| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||
BOOK REVIEW |
University of California, Los Angeles Los Angeles, CA 90095-1772
Aging, Health, and Public Policy: Demographic and Economic Perspectives, edited by Linda J. Waite. Population Council, New York, 2004, 265 pp., $21.00 (paper).
Powerful Medicines: The Benefits, Risks, and Costs of Prescription Drugs, by Jerry Avorn. Alfred A. Knopf, New York, 2004, 448 pp., $18.15 (cloth), $10.20 (paper).
Almost everybody wants policy and practice to be based on the best science. Maximizing policy effectiveness and efficiency depend on an evidence base, and few would argue against guiding clinical decisions with state-of-the-art knowledge. Dominant philosophies of science posit that information that most closely reflects our experience in the real world will win out in the marketplace of ideas. Yet a persistent theme in the literature is that policy makers and practitioners often seem oblivious to the best science. As an article in The Gerontologist put it, "researchers are from Mars; policy makers are from Venus" (Feldman, Nadash, & Gursen, 2001, p. 312). In a cultural environment where intelligent design is promoted by some as having a legitimate place in science education, what is the proper role of empirical research in decision making?
Two recent books deal with these issues as they discuss the policy implications of state-of-the-art demographic research (Aging, Health, and Public Policy: Demographic and Economic Perspectives) and the implications of the misuse of information in prescription medication use (Powerful Medicines: The Benefits Risks, and Costs of Prescription Drugs). Both books are unusual in that they include first-person accounts of how the respective authors came to their research questions, how their findings led to other lines of research, and how the context (funding, politics, and peers) shaped the type of research that was possible for them.
Aging, Health, and Public Policy, edited by Linda J. Waite, is a collection of articles that originally appeared in Population and Development Review. They are written by leading researchers in the demography of aging and provide interesting insights into the process of research, such as how the Wisconsin Longitudinal Study developed out of what was intended to be a cross-sectional study of high school seniors in the late 1950s. Several chapters provide summaries of innovative demographic methods and information about new demographic findings that can inform governmental budget forecasts. Other chapters provide syntheses of demographic research about the pathways that link disparities in socioeconomic status to unequal health status. An early chapter offers a fascinating review of historical demography that documents surprisingly high rates of chronic conditions in the 1800s and shows that in more recent times the onset of those conditions has been pushed into ever-older ages. Combined with a chapter that tests, and refutes, James Fries's well-known theory of a maximum potential lifespan, the research in this volume suggests that policy makers should look to a continued growth in the longevity and good health of the older population. These chapters provide good coverage of the issues, are logically related, and are written in a manner that is understandable to nondemographers. Any academic interested in learning about the leading edge in the demography of aging will find this volume useful.
Powerful Medicines: The Benefits, Risks, and Costs of Prescription Drugs, by Jerry Avorn, successfully targets the wider audience of educated, but not necessarily academic, readers. While the Waite book emphasizes new developments in methodology, the Avorn book includes a focus on practice—the practice of pharmaceutical research, government oversight, marketing, and physician prescribing. This book is written in an engaging essay style that mixes research findings, stories (case studies), and commentary, with the documentation tucked away at the end. The information is particularly credible and interesting because the author was involved in many of the studies and incidents that he reports in the book. So, as with several of the chapters in the Waite book, the reader learns about the process of the research. The Avorn volume delves into the politics and economics behind pharmaceuticals and provides more details about some of the trials and tribulations behind his research. The final chapters of the book contain a clear set of policy recommendations to improve the clinical and cost effectiveness of prescription drugs.
The Importance of Data
Both books suggest that information is power. Avorn makes a forceful argument for clinical trial data on the comparative benefits of new medications versus existing treatments. He explains that the Food and Drug Administration (FDA) only requires that new medications show effectiveness in comparison to an inert placebo. For example, expensive and heavily promoted new brand-name calcium channel blockers were required to have some real benefit in reducing blood pressure to obtain FDA approval. But the medications were not required to work better (which they do not) in uncomplicated cases than older and much cheaper generic diuretics. The result is a new class of medications that is widely used, significantly increases costs, and offers no commensurate improvement in clinical outcomes for most persons. Avorn also notes that safety data on new pharmaceuticals are usually limited to findings from short clinical trials in selected populations and that population-level data needs to be collected and analyzed to identify all dangerous side effects and to determine if the benefits outweigh the risks. Vioxx is one of the more recent examples of how the hype used to introduce a new high-profit drug results in the widespread use of a medication in circumstances in which it provides little or no clinical advantage to older and less expensive treatments, while its side effects in a general population of users are clearly worse.
But even when data are available, they are not necessarily used. Avorn despairs that some clearly efficacious medications are not being sufficiently used, such as cholesterol-lowering drugs, anticoagulants for those with atrial fibrillations, and Fosamax to reduce hip fractures. Avorn identifies the pharmaceutical industry as using raw political power and distorting information in its marketing to maximize profits, even at the expense of patient safety. In a sea of promotional marketing, important but boring facts are likely to get overlooked. Most readers are likely to know that drug companies sponsor lavish "educational" seminars to push their latest products and spend heavily on direct-to-consumer ads. Fewer are likely to know that doctors who are heavy prescribers may be paid as "experts"—a thinly veiled type of kickback. And even fewer are likely to know that some articles that appear in academic journals extolling the virtues of new drugs are ghost written by the manufacturer that pays the academic "author" to submit the piece under the academic's name (see Mathews, 2005).
