TIME & RISK : The Future of Healthcare. J. B. SILVERS, PhD

Transcription

1 TIME & RISK : The Future of Healthcare J. B. SILVERS, PhD Mannix MMO Prof of Health Finance Weatherhead School of Management Case Western Reserve University Copyright 2016 J. B. Silvers Weatherhead School of Management. Case Western Reserve University

2 Key Concepts and Skills 1. Understand the redistribution of risk that is occurring under the ACA in alternative payment models and new organizational arrangements 2. Determine the first principles that will govern the steady state allocation of responsibility to the parties in the value chain that can best manage various elements of risk. 3. Translate this changing environment into its strategic implications for providers, payers and patients.

3 Time and Risk THE CHALLENGE Finance is all about time and risk Moving money back and forth in time Reallocating risk among the parties One could add flexibility to this dual goal Creating and managing options Ability to react to/anticipate changing conditions If we do this well, value is created If we fail, opportunities are lost and financial distress ensues If we are creative, each party in the chain gains And the patient/society is better off

4 What s the difference: insurance vs. financing? Insurance moves money across population pools as adverse events occur $$ $$ Savings (HSA) $ $ Borrowing (VISA Card) Unpredictable high cost events drove insurance need $$ $$ $ Higher cost, chronic care & high deductibles drive financing need Financing moves money back and forth through time

5 But how can we design a system that does this well? We have a choice of what to insure and what to finance Basics of Insurance We insure events that are not predictable individually but are in large numbers The key is getting a large group to spread the risk Basics of financing We finance events that are known but occur over multiple periods The key is matching differences in the timing of cash needs between borrower and lender (intermediation)

6 What have we done with both of these in health? Mostly we have reallocated risk High deductible plans move risk to patients Value based payment moves risk to providers But the ACA also has massively refinanced care Premium subsidies shift funds from taxpayers to enrollees (much like employers always have done for employees) The question is how well have we done these

7 Who has the risk?...where is the pig in the python? Originally Insurers/govt. were at the tail end of the python bearing all the risk PAYER First dollar coverage + cost-based payment meant that only the payer bore any risk We have been gravitating away from this simple model (invented in Cleveland by John Mannix) for 50 years

8 Who has the risk?...where is the pig in the python...now? Some people think that the patient should have most of the risk (skin in the game) Should even poor people on Medicaid make payments into their HSA (i.e., finance their care over time) with high deductible plans? Most employers agree employees should (HDHPs) The Rand Health Insurance Experiment (HIE) confirms the impact of this.. for ambulatory care PATIENT

9 The problem is how much is too much? Too much risk to the payer or provider? PAYER Bankruptcy, Failure & Exit Too much risk to the patient? Loss of Access, deferral of care PATIENT

10 So should the pig be somewhere in the middle?

11 Alternative objectives in financial contracting 1. Stick it to weakest party...use market power to avoid bearing risk 2. Induce change in behavior to reduce risk...assure skin in the game to create incentives 3. Allocate risk re ability to manage it...judge competence of each link in the value chain 4. Create cushion to absorb residual risk...make sure there are deep pockets backing us up

12 It all depends on what kind of risk exists Very wide distribution of overall risk 18% 16% 14% Actual silver plan total cost per member per year 16% with no claims in year 12% 10% 8% Of those with claims, median = $1500 6% 4% 5% of claims over $12,000/yr 2% 0% $0 $200 $400 $600 $800 $1,000 $1,200 $1,400 $1,600 $1,800 $2,000 $2,200 $2,400 $2,600 $2,800 $3,000 $3,200 $3,400 $3,600 $3,800 $4,000 $4,200 $4,400 $4,600 $4,800 $5,000 $5,200 $5,400 $5,600 $5,800 $6,000 $6,200 $6,400 $7,500 Source: AV Calculator, CMS 2015

13 But where did these costs come from? Incidence Rate Patient Initiation Diagnosis Treatment & Service Cost Function & Productivity Outlier (>5%ile cost) Post Acute/ Home Care Acute Normal (median cost) Post Acute/ Home Care start >$0 Chronic Focused Factory (very low cost) Normal (median cost) Maintenance (low cost) Home Care/ Telemedicine Telemedicine/ Retail Telemedicine/ Retail $0 Routine Normal (median cost) Focused Factory (very low cost) Telemedicine/ Retail Telemedicine/ Retail

