals - Incentives can reduce hospital s propensity to invest in additional equipment. - Given challenging cost benchmarks, hospital passes on all incentives to physicians, which results in largest CT scan reduction. 28
In summary ACOs - The incentive distribution mechanism can be designed to maximize the payoffs of their members. Policy maker / CMS - The cost benchmark has to be set just right to induce desired behavior, or otherwise the incentive is not effective. 29
Next steps: YOU can help 1. Revise model and remove assumptions to get closer to clinical practice / reality 2. Calibrate and validate model through data 30
Project 2: Helping hospitals make better technology investment decision D D O Medicare O O 6 7 D O Hospital Board of Directors D,O Physician D O O D O O 4 5 D D O 1 2 3 O Patient D D Departments Physicians 31
Large number of capital requests compete for small budgets < 32
Current investment decision-making is informal and unstructured Pressure from physicians, patients and donors Limited information, high uncertainty Too many variables to consider Ad-hoc, heuristic, political decisions Multiple objectives, some hard to quantify 33
Decision analysis can make this process more transparent and systematic All objectives are taken into account Analysis of different trade-offs Objective and balanced decision making Getting stakeholders involved 34
Project objectives: Provide hospital executives with a structured decision-making framework Apply SMART (Simple Multi-Attribute Rating Technique) in a hospital setting using real investment alternatives Prepare for SMART session through student mock panels 35
We identified investment objectives reviewing literature and best practices Objectives Attributes Financials Quality Strategic importance Infrastructure Ease of implementation Net present value Quality Adjusted Life Years Growth in market share Productivity increase Increase on patient satisfaction Low level of disruption, high usability, short learning curve 36
SMART in action: Session with hospital executives Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis Alternatives: 1. CT scanner dose reduction software, $192K 2. New CT scanner, $732K 3. CT scanner lease buyout, $292K 4. New Mammography unit, $468K 5. Mammography refurb, $160K 6. Da Vinci surgical robot, $2,000K Budget: 2.5 million $ 37
Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 1. Investment alternatives were scored across five objectives Financial impact Clinical impact Market share Routine infrastructure Staff-physician relationships Linear relationship between NPV values Direct Assessment Value Scale 100 90 40 15 0 38
Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 2. We derived weights for each objective Financial impact Clinical impact Market share Routine infrastructure Staff-physician relationships Points 10 10 10 10 10 Weights (%) 20 20 20 20 20 39
Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 3. We calculated weighted values for each alternative on each objective 40
4. The alternatives were ranked based on Value/$ Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 41
Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 5. Hospital executives chose a portfolio of alternatives that meets the budget 42
Scoring of alternatives Assessment of Weighted Single Dimensional Values Selection of investment alternatives Sensitivity Analysis 6. We performed a sensitivity analysis around the objectives weights 43
Participants confirmed feasibility and value of SMART for hospital s decision process Survey scores Intuitive Can be incorporated into hospital s decisionmaking practice Feedback from executives this is the best thing I ve done all week I need more and better information on proposed investment alternatives 44
In summary We implemented SMART in a hospital using actual investment alternatives Hospital executives found the method intuitive They believe it can be incorporated into their organization s practice Participants realized information availability and accuracy are critical 45
Thank you! www.wernz.com 46