THE IMPACT OF ACO PARTICIPATION ON POST-ACUTE CARE UTILIZATION AMONG MEDICARE BENEFICIARIES Amol Navathe, MD, PhD Academy Health Annual Research Meeting June 26, 2017 Department of Health The Leonard Davis Care Management Institute of Health University of Pennsylvania Economics School of Medicine Amol Navathe MD, PhD – not for reproduction without permission
Background • Medicare Post Acute Care (PAC) spending is BIG ($60B in 2015) • PAC is the largest driver of spending variation in Medicare • Medicare has introduced alternate payment models such as bundled payment and ACOs, in part, to reduce unwarranted variation in PAC 2
Study Motivation • Early evidence suggests ACOs achieved small decreases in institutional PAC use and spending for beneficiaries attributed to ACOs • However, only 15% of beneficiaries are attributed to ACOs Do the effects of ACOs spill over to all beneficiaries? 3
Objective • To evaluate whether hospitals participating in the Medicare Shared Savings Program (MSSP) experienced hospital-wide changes in PAC use and Medicare spending 4
Methods • Design: Difference-in-differences study comparing changes in PAC use and Medicare spending among patients admitted to ACO hospitals versus changes at non-ACO hospitals • Setting : National sample of hospitals and Medicare-certified PAC providers between 2010-2013 • 233 MSSP ACO hospitals • 15,276 SNFs initiating in 3 waves • 1,216 IRFs • 3,103 non-ACO • 9,651 HHAs hospitals 5
Methods • Study Population: 11,683,573 Medicare beneficiaries experiencing 26,503,086 hospital admissions • Outcomes : Proportion of discharges and Medicare payments to SNFs, IRFs, and HHAs • Sub-groups: 6 of the most frequent conditions for PAC use and highest-severity 6
Sample Characteristics ACO Non-ACO Unique Beneficiaries 1,290,648 10,392,925 Total Number of admissions 2,933,534 23,569,552 Age 79.2 79.0 Male 42% 42% Black 8% 10% Elixhauser comorbidity score 9.0 8.8 Primary Diagnosis Lower extremity joint replacement 5.0% 5.1% Heart Failure 5.0% 5.2% Pneumonia 3.7% 4.1% 30-day Readmission Rate 16.3% 16.2% 30-day Mortality Rate 4.7% 4.9% 7
Hospital and Market Characteristics Hospital Number of beds Teaching status Ownership PAC in same hospital ACO HHA SNF IRF Market Number of ACOs ACO hospital admission share Number of PACs SNF IRF HHA Herfindahl-Hirschman index (HHI) 8
Unadjusted Trends in Institutional PAC (SNF/IRF) Use 70% Non-ACO Hospitals 60% ACO Wave 1 Hospitals Percent of Discharges ACO Wave 2 Hospitals 50% ACO Wave 3 Hospitals 40% 30% 20% 10% 0% Quarter 9
Risk-adjusted Change in Probability of Discharge to Institutional PAC from ACO vs. Non-ACO Hospital 5% 4% 3% Change in Probability 2% 1% 0% -1% -2% -3% -4% -5% High- All Hip Fx Joints CABG CHF COPD AMI Risk Patient Population 10
Risk-adjusted Change in Probability of Discharge to Any PAC from ACOs with Co-Participating SNF, IRF, HHA 5% 4% Change in Probability 3% 2% 1% 0% -1% -2% -3% -4% -5% SNF IRF HHA 11
