A Path to Accountable Care Organizations: How Do We Get From There to Here? Financial Considerations for Accountable Care Entity Engagement Presented by Milliman, Inc. San Francisco, CA susan.pantely@milliman.com 415-394-3756 Six Considerations Related to Financial Success 1. Set budgets/targets accurately 2. Use appropriate p gain/loss sharing reflecting capabilities of provider group 3. Timely, easy to use, reports to providers and possibly to members 4. Define how members will be assigned 5. Define scope of services 6. Quality measurement 2 1
Simple Secrets to ACO Success To be successful, an ACO must preach and practice two challenging and often-overlooked imperatives: 1. Establish actuarial cost and utilization targets appropriate for the ACO s designated d business model (Medicare, commercial, Medicaid) id) 2. Provide medical management to achieve those targets 3 Bending the Cost Curve Establish baseline per member per month (PMPM) claim cost Historical analysis of claims Establish trend, based on historical or concurrent information. Project to contract period for target PMPM. Savings are shared with providers. Losses may be shared with providers also. Baseline PMPM x Trend = Target PMPM Projected PMPM Actual PMPM = Savings PMPM Savings PMPM Payer Provider 4 2
Performance Needed for Success Reduce utilization below level achieved by plan Varies but in general a 5% to 10% reduction in PMPM is required This typically translates into a >10% reduction in utilization of key services (IP, OP hosp, imaging, specialty referrals, etc.) reason is need to cover cost for ACO infrastructure Can be a consideration for plan wanting to provide infrastructure to ACO as plan may not have infrastructure required to achieve the additional utilization reduction Minimize leakage 5 Building an ACO Budget The overall financial budget/target will be established by the payer or regulator (CMS, local health plan, state Medicaid plan, etc.). The ACO will need to convert these budgets/targets into utilization targets and validate the reasonability of the targets by performing its own actuarial analysis and building an ACO actuarial model Utilization reduction and improved efficiency will likely determine whether an ACO stays within the budget or achieves the established targets. t 6 3
Creating and Using an ACO Actuarial Model Use the designated population s historical data to build an ACO actuarial cost and utilization model consistent with the base period that the payer will use to evaluate performance Use the ACO actuarial model to compare the historical data to achievable benchmarks, appropriately adjusted for demographics and risk Categorize the utilization data by Meaningful and impactable service categories Site of service where relevant (e.g., hospital outpatient versus ambulatory surgical center) ACO and non-aco providers (leakage) 7 Evaluating Financial Feasibility Identify and prioritize potential opportunities for Reducing utilization by service category Shifting utilization to alternative lower cost sites of service Steering utilization to ACO providers (if lower cost) Monetize those opportunities and calculate the overall financial impact include the cost needed to implement management programs (hospitalists, util. management, referral management, etc) Determine if the calculated financial impact will meet the financial budgets/targets. 8 4
