Summary of the "Hsiao" Report

Note: This summary references the draft of 21 January 2011. A final version of the report is due by 17 February, and this summary will be updated to reflect changes in that version.

The Hsiao team’s approach to healthcare reform in Vermont is fairly simple:

  • improve the efficiency of the healthcare system by:
    • implementing a standard benefit package that encourages primary and preventive care;
    • reducing administrative waste by implementing a single provider payment channel;
    • encouraging integrated delivery of healthcare and use of information technology;
  • maximize the amount of federal money coming to the state;
  • use the funds made available to cover everyone and to further improve the system.

Background

Act 128, which passed in the spring of 2010, calls for the design of a new healthcare system for Vermont. To develop this design, the Legislature hired a team of experts led by Dr. William Hsiao of Harvard University.

To guide the process of developing a new healthcare system for Vermont, Act 128 spells out the principles and goals of healthcare reform. Among these principles are those of the Healthcare Is a Human Right campaign:

  • Universality — All Vermonters must have access to comprehensive, quality health care.
  • Equity (in access) — Systemic barriers must not prevent people from accessing necessary healthcare.
  • Equity (in financing) — The financing of health care must be sufficient, fair, sustainable, and shared equitably.
  • Transparency & Accountability — The healthcare system must be transparent in design, efficient in operation, and accountable to the people it serves.
  • Participation — The state must ensure public participation in the design, implementation, evaluation, and accountability mechanisms in the health care system.
  • Public Good — Healthcare is a public good for all Vermonters.

Instead of just a single design, Act 128 calls for three options and specifies the healthcare system components that must be incorporated into each option:

  • a payment system;
  • coordinated regional delivery systems;
  • health system planning, regulation and public health;
  • financing;
  • compliance with federal law.

Structure of the Report

The “Hsiao” report consists broadly of three parts:

  1. five introductory sections (#1-5):
    • a very cursory overview of the last 70 years of legislative healthcare reform efforts in Vermont, the team’s interpretation of the principles and goals of Act 128 and a reiteration of the current problems in Vermont’s healthcare system;
    • descriptions of the constraints imposed on healthcare reform in Vermont;
    • design principles & strategies;
    • methods of estimating savings, costs and micro- and macro-economic effects;
    • the effects of the Patient Protection and Affordable Care Act (PPACA);
  2. three sections (#6-8) describing the healthcare system designs called for in Act 128:
    • publicly-administered “single payer” options — “comprehensive” (Option 1a) and “essential” (Option 1b) benefit packages;
    • a “public option” (Option 2);
    • a “public/private single payer” option with “essential” benefit package (Option 3);
  3. two concluding sections, the first (#9) discussing implementation issues related to single-payer systems and accountable care organizations (ACOs), the second (#10) very briefly comparing the options described in sections 6-8 and making the Hsiao team’s recommendation, which is Option 3.

The Constraints

A great deal of the work of the Hsiao team consisted of studying the constraints imposed on healthcare reform in Vermont. Nearly a quarter of the body of the report is devoted to describing the constraints as perceived by the consultants. Some of these perceived constraints are political or financial; others are legal, and others are practical. Much of Hsiao’s realpolitik follows from the constraints that he describes:

  • benefits — few people want to see their current healthcare benefits reduced;
  • costs — few people (or businesses) will accept an increase in the amount spent on healthcare;
  • federal law — Vermont must comply with laws and regulations related to Medicaid and Medicare, as well as the Employee Retirement Income Security Act (ERISA) and PPACA;
  • payment to providers (hospitals and physicians) — most Vermont hospitals have very slim financial margins, and most physicians would not accept a reduction in the amount they are paid;
  • supply of providers and healthcare capacity — there is a shortage of primary care physicians and nurse practitioners in Vermont, and some hospitals need renovation;
  • grassroots concerns — people want healthcare as a human right;
  • operational — government must be able to execute the plan for reforming the system.

Design Principles & Strategy

The Hsiao team’s design principles are somewhat different from the principles enshrined in Act 128, although the report was supposed to reflect all of Act 128’s principles. This is how the Hsiao team states their design priniciples:

  • shifting the focus of care from intervention to prevention and wellness;
  • shifting the structure of care from fragmentation to organization;
  • providing both coverage and access to care for all Vermonters;
  • financing care in an equitable and sustainable way;
  • maximizing efficiency and transparency throughout the system.

The Hsiao team describes their strategy for implementing these principles in a single page. Here is that description in its entirety.

