and Adverse Health Events Amy M. G. Kandilov, Vince Keyes, Noelle - PowerPoint PPT Presentation

RTI International The Effect of the Vermont Support and Services at Home (SASH) Program on Medicare Expenditures and Adverse Health Events Amy M. G. Kandilov, Vince Keyes, Noelle Siegfried, Patrick Edwards RTI International Alisha Sanders, Robyn Stone LeadingAge Center for Applied Research Martijn van Hasselt UNC - Greensboro www.rti.org RTI International is a trade name of Research Triangle Institute. 1

RTI International Funding Sources Funding for this research came from HHS Office of the Assistant  Secretary for Planning and Evaluation (ASPE), in partnership with the Department of Housing and Urban Development (HUD) and the HHS Administration for Community Living (ACL), under contract number HHSP23337006T. The statements contained in this presentation are solely those of the  authors and do not necessarily reflect the views or policies of ASPE, HUD, or ACL. 2

RTI International Introduction to the SASH Program  The Support and Services at Home (SASH) program in Vermont was developed by the nonprofit Cathedral Square Corporation to help residents of publicly-assisted housing access the health care and support services they need to stay healthy and age comfortably and safely at home.  Each SASH “panel” contains up to 100 participants, who are served by a full-time SASH coordinator and a quarter time wellness nurse  SASH staff offer participants many services, including care coordination and wellness programs  Most panels are operated within a non-profit, affordable congregate housing site 3

RTI International Introduction to Multi-Payer Advanced Primary Care Practice Demonstration  Primary source of funding for the SASH program comes from the Centers for Medicare & Medicaid Services (CMS), through the MAPCP Demonstration  In the MAPCP Demonstration, CMS joined eight state initiatives (including the Blueprint for Health in Vermont) by providing financial incentives for physician practices to become patient-centered medical homes  CMS provides $68,000 per year per 100-participant panel, which is roughly 2/3 of SASH panel operating costs 4

RTI International Research Objective  Our objective is to estimate the impact of the first three years of the SASH program on the Medicare expenditures and adverse health events of SASH participants living in affordable congregate housing  Outcomes of interest: Medicare expenditures, all-cause hospitalizations, and ER visits (all-cause, and ER visits not leading to hospitalization) 5

RTI International Study Design  We use a difference-in-differences linear regression model for the expenditure outcomes, comparing the change in the quarterly outcomes among the SASH participants with the change in the quarterly outcomes for a comparison group of Medicare beneficiaries in affordable congregate housing who are not participating in SASH.  Our model controls for beneficiary-level demographic and health characteristics, and it includes fixed effects for the time quarters and for each of the congregate housing properties.  Comparison group is chosen through propensity score matching 6

RTI International Population Studied  Across the first three years of the SASH program (July 2011 through June 2014), there were 3,485 SASH participants  The intervention group for this study consists of 1,602 Medicare fee-for-service beneficiaries participating in the SASH program and living in SASH housing sites  Both intervention and comparison groups are residents of properties that receive funding assistance through the U.S. Department of Housing and Urban Development, or tax credits through the Low Income Housing Tax Credit. 7

RTI International SASH Participants Included in the Regression Analysis 1,602 included in sample: 1,252 matched to PIC/TRACS 350 matched to LIHTC only 2,260 Medicare FFS 3,485 SASH participants with beneficiaries attributed to start date before Blueprint for Health 7/1/14 practices 658 excluded: not found in PIC/TRACS or LIHTC 1,225 excluded: housing records 313 not Medicare FFS 912 not attributed to Blueprint practices NOTES: SASH, Support and Services at Home; FFS, fee-for-service; PIC, Public and Indian Housing Information Center database; TRACS, Tenant Rental Assistance Certification System; LIHTC, Low Income Housing Tax Credit 8

RTI International Subgroups of SASH panels We examine how the effects of SASH differ between early SASH  panels (starting before April 1, 2012) and late SASH panels (starting on or after April 1, 2012) Given considerable start-up efforts, early panels may be more  effective at reducing health care expenditures/utilization We also consider how the effect of SASH differs in site-based  panels vs. mixed panels. The majority of participants in site-based panels live in the SASH  housing host site; the majority of participants in mixed panels live in the community Panels will a high concentration of community participants may have  fewer resources available to assist each participant 9

