Post Acute Care Services Are the Fastest Growing Category of - PDF document

Post ‐ Acute Care: The Next Frontier for Controlling Medicare Spending Robert Mechanic, MBA Brandeis University Estes Park Institute November 4, 2014 Post ‐ Acute Care Services Are the Fastest Growing Category of Medicare Spending Why Should Hospitals and Physicians Care? Brandeis University 2 1

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 $ $ $ $ $ What’s in an Episode? Outpatient Professional Professional Inpatient Professional services Inpatient Readmission SNF Index Hospitalization Stays 90 day look ‐ forward Brandeis University 2

CMMI Bundled Payment Pilot Model 1 Outpatient Professional Professional Inpatient Professional services Inpatient Readmission SNF Index Hospitalization Stays 90 day look ‐ forward Brandeis University CMMI Bundled Payment Pilot Model 2 Outpatient Professional Professional Inpatient Professional services Inpatient Readmission SNF Index Hospitalization Stays 30 ‐ 90 day look ‐ forward Brandeis University 3

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

Medicare Spends a Tremendous Amount in the 30 – 90 Days After Patients Are Discharged from the Hospital 9 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 10 5

Medicare Payment Methods • SNF: Per ‐ diem payment with therapies billed separately – Patients covered for up to 100 days • Home health: 60 ‐ day bundle • Inpatient Rehab: Prospective per case payment (similar to DRG method) – 60 percent of patients must have one of 13 conditions Brandeis University 11 Avg. 2008 Medicare Inpatient Payments for Select DRGs 20,000 15,000 10,000 $11,079 5,000 $6,437 $6,075 $5,347 $5,322 0 470 ‐ Maj. Joint 194 ‐ Pne w/CC 292 ‐ Heart Fail 683 ‐ Renal 190 ‐ COPD w/CC Failure w/CC w/MCC Index Admission Source: RTI Inc, Post ‐ Acute Care Episodes: Expanded Analytic File, June 2011 12 6

2008 Medicare Acute and Post ‐ Acute Payments for Inpatient ‐ Initiated 30 ‐ Day Episodes 20,000 $18,414 15,000 $12,456 $10,636 $10,470 $9,732 10,000 5,000 0 470 ‐ Maj. Joint 194 ‐ Pne w/CC 292 ‐ Heart Fail 683 ‐ Renal 190 ‐ COPD w/CC Failure w/CC w/MCC Index Admission Post Acute Source: RTI Inc, Post ‐ Acute Care Episodes: Expanded Analytic File, June 2011. Thirty day fixed episodes include the full 13 amount of all claims incurred within 30 days of discharge even if they extend beyond the 30 days period. 2008 Medicare Acute and Post ‐ Acute Payments for Inpatient ‐ Initiated 90 ‐ Day Episodes $19,745 20,000 $16,589 $14,692 $14,910 15,000 $12,479 10,000 5,000 0 470 ‐ Maj. Joint 194 ‐ Pne w/CC 292 ‐ Heart Fail 683 ‐ Renal 190 ‐ COPD w/CC Failure w/CC w/MCC Index Admission 30 day Post Acute 90 day Post Acute Source: RTI Inc, Post ‐ Acute Care Episodes: Expanded Analytic File, June 2011. 30 ‐ 90 day amounts are estimated based on RTI, Analysis of Acute Care Episode Definitions Chart Book, November 2009. 14 7

There is Significant Variation in Post ‐ Acute Care Spending Across Hospitals …. …. And Many Opportunities to Reduce Post ‐ Acute Care Spending 15 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 16 index. 8

A Tale of Two Hospitals: Joint Replacement Episode 17 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. 18 9

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. 19 Post Acute Strategy Components 1. Right setting 2. Right partners 3. Right relationships – Patient & Family – Primary Care Physician Post ‐ Acute Providers – Brandeis University 20 10

2008 Medicare Post ‐ Acute Care Payments Per User by Site of Service: DRG 470 (Total Joint) Within 30 Days of Hospital Discharge $25,000 $20,000 $15,000 $23,017 $10,000 $12,596 $11,079 $5,000 $9,496 $8,562 $3,132 $0 Admission Home Health SNF Rehab LTAC Readmission Percent with 100% 60% 40% 7% 0.2% 9% Service: 21 Source: RTI Inc, Post ‐ Acute Care Episodes: Expanded Analytic File, June 2011 Variation in 2010 Medicare Average Length of Stay for Skilled Nursing Facilities 61 60 Difference Between Top & Bottom Quartile 10 Days = $4,000+ 50 34 40 29 34 24 29 30 24 20 5 10 0 Quartile 1 Quartile 2 Quartile 3 Quartile 4 Source: Adapted form Office of HHS Inspector General December 2010. 22 11

