Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner
Tools for Monitoring the Health Care Safety Net

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Presenting Information to Decisionmakers: A Guide For Policy Analysts

By Lawrence S. Lewin and Marion Ein Lewin


Policy Analysis, Politics, and Policy
The Policy Analysis Framework


One of the more interesting and stimulating tasks of health policy consultants is bridging the gap between health services research and public decisionmaking. In my many years in the consulting business, my colleagues and I were often asked to help States, cities, counties, health care providers, and Federal health officials grapple with issues related to expanding access to care, reorganizing existing services and resources to make them more effective, or developing a new program to broaden the reach of health care services. A major challenge in almost every case was translating the findings of health services research and policy analyses into potentially viable policy and program options for decisionmakers.

Expanding programs for uninsured and underserved populations is difficult, since these groups usually lack a strong political constituency. Elected officials and their staffs often must choose among competing demands for limited dollars. Further, unlike policy researchers whose reward structure is linked to statistically valid and scientifically defensible findings, decisionmakers must, by the nature of their jobs, emphasize relevance, timeliness, and answers to specific questions. Their need is to find a reasonable direction right now rather than a perfect answer later.

The effective use and translation of health services research for public policy decisionmaking requires a thorough understanding of the constraints and opportunities of the policy world. This kind of work requires strategic focus, sensitivity to the political preferences of key decisionmakers, and, as a critical first step, effective and compelling presentation of the data, analyses, and policy options. The analyses and arguments must be articulated in ways that can be easily understood and effectively communicated.

The availability of accurate and adequate data is usually a key requirement to completing these kinds of assignments successfully. However, all of the critical information is often not readily available. Without all the needed data, the process of analyzing problems and developing potential solutions may require the use of surrogates and "creative" analyses to describe the nature of the problem and how it might be addressed.

The purpose of this chapter is to provide policymakers with some of the tools and lessons my colleagues and I have learned from numerous consulting engagements on how to use and present key information and data to help build the "compelling case" that can help convince key decisionmakers to support a new and desired policy targeted to enhancing access for vulnerable populations.

Return to Contents

Policy Analysis, Politics, and Policy

Underlying this chapter is the notion that all policy decisions are, and appropriately should be, political decisions. In our system of government, disputes among parties with differing interests are usually resolved through a defined political process. The process is run by individuals elected by the public or by those they appoint. The challenge for a policy analyst, therefore, is to present to elected officials and major stakeholders the critical elements of policy options in a way that:

  1. Is clearly understandable.
  2. Illuminates the tradeoffs that may be involved.
  3. Can assist in educating the larger public as well.

Return to Contents

The Policy Analysis Framework

During the late 1970s, my colleagues Jack Needleman and David Helms and I developed a tool that has proven extremely helpful in assisting policymakers in thinking about the role of information in policy decisionmaking. The tool or framework, which was originally developed for the Agency for Healthcare Research and Quality's User Liaison Program (ULP), became known as the Policy Analysis Framework (PAF) (Figure 1, 22 KB).

The PAF provides a logical and sequential model for the key steps in the policy decisionmaking process. Although in practice the process rarely follows this logical sequence, the key elements can serve as a useful and relevant guide for action. A number of the PAF's key aspects deserve further explanation.

Analyzing the Problem

This is perhaps the single most important element. Its components include the following:

  • A precise and quantitative definition of who is affected, including both individuals and institutions.
  • An analysis of the financial and human consequences of not solving the problem.
  • The likelihood that the problem will either worsen or improve over time if little is done to address it.
  • A clear understanding of how different groups view the problem and its likely solution.
  • Where possible, an assessment of the factors causing the problem.

Failure to identify and achieve agreement on the problem to be solved often produces a chaotic process. Beginning with some consensus on the nature of the problem fosters more open and unbiased debate among the interested parties than if the process begins with arguments over solutions.

Assessing and Selecting Options

This is often a challenging but important task that is made much easier if the problem is well-defined. Assessing and selecting options can be facilitated by linking them with specific goals and objectives. Defining and assessing the strategies and resources that will be required to achieve the desired goals and objectives helps ground this important exercise in reality.

In assessing and selecting options, it is extremely helpful, whenever possible, to include some quantitative analyses to help predict the likely impact of each of the options. It is important that such analyses be presented in a way that is clearly understandable to the target audience. Wherever possible, analyses should reveal the expected impact on key stakeholders in a "winners and losers" format, not just the fiscal impact on the relevant agency. In some settings, such as the Federal legislative process, proposals for new or modified programs are required to be "scored" for their budget implications. The ability to accurately estimate the programmatic and financial consequences of proposed policy options in advance always contributes to a more credible and effective presentation.

Supporting and Evaluating the Policy Choice

While a full report is usually required, it is almost always preferable to supplement the written report with the three following devices to communicate with decisionmakers and the public:

  • A brief (5- to 10-page) Executive Summary.
  • An electronic slide presentation that is specific enough to serve as a stand-alone report (i.e., uses full sentences rather than phrases).
  • A press release that summarizes what the public and the program implementers need to know.