Priorities Focus initial management efforts on Reducing leakage to hospitals and specialists that are not part of the ACO. (This will increase volume to ACO providers and help offset revenue loss due to improved utilization management.) Inpatient utilization management. (Inpatient costs make up approximately 30% of total costs for a commercially insured population and 37% of total Medicare Part A and B spend) Secondary focus for management efforts should include high tech imaging, ER visits, specialist visits & generic/brand drug distribution 9 ACO Financial Feasibility & Actuarial Utilization Model (MAFF) MAFF does all of the following: 1. Produce a financial feasibility for developing an ACO model of healthcare delivery. Includes impact and cost of required infrastructure Utilization management (IP & OP) Case/Chronic disease management Plan management support (e-visits, e-consults, urgent care/expanded access, physician incentives) Group practice support (reporting, management support) Interactive input allows model to be matched to characteristics of specific payers, providers and geographic locations Provides a bottom line answer to financial feasibility and also what needs to be in place to achieve this Anchored by Milliman s Health Cost Guidelines benchmarks 2. Allows ongoing measurement of results 10 5
Sample ACO Actuarial Model Showing All Admissions 11 Milliman Health Cost Guidelines 2010 Commercial Loosely Managed Sample ACO Actuarial Model Showing Impactable Admissions Targeted Utilization Reduction for a Hypothetical Loosely Managed Medicare Population* Target Types of Admissions Admits/1,000 Target Reduction Admits/1,000 Ambulatory Case Sensitive Admissions 47 (14% of total) 15% 40 Preference Sensitive Admissions 40 (12% of total) 10% 36 Readmissions 40 (12% of total)** 10% 36 Other Admissions 210 0% 210 Total Medical/Surgical Admissions 337 4.5% 322 * Defined for this example as having approximately 340 admissions per 1,000 enrollees ** 12% excludes a portion of readmissions classified as ambulatory care sensitive admissions Source: This is a hypothetical example based on Milliman Health Cost Guidelines Ages 65 and Over 12 6
Leakage and Site of Service Analysis Analyze ACO versus non-aco provider utilization by DRGs Ambulatory Surgery Specialists Identify specific leakage facilities Identify leakage services outside the expertise of the ACO Analyze site of service for OP procedures (hospital OP versus ambulatory surgical centers) 13 Essential Management Services Supply Side The more challenging side of medical management but produces majority of savings Intended to reduce utilization and payment for medically unnecessary services and ensure that care is delivered in the most appropriate setting Clinical guidelines help evaluate the medical necessity of requested (or, retrospectively rendered) services. Demand side Optimize a population s health so that demand for services will be lower. Impact ambulatory care sensitive admissions, preference sensitive admissions, readmissions, and ER visits 14 7
Case Study - Using Hospital Employee Benefit Data Examples from General Medical Center Employee benefit data captures full scope of services in detail. This data is often not otherwise available to providers unless you work closely with a payer Excellent source for benchmarking 15 ACO Actuarial Model: Start w/ Demographics Member Demographics July 2006 to Sept 2008 GMC Members Standard Members Age Group Male Female Male Female To 25 16.9% 18.2% 19.5% 19.4% 25-29 1.7% 4.5% 3.3% 3.3% 30-34 2.2% 4.1% 4.1% 4.2% 35-39 3.6% 5.3% 4.6% 4.8% 40-44 3.8% 5.9% 5.0% 5.2% 45-49 4.0% 6.0% 4.6% 5.0% 50-54 4.3% 6.0% 3.9% 4.4% 55-59 3.0% 4.7% 2.2% 2.5% 60-64 1.8% 2.6% 1.4% 1.6% 65+ 08% 0.8% 06% 0.6% 05% 0.5% 05% 0.5% Total 42.0% 58.0% 49.2% 50.8% Demographic Adjustment Factors Inpatient Admits: 1.11 Inpatient Days: 1.11 Medical Claims Costs: 1.11 Pharmacy Claims Costs: 1.16 GMC is Illustrative Hospital Employee and Dependent Population Sources: Milliman Health Cost Guidelines. 16 8