14 Compounding elements of financial risk DRIVER: Incidence: -population -disease Pt.Demand: -net price -knowledge RISK MGT TOOL: Prevention: Cost Sharing: -wellness -copay/deduct -public hlth. -HSAs Diagnosis: -knowledge -technology Audit: -peer review -second opinion Treatment: Service Cost: -doc educ. -scale/scope -incentives -organization -org. structure -mgt. ability Grouping: -DRG -bundled/vbp Fixed Pricing: -MFS/PPS -incentives Functionality: -rehab serv -pat. attitude Penalties: -readmission -incentives

15 How do we reallocate risk: mechanisms Fundamental building block is a contingent claim: Option that pays in one state but not in another Incentive payment Option/insurance/limitation Resultant impact on me Exclusion range Underlying outcome (cost) Hold harmless/maximum limit Full incentive (or partial)

16 Examples of financial contracts in healthcare Inpatient DRG Payment Resultant impact on me Typical Insurance Policy Mandatory Preventive Care Out-of-Pocket Maximum Deductible Underlying outcome (cost) Resultant Loss Outlier (Stop-Loss) Payments gain loss Cost DRG Pmt

17 Compound set of contingent claims Resultant Loss Cost We can put together combinations of these financial options to construct any payoff incentive desired...that s the easy part! Resultant Loss Too little risk of loss? Too much risk of loss? Cost The tough part is predicting how the subjects of the incentives will react... and design the risk allocation accordingly

18 Examples: High Deductible Plans: Huge assumptions re patient s ability to finance and manage all forms of risk Assumes insurance and financing situation is the same for all income groups High income can finance front end, low income cannot Resultant Loss Too little risk of loss? Too much risk of loss? Cost

19 Examples: Outlier Payments for high cost cases under DRG: Payment assumes ability to manage Financial signal too small and too distant from case to gain attention Resultant Loss Too little risk of loss? Cost

20 Examples: Value-based Payments: Zero-sum assumes only question is allocating risk (rather than changing behavior) Risk adjustment is very limited Forces demographic risk that cannot be managed Metrics are myopic and limited to very imperfect electronic measures Resultant Loss Too little risk of loss? Too much risk of loss? Cost

21 Examples: Accountable Care Organizations: Huge incentives to bring outlier practices back to mean ( outliers anonymous support group?) Little ability to gain if already doing well BUT huge impetus to focus on reorganization across the boundaries (i.e., post discharge, etc.) Resultant Loss Too little risk of loss? Too much risk of loss? Cost

22 Examples: Bundled Payment (BPCI, CJR): Huge incentives to standardize paths and create focused factories Big temptation to exclude more difficult cases Resultant Loss Too little risk of loss? Too much risk of loss? Cost

23 So how can we design systems to do this well? First principles of financial design depend upon-- Origin/driver of risk (demand, diagnosis, etc.)? Size and nature of the risk? Binary (incidence rate, patient presentation, etc.) Continuous (no event, median, 95 th %-ile) Tools available to manage risk Education, information, feedback, incentives Increase ability to manage (QI, 6 Sigma, etc.) Cushion required to avoid distress Likelihood of adverse events compounding Cash balances, contingency plans and sources, exit

24 We need to understand the many facets of risk Actuarial Need to Spread Risk There will always be residual risk to be insured But insurance is a poor way to finance known risks We need to find a way to finance known costs over time Chronic conditions and genetic knowledge reduce the random nature of insurable events and require financing Pulling these out of the risk pool is one way (a la France) to avoid the death of health insurance as we know it now Behavioral Economics Understanding of Decisions/Reactions It matters how decisions are framed (gain/loss, sequence) & anchored (recent experience, current endowment) Organizational Approach to Threats/Opportunities re Risk Team versus individual Near term versus distant

25 Where will we be in steady state (if that s possible)? Silvers Rules of Equilibrium: Risk is best handled at the level of the system and by the responsible party with the most knowledge regarding and ability to manage that particular element of risk. Don t ask docs to be insurance companies or insurers to make medical decisions Risks can be either objective (actuarial) or subjective (behavioral) and must be managed accordingly. It s not about just spreading risk around but about encouraging appropriate behavior Systems will evolve towards an economically appropriate distribution of responsibility and risk in a competitive market based on value maximization. Whoever does the best job of sharing risk will create the most value for payers, providers, patients and society

26 So where will we wind up? Less risk to patients and more information to help with decisions More risk to systems (not to components) Physician ACOs, partnered bundled payment have initial advantage But integrated systems with good physician management/ information/incentives should win Some too large to be agile enough to manage risk (Coase s Law) More government and system assumption of financing where insurance can t work Financing high deductibles, subsidizing cost sharing But managers will have to be very creative in distributing risk within organizations to get results...not good at it now

27 Thank you!