  • First and most importantly, we wanted to design a system that could achieve universal coverage for residents of Vermont, providing everyone with financial risk protection and access to care. This came with one important caveat, however: that the cost of covering the uninsured and underinsured would be paid for entirely with the savings generated by our reforms.
  • We examined a multitude of potential overarching designs in order to maximize the savings that could be generated by health system reform in Vermont. We explored potential savings from several avenues: administrative savings, a reduction in fraud and abuse, the move towards an integrated delivery system, and malpractice reform. At the same time, we analyzed various methods of financing that would help maximize these savings, achieve universal coverage, and satisfy legal constraints.
  • We designed an Essential Benefit Package with an eye to the average level of benefits currently enjoyed by Vermonters to ensure that they are not losing coverage. Furthermore, the benefit structure was designed to promote not only preventive care and early detection, but also early treatment and wellness services.
  • We designed payment methods to promote the integration of care and reduce clinical waste and overuse.
  • We aimed to increase the supply of and access to physicians and high quality health care. We achieved this by recommending investments – again, financed solely from savings to the system representing at no additional overall spending – to improve health care facilities and increase the number of physicians. Our reforms further aimed to increase current physicians’ patient care time by reducing unnecessary paperwork and administrative burdens, and ensuring that, on average, overall physician net income does not change.
  • Furthermore, our designs attempted to always maximize and protect the federal revenue that can be obtained for Vermont. This applies to our designs with respect to Medicaid and Medicare payments for Vermonters and the potential payments from PPACA. This led us to recommend that Vermont Medicaid raise its payment rates to providers to maximize the federal matching funding. As detailed in Section 4B, if implemented today, this could bring in additional $40 million in 2010 federal funds.
  • PPACA has the potential to annually bring in more than $400 million in 2010 dollars of new federal funding into Vermont when it is fully implemented. As such, we believe that Vermont should continue with Exchange planning and that furthermore the state should begin the implementation of any system reforms in 2015 to lock in these funds and provide the basis for negotiating a reasonable wavier from the Exchange requirements in this year (See Section 2B).
  • This timeline is also consistent with our analysis of the time it might take to create and reorganize the current infrastructure to implement a single payer system. Indeed for all our estimates and design elements, we restrained ourselves to evidence-based, achievable figures and realistic timeframes and assumptions.

Estimating Savings

The report describes four general areas of cost savings, totaling 24.3 - 25.3% (with a ±15% margin of error) for the “single payer” options (Options 1 and 3):

  • reduction in waste and in duplication of care, generated by shifting to an integrated delivery system with health information technology — 10%;
  • administrative savings (to insurer, physicians and hospitals) from shifting to a “single payer” system (a single payment channel and a single benefit package) — 7.3 - 7.8%;
  • reduced fraud and abuse, due to heightened detection ability resulting from the implementation of a state-wide claims database — 5%;
  • savings from implementing a no-fault medical malpractice system — 2%.

The report also asserts that an improved management structure — the result of a competitive bidding process for the claims administration role — could reduce costs by an additional 0.5%.

They estimate that much of these savings would accrue in the first two years after implementation, with the remainder following over five additional years.

The savings from the “public option” (Option 2) would be much smaller, totaling 16.1% (±15%), due to much smaller administrative savings and much less opportunity to reduce waste and duplication of care.

Estimating Costs

One of the inputs into the micro-economic model (below) is the current cost of health insurance in Vermont. The Hsiao team found that the average annual health insurance premium in 2009 was between $4,670 and $5,368 (for individuals between 18 and 64).

They also estimated the average actuarial ratio, the ratio of costs paid by the insurer (as opposed to out-of-pocket payments) to total charges. Though this average varied widely across insurance plans (with some below 50%), an average of 88% for medical care and 79% for drugs was found, producing a composite average of 87%. This number falls between the “gold” and “platinum” packages defined in PPACA.

With the actual costs estimated, the Hsiao team used the micro-economic model to calculate the additional costs of the changes required by the new healthcare system:

  • covering the uninsured;
  • achieving a uniform actuarial ratio;
  • covering dental and vision care;
  • covering long-term care;
  • establishing uniform payment rates.

The Economic Models

Two economic models were used in the Hsiao team’s work: a micro-economic model and a macro-economic model.

The purpose of the micro-economic modeling is to determine (1) the change in the price of health insurance resulting from healthcare reform and (2) the effects on individuals, families and businesses of those changes. The micro-economic model used by the Hsiao team is the Gruber Microsimulation Model (GMSIM), which has been used by a wide variety of state and federal policy makers over the last decade to estimate the effects of changes in health insurance policy. Though not all of the required data are available, the basis of the model’s estimates is actual (empirical) data. The micro-economic model uses a calculation of the effects of PPACA alone (i.e. no additional Vermont reform) as its baseline. Each of the Hsiao team’s design options is compared to this baseline.

Changes to the healthcare system will also impact the state’s economy generally. To estimate these effects, the team used the services of Regional Economic Models, Inc. (REMI) to develop several macro-economic models of the effects of the design options. The REMI models used four inputs — (1) labor and capital demand, (2) populations and labor supply, (3) wages, prices and profits and (4) industry-specific market shares — to estimate the effects of different reform options on industrial production, employment rates and personal income in Vermont. The macro-economic models correspond to the same scenarios as the micro-economic models: each of the design options compared against a baseline of no reform (except PPACA).

