Legislators, policy wonks, and Team Obama need a serious set of online decision-support tools (DSTs) to understand, compare, and evaluate the myriad and conflicting proposals to reform health care. In mid-January, the RAND Corporation entered the fray, busily promoting to the media and legislative staff what it calls a global positioning system to comprehend the effects and unintended consequences of competing reform proposals. Called COMPARE (Comprehensive Assessment of Reform Efforts), this online toolkit (www.RANDCOMPARE.org ) allows health care stakeholders to kick the tires of various policy options before implementation.
Explore This Issue
May 2009RANDCOMPARE is essentially an enhanced business dashboard-a DST to communicate complex information quickly by using maps, charts, and other graphics to show multiple results together (see Effective Health Care Dashboards). For example, it compares how policy options such as individual mandates, employer mandates, purchasing pools, or refundable tax credits would affect aggregate spending, consumer financial risk, the patient’s experience, systemic waste, health coverage, capacity, and operational feasibility.
Elizabeth McGlynn, PhD, Associate Director of RAND Health and COMPARE’s Co-Director, said the tool will help make realistic assessments, based on objective standards and facts, and therefore encourage a full understanding of the immensely complicated benefits and risks of the policy choices ahead. She noted that policy options selected for the dashboard and their underlying assumptions were based on RAND’s 26-member advisory committee, who looked at 25 years of health policy proposals and reviewed relevant literature. When asked about two glaring omissions, innovative health plans stripped of onerous mandates and allowing the selling of plans across state lines, she said, We didn’t include them because we made a decision to launch and will eventually add other policy options over time.
Sreedhar Potarazu, MD, MBA, Chairman and CEO of Vital Spring Technologies, which develops software for corporations and government agencies to manage health care benefits’ costs and quality, contends that approaches such as RANDCOMPARE are likely to fail because of a disconnect between trend-spotting and reality. What’s missing in mandates is that someone has to pay for them, and our data aren’t sophisticated enough to put accurate price tags on medical services, he said. These things always cost more than their estimates because we need three key metrics-price, utilization, and the likelihood of consumers using the resources, he added.
Even though there are truckloads of data, they are scattered and incomplete. To analyze health care reform proposals options correctly, Dr. Potarazu suggests investing in the right IT infrastructure to crunch data for the most common medical conditions and providing patients with the right incentives to consume medical standards, and urges providers to maintain standardized, consistent data sets. Consumer-directed health plans are a good start to accurate pricing because people know how much they have in their accounts to spend, what they’re going to pay, and how good decision-making can improve health and lower costs.
Giving more ammunition to the argument that getting a handle on costs before the sweeping cart-before-the-horse approach favored by RANDCOMPARE’S policy changes, Winifred Hayes, MS, PhD, RN, and CEO of Hayes, Inc., a health technology assessment and consulting firm, uses evidence-based assessments of health technologies to help providers and payers calculate costs and benefits. There is ongoing debate about the value of many technologies. Having a DaVinci robot, for example, may give a hospital a marketing edge and impress physicians and consumers, but the technology is very expensive, and there’s no solid evidence of better outcomes, she said.
Similarly, while Loma Linda University Medical Center and others have invested in proton radiation therapy in addition to having a gamma knife, putting a price tag on proton technology’s higher accuracy and fewer side effects is a major decision. Dr. Hayes also pointed out that an unintended consequence of government-sponsored or mandated coverage is that it will undermine employer-sponsored coverage, which currently provides 61% of all health insurance.
Objectivity Questioned
Before the legislative gofers and the media who are actively exploring health care policy navigate heavily with RANDCOMPARE, they should carefully scrutinize the dashboard (see A Close Look at One Policy Option). Dr. McGlynn’s explanation that important policy options weren’t included in COMPARE’s launch but will be included later indicates a certain haste in pushing the product out the door. She also mentioned that a National Health Exchange similar to Massachusetts’ Connector would be a national rather than a state plan. Similarly, whereas RAND’s dashboard claims there is no evidence on the impact of purchasing pools on spending, the National Association of State Comprehensive Health Insurance Plans (NASCHIP) provides a plethora of data on purchasing pools, different state models, and what ideal plans cost and look like.
