Overview Primer on Medicare bundled payment Early experiences and - PDF document

Medicare Bundled Payment: Risks, Opportunities & Future Implications Robert Mechanic, MBA The Estes Park Institute January 27, 2015 Overview • Primer on Medicare bundled payment • Early experiences and emerging trends • Opportunities for improved efficiency • Managing financial risk • Selecting bundles • Perspectives and discussion Brandeis University 2 1

A Brief History of Bundled Payment 1983 1993 2007 2009 2010 2011 2014 Medicare Medicare Bundled Pmt. Heart Bypass Acute Care for Care Demo Episode Demo Improvement Medicare Private Efforts: Affordable Inpatient ProvenCare Care DRGs Prometheus Act 3 Will Bundled Payment Become the Next DRG System? Brandeis University 4 2

Bundled Payment Payer Single payment to cover costs of episode of care $$$ (30, 60, 90 days) Shared Group is Accountability responsible for Hospital or Integrated all care within Network the episode $ $ $ $ $ Prospective Retrospective • Target budget for each episode • All providers paid FFS $$$ • Periodic CMS settlements – Distribute surplus – Reclaim deficit • Health system decides Hospital or Integrated Network – Whom to contract with $ $ $ $ $ – How to distribute bonuses Brandeis University 6 3

Opportunities to Improve Margins Primary • Reduce supply costs (e.g. implants) • Reduce unnecessary consults & testing • Reduce errors and complications • Reduce post ‐ acute care costs Conditional (dependent on backfill) • Reduce readmissions • Reduce length of stay Brandeis University 7 CMS Innovation Center: Bundled Payment Pilot Brandeis University 8 4

CMMI Bundled Payment Pilot Model 1 Outpatient Professional Professional Inpatient Professional services Inpatient SNF Readmission Index Hospitalization Stays 90 day look ‐ forward Brandeis University CMMI Bundled Payment Pilot Model 2 Outpatient Professional Professional Inpatient Professional services Inpatient SNF Readmission Index Hospitalization Stays 30 ‐ 90 day look ‐ forward Brandeis University 5

CMMI Bundled Payment Pilot Model 3 Outpatient Professional Professional Inpatient Professional services Inpatient SNF Readmission Index Hospitalization Stays 30 day look ‐ forward Brandeis University CMMI Bundled Payment Pilot Model 4: Prospective Payment Outpatient Professional Professional Inpatient Professional services Inpatient SNF Readmission Index Hospitalization Stays 30 day look ‐ forward Brandeis University 6

Choices for Bundlers • Select bundles – 48 bundles (encompassing DRG families) – Covering about 70% of Medicare payments – Must take all patients in selected bundles • Exclusions (not optional) – Readmissions and Part B services – Revised and improved but ….. (328 for 469) • Select bundle length (30, 60, 90) • Select risk track Brandeis University 13 Gainsharing • CMS gets first 2% of savings (90 ‐ day episodes) • Additional savings can be shared between facility, physicians, post ‐ acute providers – Physician gain share capped at 50% of Medicare – No cap on other providers • Useful principals – Quality gates to ensure clinical performance – Distribute based on contribution and performance Brandeis University 14 7

Financial Model 2009 ‐ 12 2014 Historical Update Factor * Target Price Cost Per CMS Discount Episode $18,200 * For illustration $18,200 update = 1%/yr Settlement discount = 2% $900 Actual Episode Definitions FFS Cost Winsorization $17,300 15 BPCI Experience Year 1 Brandeis University 16 8

Initial BPCI Enrollment Was Modest Number of Organizations at Risk in 2014 120 100 80 60 111 89 40 20 15 10 0 Model 1 Model 2 Model 3 Model 4 Brandeis University 17 2014 Model 2 Enrollment Was Limited Hospitals by Number of Episodes Selected 60 50 40 30 56 20 10 20 16 11 9 0 1 2 3 ‐ 8 11 ‐ 18 24 ‐ 48 Number of Episodes Selected in 2014 Brandeis University 18 Source: BPCI participant file (July 24, 2014) downloaded from CMMI website. 9

BPCI Experience in 2014 • Program start delayed from Jan 13 to Jan 14. • New round announced ‐ LOI due April 2014 • Q1 reconciliation files distributed in October with many of surprises • Large bundler push back • CMS announces no downside risk for 2014 Brandeis University 19 BPCI Trend Factors: 2012 ‐ 2014 (Q2) Source: CMS M2 common template file for BPCI reconciliation 2014 Q2. Brandeis University 20 10