The Patient Protection and Affordable Care Act (PPACA) Baseline

The GMSIM and REMI models were used to calculate the effects of PPACA alone on Vermont. The main effects are as follows (compared to 2010, in 2010 dollars):

  • 32,00 Vermonters will still lack healthcare coverage in 2015, with that number decreasing to 31,000 by 2019;
  • healthcare costs will continue to rise at approximately 6% a year, though some of those costs will be shifted from employers to the federal government;
  • a great deal of additional federal money will become available to Vermont;
  • employment in Vermont will increase by about 1,700 new jobs in 2015 and 2.300 by 2019;
  • total annual economic output is expected to increase by about $125 million by 2015 and $180 million by 2019;
  • approximately 500 people are expected to immigrate to Vermont by 2015, with a total of about 1,400 by 2019, as a result of new employment opportunities (not changes in healthcare availability).

The Publicly-Administered “Single Payer” Options (1a & 1b)

Each of the two variations of Option 1 is a publicly administered health insurance system with a single insurance fund and a single benefit package, financed through employer and employee payroll taxes.* Each would cover all legal Vermont residents, with the exception of Medicare recipients, who would continue to receive Medicare benefits. Similarly, Medicaid benefits would not change, except for being improved where they fall short of the proposed benefit package. The models assume that Vermont would be able to receive Medicare, Medicaid and PPACA (exchange) waivers to facilitate financing the system.

Options 1a and 1b differ in their benefit “package” and in out-of-pocket expenses (actuarial ratio).

Option 1a’s “comprehensive” benefit package would include medical, mental health and substance abuse, drugs, vision care, dental, nursing home and homecare. Option 1a has very low out-of-pocket expenses, with approximate actuarial ratios of 97% for medical and mental health services, 90% for drugs and vision care, and 85% for dental, nursing homes and homecare.

Option 1b’s “essential” benefit package differs from that of Option 1a in not covering nursing home and homecare, and in having more limited vision care and dental services (though this “essential” benefit package is better than what the average private health insurance plan in Vermont currently covers, particularly with respect to vision and dental care services). Option 1b, with an average actuarial ratio of approximately 87%, has higher out-of-pocket costs than Option 1a.

Option 1a and 1b differ greatly in their estimated macro-economic effects.

Under Option 1a, employment in Vermont would grow significantly, with an estimated total net increase in new jobs of 8,500 by 2015 and 7,000 by 2019, compared to the baseline. The gross state product would be about $340 million higher in 2015 and $250 million higher in 2019. These economic changes (not the availability of healthcare) would result in a net migration of people to Vermont of 2,000 by 2015 and 7,000 by 2019.

The macro-economic model estimates that Option 1b would have smaller effects. Employment would increase by about 5,000 jobs by 2015 and 4,000 jobs by 2019. The gross state product would be $190 million higher in 2015 and $130 million higher in 2019. Net migration of people to Vermont is estimated to be 1,000 in 2015 and 3,500 in 2019.

*(Though Hsiao refers to these options and Option 3 as “single payer”, obviously this is something of a misnomer, since these options do not fully integrate all coverage plans — notably Medicare, TRICARE and the Veteran’s Administration.)

The “Public Option” (2)

Option 2 is a publicly administered insurance plan intended to compete with private insurance in a PPACA exchange. Because this option preserves the basic structure of the current multi-payer insurance system, it offers few opportunities for significant cost containment. And, unlike the other options, Option 2 would fail to result in universal coverage. Under Option 2, an estimated 30,000 Vermonters would lack health insurance in 2015, dropping to 28,000 in 2019.

The macro-economic effects of Option 2, in contrast to the other options, are estimated to be largely negative. Compared to PPACA alone, implementation of Option 2 would result in a net loss of 1,200 jobs by 2015 and 3,000 jobs by 2019. The gross state product would decline by $90 million by 2015 and by $230 million by 2019. And Vermont would suffer a net loss of residents — 500 by 2015, 2,200 by 2019.

The “Public/Private Single Payer” Option (3)

Option 3 is identical to Option 1b except for two differences:

  • instead of being purely publicly-administered, Option 3 would be managed by an independent board representing payers (employers, state government and people) and recipients (healthcare providers); this board would have responsibility for negotiating payment rates to providers and changes to the benefit package;
  • claims administration and adjudication would be contracted out through a competitive bid process, rather than being performed by a government agency.

The estimated macro-economic effects of Option 3 are very close to those of Option 1b, though with a slightly smaller effect on the gross state product.

As mentioned above, Option 3 is the option that the Hsiao team recommends.

Implementation

The Hsiao team has proposed the following rough timeline for implementation of the new healthcare system.

  • In the 2011-2012 biennium, the Vermont legislature should draft and pass a health care reform law that institutes a single payer system with integrated service delivery.
  • Also during 2011, work should continue on developing an insurance exchange as dictated by PPACA. Vermont should continue expanding the medical homes programs legislated by the Blueprint for Health.
  • In 2012, Vermont must begin developing a state agency to act as the single payer for the health system.
  • In 2014, the state should establish an Insurance Fund and prepare the appropriate state health agencies for going online with the fully reformed, single payer system in 2015.