2013 Average SNF Spending Per Admission for Total Joint Replacement Patients One Large Hospital’s Top 6 SNFs by Number of Admissions $25,000 $20,000 $15,000 $20,717 $10,000 $15,961 $12,835 $5,000 $9,336 $9,299 $7,929 $0 SNF A SNF B SNF C SNF D SNF E SNF F Source: Brandeis University analysis of Medicare claims data. All SNFs have 10+ cases. 23 Variation in 2009 Risk Adjusted Readmission Rates from Skilled Nursing Facilities 30.0% 25.0% 20.0% 15.0% 22.0% 10.0% 18.1% 14.4% 5.0% 0.0% 25th Percentile Median 75th Percentile Readmissions Source: MedPAC Report to Congress, March 2012. 24 12

Expert Panel Ratings of Whether Hospital Admissions from Nursing Home Were Avoidable NH Resident Group Yes No Medicare (n=94) 69% 31% Medicaid/Other (n=106) 65% 35% High Readmit NHs (n=101) 75% 25% Low Readmit NHs (n=99) 59% 41% All Residents (n=200) 67% 33% Source: Ouslander et al. Potentially avoidable hospitalizations of nursing home residents: frequency, causes and costs. JAGS 58:627 ‐ 635, 2010. 25 Causes of Potentially Avoidable Admissions Diagnoses for Potentially Avoidable Admissions n=100 Cardiovascular (mostly CHF, chest pain) 22% Respiratory (mainly pneumonia, bronchitis) 21% Mental status change 13% Urinary tract infection 11% Sepsis or fever 8% Skin (cellulitis, wound, pressure ulcer) 8% Dehydration 7% Gastrointestinal (bleeding, diarrhea) 7% Source: Ouslander et al. Potentially avoidable hospitalizations of nursing home residents: frequency, causes and costs. JAGS 58:627 ‐ 635, 2010. Brandeis University 26 13

Ratings of Factors Associated with Potentially Avoidable Admissions Important Somewhat Factors that could have prevented Important hospitalization NH should have been able to do everything 50% 34% done by hospital Better quality by NH physician or AP 55 28 One MD visit could have prevented transfer 37 43 Better quality by NH staff 24 48 Better advance care planning 38 24 Resident’s condition limited ability to 19 28 benefit from the transfer Source: Ouslander et al. Potentially avoidable hospitalizations of nursing Brandeis University 27 home residents: frequency, causes and costs. JAGS 58:627 ‐ 635, 2010. Resources Rated as Potentially Helpful in Preventing Avoidable Admissions Would Very or Prevent Somewhat Transfer Helpful Exam by MD or APC within 24 hrs 40% 52% MD or APC in NH 3 days/week 16 80 RN providing care vs. LPN or Aide 6 85 IV therapy available in NH 22 58 Lab tests within 3 hours 15 74 Source: Ouslander et al. Potentially avoidable hospitalizations of nursing home residents: frequency, causes and costs. JAGS 58:627 ‐ 635, 2010. Brandeis University 14

Assessing Quality is Difficult 29 Developing a Preferred SNF Network Right Right Geography Geography Ability to Ability to Willing to Willing to Manage Manage Complex Complex collaborate collaborate Patients Patients on QI on QI Criteria Criteria On ‐ Site On ‐ Site Strong Strong MD MD Performance Performance Metrics Metrics Coverage Coverage Source: Atrius Health. Brandeis University 15

Performance in One Health System’s Preferred SNF Network Average Length of Stay 35 32 6 Days = $2,400+ 30 21 25 20 Medicare Advantage 15 Pioneer ACO 15 Market Avg. 10 5 0 Brandeis University 31 Source: Atrius Health, 2013. Performance Expectations • Appropriate staffing ‐ low staff turnover • Able to manage complex patients • Able to treat acute exacerbations in place • Use “preferred” MD/APC provider with 24/7 coverage Brandeis University 32 16