What’s missing in mandates is that someone has to pay for them, and our data aren’t sophisticated enough to put accurate price tags on medical services.
-Sreedhar Potarazu, MD, MBA
While RANDCOMPARE deals with health care’s growth in spending that annually outpaces inflation consistently, it identifies widespread systemic waste in administrative, operational, and clinical areas. It hints that switching to a Canadian-style system might cut waste substantially. However, it never addresses the systemic waste caused by state mandates that cover chiropractic, acupuncture, massage therapy, wigs, mental health parity, drug and alcohol rehabilitation, in vitro fertilization, and other services that politicians find difficult to oppose for fear of alienating motivated voters.
Effective Health Care Dashboards
Navigating the increasingly complex health care system requires a set of tools that smooth the way. Dashboards are a subset of decision-support tools that can help health care administrators and providers filter the staggering amount of data bombarding them.
Although dashboards are intuitive and user-friendly, they are only as good as their underlying assumptions and data. To be effective a dashboard:
- Has a balanced set of performance measures.
- Selects a manageable set of measures (15-30).
- Presents data in graphic displays.
- Highlights action triggers.
Health care dashboard categories usually measure:
- Financial performance.
- Operational effectiveness/efficiency.
- Quality.
- Patient satisfaction.
A Close Look at One Policy Option
Closely examining the RAND dashboard for the policy option labeled individual mandate reveals questionable assumptions and outright disregard for the actual performance of the largest experiment with the individual mandate, the Massachusetts plan enacted in 2006.
RANDCOMPARE said that individual mandates will have no effect on spending in the aggregate, which will increase by only $7 billion to $26 billion, or 0.3% to 1.2% of total national spending, indistinguishable from zero. Tell that to the people of Massachusetts, who faced a $147 million shortfall in the program’s first year and a $130 million deficit halfway through the second year.
The mandate’s price tag in 2006 dollars-$3500 and $10,000 for individual and family insurance, respectively, including deductibles of $2000 and $4000, respectively-contradicts RAND’s estimate that individual mandates show no discernable change in consumer financial risk for the non-elderly. The analysis contradicts itself by adding: the median proportion of income spent on health care increases substantially among those who become newly insured (emphasis added).
Moving along the dashboard, RANDCOMPARE makes these predictions for the next four variables: reliability (no effect), patient experience (improve), health (improve), and coverage (increase). The underlying assumption-that having health insurance will improve access to care-is accepted, although no empirical evidence exists. It boldly continues, projecting nationally, that an estimated four million life years will occur and the mandate will increase coverage by 9-14 million. It assumes that there will be no effect on the health care system’s capacity, without indicating who will care for those additional millions of patients. On operational feasibility, it concludes, difficult, because determining compliance with the mandate and enforcing penalties for noncompliance are large tasks.
Since one of RANDCOMPARE’s stated goals is to detect the unintended consequences of policy options, it should have reported that Massachusetts’ two-year experience with the individual mandate has increased costs for individuals and the state, reduced revenues for doctors and hospitals, and decreased access in a state with an abundance of physicians. As of the end of 2008, state costs rose more than $400 million (85% above projections) and new patients waited an average of three months to see a doctor. This is not for lack of doctors but for physician unwillingness to accept the Connector’s pay scale, and lose money every time they see a new patient.
Patients and doctors are now trapped in a vicious circle. In 2009, insurance prices for individuals will rise 10% to 12%, twice the national average. Costs to Massachusetts continue to rise; the state has asked the federal government to help make up the shortfall of hundreds of millions of dollars. The state is also considering raising taxes and requiring businesses, hospitals, and insurers to pay more to fund the program. Meanwhile, the state plans to slash payments to doctors and hospitals by 3% to 5%, making it that much harder for patients to find a physician willing to see them.
One hates to contemplate: If RANDCOMPARE finds that an individual mandate is the most cost-effective strategy for decreasing the number of uninsured, what are some of the really expensive strategies?
-Marlene Piturro, PhD, MBA
Contact the Editor
ENToday welcomes your feedback about our publication! Contact us using the information below:
©2009 The Triological Society