To ensure that the results of the policy analysis efforts reach the right people in time to impact key decisions, it is essential to understand how the process of presenting the evidence and the overall communications process will work. What parties and individuals will need to be involved and influenced, the scheduling of hearings and key meetings, the development of the budget and final policy decisions, and issues related to implementation are critical factors that need to be considered and decided.

Return to Contents


To make the Policy Analysis Framework more directly relevant, this section presents and describes eight of the graphics that were particularly useful in our work with policy and decisionmakers.

Example 1. Who are the Poor and Uninsured?

Figure 2a (11 KB) was originally developed for a ULP-sponsored workshop for State legislators to illustrate how Medicaid eligibility is a function of income and categorical eligibility, and how income and eligibility varied among the States. The chart is especially powerful as a means of illustrating the impact of different income and eligibility levels and how these are affected by various coverage options. It also clearly illustrates a point often misunderstood by elected officials, namely, that certain groups of individuals are not eligible for Medicaid, no matter how poor they are. The numerous Medicaid expansions in recent years have made this graphic more complicated. The version presented here does not incorporate all these changes, nor does it reflect the wide variations among States in their coverage policies. This graphic helped policy and decisionmakers (many for the first time) understand how Medicaid works, and provided not only a schema for understanding who in their State or region remained ineligible but also how their State or locale differed from others with respect to Medicaid coverage.

With some modifications, we were able to use this tool, which was designed for analyzing the problem, for planning and priority setting (i.e., assessing and selecting options) as well. By using Figure 2a to illustrate the "base case" of current Medicaid and other coverage programs, we were able to identify and categorize blocks of remaining uninsured persons by status and income (Figure 2b, 15 KB). By displaying estimates of the number of persons in each block and the actuarial cost of covering the various groups, we were able to encourage policymakers to begin discussing priorities for expanding coverage and their potential costs.

Example 2. Comparative Health Status Indicators

In association with our work on improving access to care for underserved people in a large urban area, we developed Figure 3 (4 KB) to help gain the attention of policymakers regarding the importance and urgency of addressing shortfalls in the public health and care delivery system in their communities. In this case, the chart compares the incidence of syphilis in St. Louis, MO, with the incidence in other metropolitan areas. In using this kind of chart, it is important to compare groups by ethnicity, insurance status, or place of residence whenever possible to ensure that comparisons are valid. This graphic tool helped local officials understand the need for greater efforts to combat socially transmitted diseases.

Example 3 Sources and Uses of Financing

When analyzing a problem, it is always useful and instructive for decisionmakers to know how current health care funding for the poor and uninsured is being allocated. In our work in St. Louis, this process began by showing the scope of coverage by payer source for the city and surrounding county (Figure 4a, 8 KB). Figure 4b (12 KB) shows the actual sources and uses of funds for the St. Louis region.

Since most program dollars are fungible, it is difficult to track the precise connection between sources and uses. Nevertheless, it is possible and instructive to illustrate (usually in the form of side-by-side pie charts) the sources and uses of funds for health care at the city, regional, or national level. If nothing else, constructing the pie charts requires a fairly rigorous and certainly useful process for identifying both sources and the levels of funding.

In many cases, it is helpful to construct these charts for several points in time. Doing so helps illustrate how the number of beneficiaries and the burden of financing change over time. We used similar charts for a city in California and were able to demonstrate that while Medicaid spending had increased as a source of funding in recent years, city and county payments for care of the uninsured had declined in real terms. For those working in the area of access, the ability to compare funding for health care by municipalities would be very helpful. Despite their value, unfortunately, these data are not readily available for comparative purposes.

Example 4. The Cost-Shifting Hydraulic

Many of the projects with which my colleagues and I were involved looked at the secondary effects of financing, particularly the scope and impact of cost shifting to cover losses from uninsured and publicly insured patients. One of the most creative and effective charts we developed is one that demonstrates graphically how cost shifting works (Figure 5, 10 KB). This chart shows how the costs of covering the self-pay population (uninsured and bad debt) are shifted to privately insured patients, and thereby passed on to employers purchasing coverage for their employees. To the extent that Medicare and Medicaid also pay at a level below average costs, these underpayments also increase the amount that must be recovered from other payers.

Using this chart, we conducted an analysis for a group of large manufacturers, and were able to demonstrate that expanding Medicaid to cover more uninsured individuals would cost these companies less (via a tax contribution) than what they were currently paying for the same group via the cost-shift to their premiums.

Another interesting and revealing message that the "hydraulic" tool can help illustrate is the impact of uncompensated care on incentives for hospitals to care for indigent patients. The higher the proportion of uninsured individuals, the more hospitals have to cost shift to recover these costs. The chart can be used to illustrate how hospitals and systems that care for underserved populations are at a disadvantage in price-competitive markets and thus have an incentive to move to the suburbs or to close or limit their emergency rooms.