Example of Ambulatory Care Sensitive Admits National Well Managed National Loosely Managed Ambulatory Care Sensitive Admissions (ACSAs) (Admits/1000 Commercial) GMC Congestive Heart Failure 1.56 2.52 3.47 Bacterial Pneumonia 1.64 2.06 2.77 COPD 0.52 0.83 1.30 Urinary Infection 0.63 0.87 1.22 Dehydration 0.21 0.35 0.78 Diabetes Long Term Complications 0.14 0.24 0.47 Adult Asthma 0.12 0.19 0.35 Hypertension 0.06 0.14 0.21 Angina 0.09 0.10 0.14 Lower Extremity Amputation 0.03 0.05 0.14 Diabetes Uncontrolled 0.02 0.04 0.05 Diabetes Short Term Complication 0.02 0.04 0.05 Total ACSAs/1000 5.04 7.44 10.97 ACSAs As Portion Of Total Non-Mat Ad 14% 15% 18% 17 GMC is Illustrative Hospital Employee and Dependent Population Example of Preference Sensitive Procedures Utilization of Potentially Avoidable Preference-sensitive Admissions Admits / 1,000 Comparison GMC National National Members Loosely Well DRG Narrative Managed Managed Spinal Fusion 496 COMBINED ANTERIOR/POSTERIOR SPINAL FUSION 0.00 0.06 0.03 497 SPINAL FUSION EXCEPT CERVICAL W CC 0.71 0.14 0.06 498 SPINAL FUSION EXCEPT CERVICAL W/O CC 0.83 0.24 0.11 519 CERVICAL SPINAL FUSION W CC 0.24 0.09 0.04 520 CERVICAL SPINAL FUSION W/O CC 0.35 0.38 0.18 Total 2.13 0.92 0.43 Joint Replacement BILATERAL OR MULTIPLE MAJOR JOINT PROCS OF 471 LOWER EXTREMITY 0.00 0.09 0.02 MAJOR JOINT REPLACEMENT OR REATTACHMENT OF 544 LOWER EXTREMITY 2.01 1.39 0.73 545 REVISION OF HIP OR KNEE REPLACEMENT 0.24 0.12 0.07 Total 2.25 1.61 0.82 18 GMC is Illustrative Hospital Employee and Dependent Population Data Sources: Source: Milliman Milliman Health Medical Cost Guidelines Management and DRG Guidelines models 9
Example of LOS Distribution Length of Stay Continuance: GMC versus Benchmark GMC National July '06 - Sep '08 Benchmark % Inpatient Admits <= 3 Days 72.9% 64.0% % Inpatient Admits Greater than 3 Days 27.1% 36.0% % Inpatient Admits Greater than 5 Days 13.8% 18.0% % Inpatient Admits Greater than 10 Days 6.0% 6.6% % Inpatient Admits Greater than 20 Days 3.0% 2.5% % Inpatient t Admits tsgreater than 30 Days 1.8% 1.2% % Inpatient Admits Greater than 60 Days 0.1% 0.3% Average Length of Stay (excl. SNF) 4.40 4.20 GMC is Illustrative Hospital Employee and Dependent Population Sources: Milliman Health Cost Guidelines. 19 Example of Examining Primary/Specialty Visit Ratios Office/Home Visits Primary Care Physician (PCP)/Specialist Care Physician (SCP) Split National Loosely Managed National Well Managed % total visits GMC Experience % total visits visits per 1,000 3,241 2,917 3727 PCP 2,062 1,857 63.6% 1,650 44.3% SCP 1,178178 1,061 36.4% 2,077 55.7% GMC is Illustrative Hospital Employee and Dependent Population Sources: Milliman Health Cost Guidelines. 20 10
Global Capitation Feasibility for GMC (simplified) Global Capitation Payment from Payer $270 PMPM Current Cost @ GMC s Fees $280 PMPM Needed Improvement for Breakeven 4% Goal for GMC $250 PMPM Needed Improvement for Goal 10.7% How to Reach Goal Utilization Service Reduction Contrib. to Total Cost Reduced Hospital Admits 9% 2.7% Reduced Imaging, Surgeries 10% 1.0% Better Rx Management 10% 15% 1.5% Subtotal Utilization 5.2% Reduce GMC/Physician Fees 5.5% Total Reduction in Cost 10.7% 21 Reporting Data warehouse and provider profiling systems need to easily produce an up-to-date blend of, Health insurance type population based reports showing utilization, cost and quality per patient per month Physician report cards Hospital based reports showing inpatient cost and resource utilization by diagnosis and attending physician Physician group based reports showing risk adjusted panel management results such as panel size, referrals, ER visits, etc per patient All these report need to have outcome targets linked to financial success. 22 11