Risk-adjusted Change in Medicare Payments by PAC type from ACOs vs. non-ACOs $400 $300 Change in PAC Payments $200 $100 $0 -$100 -$200 -$300 -$400 IRF HHA SNF 12
Conclusions • Hospital participation in an ACO did not result in changes in PAC utilization or payments – statistical zeros • This was true even when considering high PAC-use conditions like joint replacement and heart failure • Inclusion of hospitals and PAC providers in the same ACO did not alter the effect 13
Limitations • ACO hospitals may differ from non-ACO hospitals in unobservable ways because the program is voluntary • We focus on a sub-set of ACOs that include a hospital • Use of administrative data may limit risk- adjustment 14
Implications • ACO hospitals are not redesigning PAC referral patterns hospital-wide, but selecting for attributed beneficiaries • Integration of hospitals and PAC providers within ACOs may not be enough • This suggests a certain degree of caution on the expectations for ACO programs to curb high PAC use and variation nationwide 15
Acknowledgements • Co-investigators: • Rachel Werner, MD, PhD, University of Pennsylvania • Alex Bain, BS, University of Pennsylvania • Funders: • AHRQ and NIA 16
Questions? amol@wharton.upenn.edu 17
APPENDIX 18
Hospital and Market Characteristics Hospital (n = 3,336) ACO Hospitals (N=233) Non-ACO Hospitals (N=3103) No. certified beds, mean (SD) 356 (279) 222 (220) RB ratio (Teaching status) (%) Non-teaching (RB ratio = 0) 65.67 81.31 Minor teaching (RB ratio < 0.25) 21.89 13.66 Major teaching (RB ratio ≥ 0.25) 12.45 4.99 Ownership (%) Government 10.30 18.34 Not for profit 82.40 58.06 For profit 7.30 23.60 PAC in same hospital ACO (%) HHA 25.32 NA SNF 13.73 NA IRF 51.93 NA Market (n = 307) b With ACO (N=95) Without ACO (N=212) No. ACOs in market 1.38 (0.74) N/A ACO hospital admission in market (%) 32.38 (22.10) N/A No. PACs in market (%) SNF 73.43 (63.86) 39.15 (35.81) IRF 5.91 (6.01) 3.09 (3.36) HHA 51.69 (78.74) 22.36 (27.04) HHI, mean (SD) c\ Hospital 2171.96 (1844.53) 3185.14 (2022.89) Aggregate PAC 392.31 (275.00) 622.50 (423.50) 19
Unadjusted Trends in Home Health Agency Use 70% Non-ACO Hospitals 60% ACO Wave 1 Hospitals Percent of Discharges ACO Wave 2 Hospitals 50% ACO Wave 3 Hospitals 40% 30% 20% 10% 0% Quarter 20
Risk-adjusted Change in Probability of Discharge to Any PAC from ACO vs. Non-ACO Hospital 5% 4% 3% Change in Probability 2% 1% 0% -1% -2% -3% -4% -5% All High-Risk Hip Fx Joints CABG CHF COPD AMI Patient Population 21
Change in Probability of Discharge to Post Acute Care from ACO Hospital, by Location Unconditional Restricted to DRG group b Acute Joint Hip and Femur Coronary Chronic Lung Myocardial Unrestricted High Risk, Replacement Fracture Artery Bypass Heart Failure Disease Infarction (n=25,528,375) (n=5,173,403) (n=1,275,791) (n=408,487) (n=165,139) (n=1,319,929) (n=1,101,764) (n=996,440) Discharge to 0.000 (-0.003, -0.003 (-0.007, -0.009 (-0.018, -0.005 (-0.011, 0.006 (-0.018, -0.002 (-0.011, -0.005 (-0.014, -0.001 (-0.001, PAC (95% CI) 0.003) 0.002) 0.001) 0.002) 0.030) 0.006) 0.003) 0.007) p-value 0.89 0.23 0.08 0.15 0.61 0.58 0.22 0.83 Discharge to 0.000 (-0.003, -0.004 (-0.008, -0.006 (-0.019, -0.004 (-0.010, 0.002 (-0.019, -0.002 (-0.009, -0.005 (-0.012, -0.003 (-0.008, SNF (95% CI) 0.002) 0.000) 0.006) 0.011) 0.022) 0.005) 0.001) 0.003) p-value 0.73 0.06 0.33 0.58 0.87 0.59 0.11 0.39 Discharge to 0.000 (-0.002, 0.000 (-0.002, 0.001 (-0.007, 0.0006 (-0.011, 0.009 (-0.006, 0.001 (-0.001, 0.000 (-0.002, 0.002 (0.000, IRF (95% CI) 0.002) 0.002) 0.010) 0.012) 0.023) 0.003) 0.001) 0.003) p-value 0.96 0.69 0.80 0.92 0.25 0.19 0.73 0.08 Discharge to 0.001 (-0.002, 0.001 (-0.002, -0.004 (-0.017, -0.001 (-0.006, -0.004 (-0.010, -0.002 (-0.008, 0.000 (-0.006, 0.000 (-0.006, HHA (95% CI) 0.003) 0.004) 0.009) 0.003) 0.031) 0.005) 0.007) 0.006) p-value 0.57 0.68 0.58 0.53 0.72 0.60 0.94 0.99 Conditional on Discharge to PAC Restricted to DRG group Acute Joint Hip and Femur Coronary Chronic Lung Myocardial Unrestricted High Risk Replacement Fracture Artery Bypass Heart Failure Disease Infarction (n=10,218,784) (n=2,429,937) (n=1,008,231) (n=370,299) (n=101,759) (n=513,743) (n=313,249) (n=220,427) Discharge to 0.000 (-0.005, -0.003 (-0.009, -0.002 (-0.016, 0.000 (-0.014, 0.006 (-0.019, 0.000 (-0.012, 0.000 (-0.015, -0.004 (-0.021, SNF (95% CI) 0.005) 0.003) 0.013) 0.014) 0.032) 0.012) 0.015) 0.013) p-value 0.98 0.32 0.80 0.99 0.63 0.99 1.00 0.64 Discharge to 0.000 (-0.034, 0.001 (-0.003, 0.002 (-0.009, 0.001 (-0.012, 0.009 (-0.013, 0.003 (-0.002, -0.001 (-0.005, 0.008 (-0.001, IRF (95% CI) 0.004) 0.005) 0.013) 0.014) 0.030) 0.007) 0.004) 0.014) p-value 0.92 0.50 0.66 0.87 0.43 0.20 0.82 0.03 Discharge to 0.000 (-0.005, 0.002 (-0.004, -0.001 (-0.015, -0.001 (-0.006, -0.015 (-0.040, -0.003 (-0.015, 0.001 (-0.015, -0.004 (-0.020, HHA (95% CI) 0.004) 0.007) 0.013) 0.004) 0.010) 0.009) 0.016) 0.013) 22 p-value 0.96 0.54 0.94 0.66 0.24 0.63 0.94 0.67
Risk-adjusted Change in Probability of Discharge from ACOs with Co-Participating SNF, IRF, HHA 23
Percent Discharge to Home 70 60 Discharge Percent (%) 50 40 30 Non-ACO Hospitals ACO Wave 1 Hospitals 20 ACO Wave 2 Hospitals ACO Wave 3 Hospitals 10 0 Quarter 24
Percent Discharge to SNF 70 60 Non-ACO Hospitals ACO Wave 1 Hospitals 50 Discharge Percent (%) ACO Wave 2 Hospitals ACO Wave 3 Hospitals 40 30 20 10 0 Quarter 25
Percent Discharge to IRF 70 Non-ACO Hospitals ACO Wave 1 Hospitals 60 ACO Wave 2 Hospitals ACO Wave 3 Hospitals 50 Discharge Percent (%) 40 30 20 10 0 Quarter 26
Percent Discharge to SNF Conditional on PAC Discharge 70 60 Discharge Percent (%) 50 40 Non-ACO Hospitals 30 ACO Wave 1 Hospitals ACO Wave 2 Hospitals 20 ACO Wave 3 Hospitals 10 0 Quarter 27
Percent Discharge to IRF Conditional on PAC Discharge 70 Non-ACO Hospitals 60 ACO Wave 1 Hospitals Discharge Percent (%) 50 ACO Wave 2 Hospitals ACO Wave 3 Hospitals 40 30 20 10 0 Quarter 28
Recommend
More recommend