Applications for 2015 Model 2 Model 3 Institutions 1,054 3,280 MD Groups 957 645 Total 2,011 3,925 Source: BPCI participant file (July 24, 2014) downloaded from CMMI website. • Must commit to “risk” for at least one episode by April 13 • May add episodes up to October 1, 2015 Brandeis University 21 Model 2: Conveners for Hospitals Source: BPCI participant file (July 24, 2014) downloaded from CMMI website. Brandeis University 22 11

Model 2: Conveners for Physicians Source: BPCI participant file (July 24, 2014) downloaded from CMMI website. Brandeis University 23 The Financial Opportunity of Bundled Payment Medicare Spends a Tremendous Amount in the 30 – 90 Days After Patients Are Discharged from the Hospital 24 12

Medicare Post Acute Care Spending 2012 Medicare Spending by Type 2008 Medicare Spending for Hospitalization plus 30 Days 21% 34% Hospital IP Professional Hospital IP Professional Post ‐ Acute Post ‐ Acute Hospital OP Source: RTI Inc, Post ‐ Acute Care Episodes: Expanded Source: MedPAC, 2014 Data Book (Charts 1 ‐ 1, 8 ‐ 2). Analytic File, June 2011 p.216. Brandeis University 25 Average 2013 Medicare Inpatient Payments Select Episodes for Sample Hospital $40,000 $35,000 $30,000 $25,000 $20,000 $15,000 $10,000 $5,000 $0 Major Joint CHF COPD Renal Fail. PCI Spinal Fusion Source: Brandeis University analysis of Medicare Claims. 26 13

Avg. 2013 Medicare 90 ‐ Day Episode Price for Index Stay & Post Acute: Sample Hospital $40,000 $35,000 29% $30,000 $25,000 45% $20,000 48% 70% $15,000 71% 69% $10,000 $5,000 $0 Major Joint CHF COPD Renal Fail. PCI Spinal Fusion Index Admission Post Acute Source: Brandeis University analysis of Medicare claims data. 27 Thinking About Post ‐ Acute Care Brandeis University 28 14

There is Significant Variation in Post ‐ Acute Care Spending Across Hospitals …. …. And Many Opportunities to Reduce Post ‐ Acute Care Spending 29 Average 2009 Post ‐ Acute Care Spending per Episode for Total Joint Replacement (90 day) $16,000 $12,000 $14,000 $12,000 $10,000 $6,000 $8,000 $6,000 $4,000 $2,000 $0 A B C D E F G H I J K L M N O P Q R “St. Minimus” “St. Maximus” Source: Brandeis University analysis of Medicare Claims data. Figures adjusted for hospital wage 30 index. 15

A Tale of Two Hospitals: Joint Replacement Episode 31 Source: Brandeis University analysis of Medicare Claims data. Unadjusted data. A Tale of Two Hospitals: Joint Replacement Episode 80.0% 70.0% 60.0% 50.0% St. Maximus 40.0% St. Minimus 30.0% 20.0% 10.0% 0.0% Readmission Rate Pct. SNF Pct. Home Health Source: Brandeis University analysis of Medicare Claims data. 32 16

Opportunities for St. Maximus • Expand home health and reduce use of SNF services where appropriate • Put a program in place to monitor patients following discharge – Medication reconciliation – Home assessment – Primary care visit within 7 days – Emergency plan for likely events • Consider preferred relationships with collaborative & high value facilities. 33 Picking Bundles for BPCI 34 17

Considerations for Bundle Selection • Commitment level – learning or major strategy • Stability of year to year bundle prices – Volume – Variation • Performance/spending relative to local market • Opportunity to reduce wasteful spending • Likely ability to effect clinical changes Spending for Top Clinical Episode: 2009 ‐ 12 Source: Claims data from sample hospital. 18

Episode Price vs. State Average Note: Hospital mean and state mean are winsorized and normalized to a common wage index. **: DRG Spending by Category Procedural Bundles Chronic Bundles $40,000 $35,000 $30,000 $25,000 $20,000 $15,000 $10,000 $5,000 $0 Inpatient Part B Post Acute 19

Illustrative Spending: Hospital vs. State Major Spending Categories for Surgical Bundles Note: All DRG bundle cost comparisons are unadjusted for geographic wage index. * Hospital spending for IRF/LTAC significantly exceeds state average. Post ‐ Acute Spend: Hospital vs. State Spending per Case for Surgical Bundles Note: All DRG bundle cost comparisons are unadjusted for geographic wage index. * Hospital spending for IRF significantly exceeds state average. 20

Post ‐ Acute: Hospital vs. State Spending per Case for Chronic/Medical Bundles Brandeis University 41 Episode Price Variance: Risk Track A 21

Managing Financial Risk in BPCI Illustration of Risk within a Bundle (or DRG) 44 Brandeis University 22