Reporting Allowed PMPM Claim Cost Member Average Inpatient Emerg. Other Grand PMPM PCP Number Months Members Hospital Room OP Hospital Total Prof. Other Rx Total Revenue Loss Ratio 27499 4,393 220 $181.93 $21.77 $67.25 $201.15 $33.99 $90.46 $596.55 $1,006.90 59.2% 32784 2,108 106 $121.66 $26.78 $97.56 $184.93 $32.40 $82.80 $546.15 $893.40 61.1% 41301 4,121 206 $152.40 $21.78 $120.03 $226.42 $35.41 $111.79 $667.85 $1,090.14 61.3% 42929 4,910 246 $107.74 $34.54 $65.26 $209.57 $27.08 $117.80 $562.00 $995.16 56.5% 44404 7,509 376 $167.64 $26.87 $106.74 $197.49 $35.86 $90.49 $625.09 $1,119.54 55.8% 45427 4,274 214 $203.93 $28.59 $105.18 $219.41 $39.98 $133.52 $730.61 $1,061.28 68.8% 45807 3,962 198 $134.89 $20.94 $100.07 $255.50 $24.09 $100.77 $636.25 $1,098.65 57.9% 51870 7,995 400 $193.93 $21.44 $88.86 $202.55 $44.16 $100.45 $651.38 $1,080.80 60.3% 52043 7,293 365 $204.60 $27.18 $96.32 $192.06 $50.44 $96.46 $667.05 $1,142.84 58.4% 53330 5,222 261 $222.28 $18.29 $96.50 $230.09 $32.23 $97.82 $697.20 $1,077.58 64.7% 53595 3,670 184 $91.49 $18.48 $44.57 $130.37 $16.14 $65.32 $366.38 $661.47 55.4% 5735 5,846 293 $301.19 $30.04 $137.41 $338.84 $46.11 $165.93 $1,019.54 $1,247.18 81.7% 23 Reporting Annual Utilization Per 1,000 Average Emerg. Office PCP Number Admits LOS Days Room Visits 27499 260.7 60 6.0 1,551.9 215.2 2 6,185.2 32784 214.8 6.4 1,377.0 252.7 5,100.3 41301 245.6 4.4 1,076.0 255.3 6,217.3 42929 179.0 4.7 838.0 222.4 6,997.8 44404 267.8 6.7 1,783.9 216.3 6,405.3 45427 292.9 6.4 1,872.4 264.8 5,193.4 45807 184.8 5.0 930.9 265.5 5,868.6 51870 291.5 7.9 2,310.0 229.8 5,226.8 52043 330.5 6.3 2,085.1 215.4 5,265.2 53330 311.1 6.4 2,004.2 196.3 5,569.4 53595 156.0 7.0 1,092.1 257.6 3,994.1 5735 384.9 7.0 2,699.1 332.5 8,784.5 5881 247.22 39 3.9 967.8 243.7 6,223.3 3 59303 159.7 4.7 746.4 184.3 4,493.9 24 12
Reporting How do you make good value choices if you don t know the cost, efficiency and quality differences among hospitals and specialists For Example: Comparison Indices ProvType Name UnitCost Efficiency Quality Hospital A 0.95 1.00 1.20 Hospital B 1.03 1.00 0.98 Hospital C 107 1.07 100 1.00 105 1.05 Cardiologist A 0.95 1.00 1.20 Cardiologist B 1.03 1.00 0.98 Cardiologist C 1.07 1.00 1.05 25 Reports Unit Cost Comparison Standard fee schedules will allow for easier comparisons among hospitals and specialists. This will help members and referrals. Non-standard fee schedules will require more complicated and less precise unit cost comparisons among hospitals and specialists. For example: Inpatient Outpatient Total Facility Medical Surgical Mat MH/SA Avg ER Surg Rad Lab Other Avg Avg A 1.16 1.11 1.08 0.54 1.12 0.91 0.74 0.86 0.66 0.83 0.79 0.95 B 1.38 1.22 1.53 0.68 1.28 1.39 0.84 1.68 2.35 1.33 1.14 1.23 C 1.32 1.13 0.97 1.11 1.11 1.23 0.98 1.02 1.15 1.12 1.02 1.06 Note: There will still be variation within each category and contract/charge master charges will impact historical based estimates. See Milliman s RBRVS for Hospitals TM for methodology. 26 13
Reports Site of Service Variation For Colonoscopy (Proc Code = 45378) Location Professional* Facility** Total Office $390.81 $0.00 $390.81 ASC #1 $216.97 $406.54 $623.51 ASC #2 $216.97 $406.54 $623.51 Hospital A $216.97 $664.70 $881.68 Hospital B $216.97 $797.65 $1,014.62 Hospital C $216.97 $997.06 $1,214.03 Hospital D $216.97 $1,130.00 $1,346.97 Hospital E $216.97 $1,329.41 $1,546.38 *All physicians in this sample accept 100% of the base fee schedule ** Each hospital has a different negotiated rate 27 Gain/Loss Sharing Sample agreement Actual paid claims minus target calculated and reported each month Exclude 90% of claims above $25,000 for each member in exchange for $xx.xx pooling charge First payment made after 1 year Gains/Losses distributed 75% to provider and 25% to plan Only 50% of gains are paid out until 5% of prior 12-month target is accrued in account to fund future losses 28 14
