Health Savings Accounts in a Medicaid Expansion Population Joseph - - PowerPoint PPT Presentation

June 14, 2017

Arkansas Experience with “Health Savings Accounts” in a Medicaid Expansion Population

Joseph W. Thompson, MD MPH Anthony Goudie, PhD Jeral Self, MPH Anuj Shah, MS

• J. Mick Tilford, PhD

Background: Consumer Driven Health Care and Medicaid

• Medicare Modernization Act (2003): Health

Savings Accounts enabled as tax shelter associated with high deductible plans

• Deficit Reduction Act (2005): 10-state

demonstration of Health Opportunity Accounts with Medicaid funding (SC only participant)

• Medicaid 1115 Waivers: cost-sharing strategies

– MI Health Accounts – IN Personal Wellness and Responsibility Accounts – AR Health Independence Accounts

Objectives:

• Profile the experience of the Arkansas

Health Independence Accounts (HIA)

• Assess characteristics of participating

individuals

• Evaluate the financial impact of the HIA
• Contribute to the ongoing policy

dialogue

Arkansas Policy Environment

• Arkansas Health Care Independence Act

– Section 1115 Waiver utilizing premium assistance to purchase plans on the private individual marketplace – Private sector cost-sharing for those >100% FPL

(e.g., $8-10 for clinic visit; $4 for generic drugs)

– Implemented January 1, 2014

• Health Independence Accounts (HIA)

– Operational 1/1/15 – Originally for individuals 50-138% FPL /

• nly implemented for those 100-138% FPL

Legislative Intent for HIA

• Citizens to gain knowledge about appropriate

healthcare services and how much those services cost

• Citizens to gain experience paying cost-sharing

requirements and introduction of the concept of paying premiums for insurance

• Promote personal responsibility for health care

decision making

• Accrue funds to offset premiums and foster a

cost-sharing model in the Insurance Marketplace when beneficiary incomes are over 138% FPL

Health Independence Accounts:

• Individual accounts for those 100-138% FPL

– Individuals required to contribute or generate a debt to the state (no collection mechanism)

• 100-117% FPL $10/mo / 118-138% FPL $15/mo
• Payment resulted in state-funded cost-sharing

protection for following month

– Activation of account gained 2 months protection followed by payment expectations – State debits account only for failed payment

• State matches individual’s contribution up to

$200 if timely payments / annual roll-over

• Funds available for premium payments upon exit

Study Data and Methods Employed

• Data:

– Enrollment files for Medicaid beneficiaries between 100-138% FPL in premium assistance – Claims data from Qualified Health Plans (2014) – Financial transactions (collection and distribution) from third-party HIA administrator (2015-2016)

• Individuals (N=57,079) profiled and compared

for those making/not making contributions on:

• Demographic characteristics
• Charlson co-morbidity index
• Payment contribution frequency and amounts
• Cost-sharing protection received
• Net economic impact – individual and programmatic

Results: Participants by Income

Federal Poverty Level Required Monthly Payment Made No Payments Made At Least One Payment Total 100-117% $10 22,184 (86.9%) 3,332 (13.1%) 25,516 (44.7%) 118-138% $15 27,823 (88.2%) 3,740 (11.9%) 31,563 (55.3%) Total 50,007 (87.6%) 7,072 (12.4%) 57,079 (100%)

Results: Demographics of Participants

Category Made No Payments Made At Least One Payment Total Age p<0.0001 19-34 24,154 (94.1%) 1,508 (5.9%) 25,664 (45.0%) 35-49 14,279 (88.0%) 1,954 (12.0%) 16,233 (28.4%) 50-64 11,572 (76.2%) 3,610 (23.8%) 15,182 (26.6%) Gender Male 20,130 (87.7%) 2,831 (12.3%) 22,961 (40.2%) Female 29,877 (87.6%) 4,241 (12.4%) 34,118 (59.8%) Race / Ethnicity P<0.001 White 32,073 (87.9%) 4,400 (12.1%) 36,473 (63.9%) Black 7,595 (87.9%) 1,050 (12.1%) 8,645 (15.2%) Hispanic 1,865 (85.2%) 324 (14.8%) 2,189 (3.8%) Other 8,474 (86.7%) 1,298 (13.3%) 9,772 (17.1%)

Results: Prior Utilization / Risk Profile

• Prior hospitalization(s) greater for participants

7.1% vs 5.1% *

• Prior emergency room less for participants

25.7% vs 27.7% *

• Charlson Co-Morbidity

Indices *

* p<0.0001

Outcomes: Cost-sharing protections

• A total of $476,843 in cost-sharing protections

associated with payments were achieved:

– median cost-sharing payment was $8 (IQR $0, $64) – mean payment of $67 (Std Dev=$141)

Type Number $$

Pharmaceuticals 31,805 $289,522 Physician 9,198 $79,482 Non-MD Clinician 7,339 $59,095 Hospitals 2,884 $41,744 Other 710 $7,000 Total 51,936 $476,843

*An additional $118,294 were paid during “free” two-month start-up period

Outcomes: Economic Impact

• Premium payments for the ~7,000 individuals

totaled $426,670

– Median total $40 (IQR $15, $90): 4 months payments – Mean total $60 (Std Dev $141): 6 months payments

• Cost-sharing protections exceeded HIA

payments by $50,172 ($476,843 – 426,670)

• Cost-avoidance (individual co-payment

protections greater than monthly payments) was realized by 23.4% of participants

• Administrative costs to maintain program and

accounts were ~$9M over 18 months

Limitations

Our results are based upon observational information related to individual actions; we lack information on beneficiaries’:

– Health literacy related to insurance design – Awareness of program design or purpose – Experiences in program participation – Perceived value or risk of participating – Ability to anticipate cost-sharing events – Relative income stability within households – Reasons for intermittent payment patterns

Conclusions:

• Introduction of “savings” accounts into

Medicaid demonstrated limited participation

• Medicaid beneficiaries that are likely to

participate in personal “savings” accounts are likely to be older, have more conditions, and previously hospitalizations

• Participating individuals demonstrated rational

economic behavior with some achieving economic advantages

• Operational and fiscal costs of such efforts

should be considered prior to implementation

Postlude:

• The Arkansas General Assembly

terminated the Healthcare Independence Accounts in 2016 and replaced them with nominal monthly premiums

• Individuals with account balances

(n=2253) received program termination checks

• No awareness of individuals directly

utilizing account balances for private sector premium payments

Thank you

Access Cost Quality