Health policy roundtable panel discussion: translating health insurance studies into policy proposals

Health Services Research, June, 2004 by Christina E. Folz

Kathryn Haslanger: We cannot just talk about coverage. We also need to discuss delivery service reform. It is time to do serious research to figure out how to change the current system, where people with complicated chronic conditions are being treated body part by body part.

Sherry Glied: How can researchers know whether to disseminate their findings rapidly on the Web, or more slowly through traditional peer-reviewed journals?

Jeanne Lambrew: It is hard to say. Researchers need to continually monitor their work against what is going on in the health policy world. They may find themselves in the midst of writing a paper when, all of the sudden, something happens that makes the findings immediately relevant, and they need to change their dissemination strategy.

Right now, I am working on a paper about Medicaid block grants that started when the block-grant debate was raging, shortly after the President issued a proposal to form a task force to consider this concept. The idea was to provide analysis to inform the debate on Medicaid reform. Well, the task force was recently disbanded, and we decided instead to go the peer-review route.

Richard Kronick: Your dissemination strategy really depends on your goals. If researchers want to effect near-term change on a piece of legislation, they obviously want to move quickly. But sometimes investigators want to change the way that we collectively think about a problem in order to drive policy change; for that, they may want to take their time and really hone their message.

Sherry Glied: Statistics and numbers play an important role in shaping policy. Does policy analysis without numbers matter? What do researchers gain and lose by quantifying?

JoAnn Lamphere: One of the things that researchers struggle with is figuring out what the "right number" is in a given debate, and how to get that number used. Policy analysts need to make their estimates based on the best available data, given a great deal of uncertainty about what is behind the current numbers and what the future holds in terms of the economy, etc. They must try to tease out what could reasonably be accomplished through a given program in the context of a changing environment and many unknowns.

Jeanne Lambrew: People underestimate the importance of numbers in health policy. My example is the current prescription drug debate. It began with a number--$400 billion, the amount of money dedicated to financing the plan over the next ten years. People think that that figure was based on analysis, but it was really just a number that fit into the budget resolution. The current Medicare drug plan was built around that number.

Thus, it was the CBO's assumptions that shaped the policy details, since Congress has had to back into a policy using its assumptions to meet the budget limit. Numerous assumptions are needed to determine a drug benefit's costs. For example, the CBO made precise estimates for the number of retirees that will be dropped from their retiree health benefits due to the drug benefits, and it incorporated into their financing the hypothesis that drug prices will rise once catastrophic care is covered for seniors. Even the decision to use a private risk-based system was partly based on an assumption the CBO made that private plans would produce deeper discounts than a non-risk-based system.