Gain/Loss Sharing Alternate agreements can vary gain/loss percentages. Less risk may be more appropriate for PCP only or Multispecialty only ACOs (as opposed to integrated systems) Corridor Plan Provider Total Losses > 10% 75% 25% 100% Losses 3% 10% 25% 75% 100% Gains/losses +/ 3% 0% 100% 100% Gi Gains 3% 10% 25% 75% 100% Gains > 10% 75% 25% 100% Provider groups may need support in distributing gains to individual physicians and hospitals 29 Medicare Proposed Rules PCP only Defined as internal medicine, geriatric medicine, family practice and general practice Must be exclusive to a single ACO Primary Care Services HCPCS codes 99201-99215; 99304-99340; 99341-99350; Welcome to Medicare (G0402); annual wellness visit (G0438, G0439) Option 1: Assign based on primary care services, assigns most beneficiaries i i (especially in regions with PCP shortages) Option 2: Assign based on provider (PCP) and primary care services Option 3: Step-wise. Assigned to specialist providing primary care services if member has no PCP visits. More complex. 30 15
Medicare Proposed Rules Prospective vs. Retrospective Prospective: Assigned at beginning of year based on historical utilization. Providers know who is assigned and can develop targeted programs such as identifying high risk members and outreach programs. Providers can track and monitor experience. Opponents contend providers should do this for all members not just assigned members. Will always be some retrospective adjustments for various reasons such as a member moving. Retrospective: Assigned at end of year based on actual utilization. Year to year movement of beneficiaries significant Decision based on higher accuracy and desire to change provider behavior for all beneficiaries rather than assigned beneficiaries 31 Medicare Proposed Rules Majority vs. Plurality Plurality: More beneficiaries assigned. Most Medicare members see multiple l providers and desire was to assign as many members as possible. No minimum threshold. Majority: Stricter, less beneficiaries assigned. Provider has more responsibility. Determination by Count of Services or Total Dollars Number of Services: Total Dollars: No tie-breaker rules. Not necessarily provider seen most often. May reflect intensity/resource use. Beneficiary Notification ACOs will post information about their participation CMS to provider educational material to beneficiaries 32 16
Scope of Services 1. Capitation agreements have Division of Financial Responsibility and often remove out of area, pharmacy, transplants, etc. 2. Typically, all covered services are included as goal is to coordinate care. 3. If not, alternative target rates need to be created allowing for various exclusions. 4. Pooling of large claims provides some protection for large uncontrollable events. 33 Quality Ambulatory Preventive Visits Eye exams for Diabetics Cancer Screening Hospital AMI: Aspirin at admission & discharge; Smoking cessation Pneumonia: Smoking cessation; Oxygen assessment; Antibiotic w/in 6 hrs Outcomes Hypertension: Controlling high blood pressure AMI or Pneumonia after major surgery Patient Experience 34 17
Common Mistakes Flawed analysis: Failure to capture trend; IBNR; stoploss; fees too high Assume claims costs will be enough (ignore admin cost) Underserved population generates surge in services Uncontrollable leakage Flawed design Rich benefits Internal competition/greed takes over instead of common enemy Payer barriers to reducing utilization removed (no more UM) utilization surges Services provided to ineligibles 35 36 18