RTI International Effect of the SASH program on monthly Medicare expenditures (1) (2) (3) (4) (5) All SASH Early Late Site-based Mixed Expenditure type participants SASH panels SASH panels panels panels (n=1,602) (n=699) (n=933) (n=1,218) (n=384) -12.31 -127.99* 62.18 -65.76 121.25 Total Medicare (57.1) (71.7) (71.55) (62.12) (95.78) 5.08 -27.97 26.36 -15.27 56.25 Acute care (33.61) (42.49) (41.7) (36.78) (54.93) 5.44 -21.91 27.56 -8.63 48.96 Post-Acute Care (17.86) (21.23) (22.54) (18.95) (31.03) -4.54 -9.18** -2.62 -6.19 -1.83 Emergency room (3.75) (4.17) (4.90) (3.95) (6.47) -10.33 -26.56* -2.13 -17.95 7.20 Hospital outpatient department (11.70) (14.33) (14.18) (12.44) (19.54) -1.69 -9.11* 3.69 -4.27 5.57 Primary care/Specialist physician (4.15) (5.26) (5.21) (4.58) (6.47) Source: Authors’ analysis of Medicare claims data January 2006 through June 2014 10

RTI International Principal findings: Monthly expenditures  Overall, there is no significant effect of SASH on monthly Medicare expenditures for the SASH participants as a whole  SASH participants in early panels have significantly lower growth in total Medicare costs of $128.  Expenditures for emergency room, hospital outpatient departments, and physicians are also lower for the SASH participants in early panels  No difference in cost growth for other types of SASH panels 11

RTI International Effect of the SASH program on monthly Medicare expenditures, annual aggregates (1) (2) (3) (4) (5) All SASH Early Late Site-based Year Mixed panels participants SASH panels SASH panels panels (n=384) (n=1,602) (n=699) (n=933) (n=1,218) -24.54 -52.55 -52.28 -22.43 -48.66 Year one (79.27) (90.35) (177.12) (84.73) (195.09) 164.51* -33.95 317.4* 108.02 253.51 Year two (81.72) (106.64) (111.43) (91.14) (156.5) -124.07 -221.25* -71.76 -188.45* 59.8 Year three (76.02) (94.24) (85.91) (81.2) (116.36) -12.31 -127.99* 62.18 -65.76 121.25 All years combined (57.1) (71.7) (71.55) (62.12) (95.78) 12

RTI International Principal findings: Annual aggregate expenditures  For the early SASH panels, the largest decrease in Medicare cost growth comes in Year Three  Site-based SASH panels also have decreased Medicare cost growth in Year Three, although not overall  Note that the majority of the early SASH panels are site- based panels 13

RTI International Effect of the SASH program on quarterly adverse health event measures (2) (3) (1) (4) (5) Early Late All SASH Site-based Mixed Utilization outcome SASH SASH participants panels panels panels panels (n=1,218) (n=384) (n=1,602) (n=699) (n=933) 2.78 -5.54 7.28 -2.05 17.76 All-cause, acute care hospitalizations (6.33) (7.59) (8.85) (6.41) (14.53) 6.72 3.23 6.53 3.23 18.91 All-cause ER visits (8.60) (10.49) (11.13) (8.90) (16.79) 2.42 1.53 1.30 1.35 6.53 ER visits not leading to a hospitalization (5.29) (6.76) (6.79) (5.62) (10.03) 14

RTI International Principal Findings: Adverse health events  No statistically significant findings for any of the three utilization measures  Further research will look at the intensity of the acute case hospitalizations and ER visits, to better explain why expenditure growth is declining but utilization is not 15

RTI International Conclusions  The SASH panels with the earliest start dates experienced lower growth in Medicare expenditures, with the strongest results in the third year of SASH implementation.  This decreased growth in Medicare expenditures over the three years for the early SASH panels is consistent with the hypothesis that the start-up activities for a SASH panel reduce that panel’s ability to make a significant impact on Medicare expenditures and utilization in the first few quarters of operation.  Site-based SASH panels exhibit lower growth in Medicare expenditures in the third year of the SASH program 16

RTI International Policy Implications  Our principal findings suggest that SASH is a promising intervention for reducing the growth of Medicare expenditures among a population of elderly and disabled residents living in affordable congregate housing.  Further research is needed to determine if Medicare cost growth can be reduced as newer SASH panels mature  This model of providing in-home services for elderly residents of affordable congregate housing properties may have viability beyond Vermont. 17