Example 5. Patient Flow and Referral Pathways

In grappling with issues related to expanding health care for the poor and uninsured, it is important to consider various financing options but also to understand how these populations access care. For example, as an alternative to using the emergency department for primary care, it may be helpful to understand the extent to which there are other adequate options for the poor and uninsured for accessing primary and other needed care. Figure 6 (18 KB) describes the health care referral pathways within the city and county of St. Louis. Once these patterns were revealed and understood, there was a strong consensus to modify them. Specifically, this chart revealed the complexity of assessment and referral patterns, with virtually all specialty referrals having to go through ConnectCare, the program established for this purpose. The other, largely hospital-based institutions (BJC, Tenet, SSM, and Unity) as well as the patients chafed under the bureaucratic weight and delay of such a complex system. The graphic also helped to pinpoint the areas where the impediments to access and the obstacles to continuous care were the greatest. In the St. Louis example, as in many others, the analysis revealed that difficulty obtaining access to specialty care was a major problem.

Example 6. Unit Cost Comparisons

When considering alternative sources of care or facility and service consolidations, it is useful to have some sense of the economic implications of the potential options. This can be done by comparing the unit cost of services offered by alternative providers (Figure 7). There are at least three caveats that must be kept in mind when making these comparisons:

  • What matters is the cost to the purchasing program, which means the price the provider negotiates or can demand.
  • Care must be taken to ensure the comparisons are valid and not a comparison of apples and oranges. For example, when comparing unit cost per primary care visit, the denominator for hospital-based clinics is often the visit. In other cases, however, unit costs are the sum of the individual services provided. Thus, a clinic visit that includes an X-ray and lab tests would count as three units of service, whereas an identical visit to a hospital outpatient clinic will count as a single unit of service.
  • Care must also be taken to understand whether the comparison price reflects the full cost of service. For example, in selecting a prenatal care package, some competitive Medicaid managed care programs will select a program based on price that may exclude services previously covered such as drug counseling, day care, or transportation. The result of selecting the "lowest price" option may include the loss of services that have to be recreated and may result in higher overall cost to the Medicaid program.

For some of the reasons stated above, unit cost comparisons within a community can be difficult to obtain. Providers may resist disclosing them. Nevertheless, where they can be obtained, they can be useful and often revealing.

Example 7. Performance Comparisons

Ideally, purchasers and consumers of care would be able to select providers and health care plans on the basis of available, understandable, and reliable outcome or other performance measures. This ideal is, however, difficult to realize in today's health care system for several reasons:

  • Proven and effective outcome measures remain few in number, and many "outcomes" measures are actually process measures (e.g., immunizations given, timely prenatal visits, mammograms). Nevertheless, such measures can serve as useful measures of progress over time.
  • Data for developing outcome measures are often hard to come by, particularly for individuals who move on and off of coverage or seek services from a variety of providers without being a part of any organized system of care.
  • The transient nature of uninsured and Medicaid populations further compromises the validity of performance data.

Current efforts to improve quality and outcome measures will enhance the ability to compare the performance of local providers, health plans, and communities against desired standards or national objectives. These efforts will contribute not only as a way for communities to define the nature and severity of the problems to be addressed, but also for setting the goals against which local providers and health systems can be measured and compared. Figure 8 (14 KB), which was developed for a joint Ford-GM inquiry into the adequacy of community-based prenatal care, demonstrates how this city compares with others that are considered comparable. Such comparisons are useful in obtaining a sense of reasonable expectations and goals for improvement.

Example 8. Winners and Losers

One of the keys to "selling" a policy change is to illustrate clearly the winners and losers resulting from the proposed change. This usually consists of an analysis of who will gain access to care, either by insurance or direct services, and who is likely to pay for these changes.

An analysis of who would have been better off if the Clinton Health Security Act had passed is illustrated in Figure 9a (7 KB). The chart was part of The Lewin Group's Green Book analysis of the Clinton Health Security Plan. The analysis of who would lose (Figure 9b) tends to be more complicated because it usually involves reallocation of funding among Federal and State payers, private purchasers (via the cost-shift), and providers of care.

Return to Contents


The graphic tools presented here represent only a few examples of how such tools can assist in intelligent decisionmaking and can have a powerful impact on the eventual outcome of the policy decisionmaking process. Tools like the ones that have been described can help present complicated ideas and dynamics in an easily understandable and dynamic way. The use of graphic tools is most effective if they are integrated into the policy debate in a timely manner with all of the relevant players at the table. Although graphic tools cannot ensure the desired outcome, they have been proven to be a compelling strategy for getting the attention of key policy and decisionmakers.

Return to Contents
Return to Tools for Monitoring the Safety Net

Current as of December 2003


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care