We can breathe a sigh of relief that there is no equivalent to the drug industry that distorts demographic data of the type produced by the authors in the Waite volume. Their work suggests how to improve population projections and projected retirement rates, both key issues in Social Security policy. Does that mean that policy makers will use the demographic work of several of the authors in the Waite volume? Not necessarily.
Academics, including the authors of both books, generally see the adoption of knowledge and the diffusion of innovation as rational processes involving a marketplace of ideas. In that marketplace we systematically compare the utility of different theories in accurately predicting or explaining the empirical world. We "look under the hood" to see if the assumptions and methods are reasonable, and we use a peer-review process to make sure that new findings are valid and useful. Truth is the primary value (even as we acknowledge that what constitutes truth and who has the power to define truth is contested). Although this may roughly approximate reality in academic settings, the policy world has different priorities. Politicians are often affected by ideological blinders, economic pressures (both in governmental budgets and in their own campaign coffers), electoral realities, bureaucratic inertia, and a host of other factors that can make good data irrelevant (Black, 2001). Just as the incentives in industry can be to distort scientific data for the sake of profits, the incentives in the political realm can be to selectively use only the data that support a predetermined position.
Disseminating Science
What can be done to bring more science to policy and practice? As a physician, Avorn's solutions mostly rely on the goodwill of physicians. He explains that physicians are disconnected from pharmacologists, so their prescribing practices are susceptible to fads and ads that may or may not have much grounding in science. If doctors had authoritative data on the risks, benefits, and costs of different medications, he argues, clinical decisions would be better. As evidence of the effectiveness of facts placed in the right hands he presents examples from Australia, Britain, and Canada. Australia has recently developed a National Prescribing Service (NPS) that commissions and disseminates evidence-based reviews of all drugs, including information on cost effectiveness. This seems like a good idea, but in the United States politically powerful forces have reacted against data produced by agencies similar to the NPS. The best-known example is when the surgeon's lobby was nearly successful in defunding what is now the Agency for Healthcare Research and Quality (AHRQ) because the agency released an evidence-based report that found that surgery was not useful in most cases of lower back pain (see Gray, Gusmano, & Collins, 2003). Perhaps in Australia medical professional and commercial interests are not as politically powerful as in the United States.
One dissemination strategy Avorn promotes is "academic detailing," a scientific equivalent to drug salespersons who make office visits to disseminate information. He has conducted research that shows this type of face-to-face dissemination changes prescribing practices of physicians. What is missing from his discussion is evaluation data on the impact of the countrywide implementation of the program in Australia, and the extent to which commercial marketing practices there are similar to those in the United States. Even good data personally placed in the hands of physicians, for example, may not be able to overcome patient demands generated by direct-to-consumer advertising. Avorn does not suggest ending direct-to-patient marketing, which he concludes is here to stay because of the power of the drug industry.
We know policy makers can be immune to facts because of political concerns. But isn't the right information in a clinician's hands enough to revolutionize practice? Avorn is not Pollyannaish about this. He summarizes the Ibsen play, An Enemy of the People, as a parable of how well-meaning medical science can be trumped by self-serving political and commercial interests. In his typical prose style he notes that "I feel a warm glow inside each time I reread this play. It's not pride—it's acid reflux" (p. 341). He then continues with several real-world examples that support this parable. What he fails to do is offer advice for how to reduce the power of self-serving political and economic institutions so that scientific findings benefit the general public good rather than particularistic material gain.
Independent of the deliberate misrepresentation of data, scientific data may not be used because appropriate data are not readily available to answer the relevant policy questions. To make data relevant for policy makers, it is useful to think about the policy process in distinct phases because each phase may use different data. One phase is defining the problem and setting the agenda. Policy makers focus on a limited number of issues at a time, and scientific information can be useful in getting an issue on the active policy agenda. Useful information describes the magnitude of the problem, the characteristics of those affected by it, and the costs (human and economic) of not addressing the problem. The way in which information is presented and argued helps "frame" the problem, identifying which are legitimate aspects of the issue and which are irrelevant.
When policy makers (or practitioners) acknowledge that there is a problem, they then have to consider a limited number of solutions. The range of possible solution is limited by the way that the problem is framed. In this part of the policy process data are important in evaluating the relative effectiveness of different solutions, establishing costs, and identifying how different groups might be affected by the solution.
After a policy is adopted, evaluation data can be useful in pinpointing failures in implementation, unintended side effects, and monitoring the movement towards achieving the policy goal. Once again, the framing of the problem will direct the policy evaluation to consider a specific set of outcomes to identify success.
Competing Frameworks
The new Medicare Part D prescription drug benefit is a good case study of different ways that an issue can be framed and how the framing drives data needs. It also shows how politics and economics shape debates and put boundaries around what information is considered relevant to an issue. For many years there have been complaints that Medicare did not cover outpatient prescription drugs. The problem was defined as one of insurance. The data marshaled to show that there was a problem included the number of older persons without prescription drug coverage along with the out-of-pocket costs that they incurred. Limiting the framing of the issue to lack of insurance coverage for outpatient medications attracts the strong support of the insurance industry (as long as the benefit is administered by private companies) and the acquiescence of the pharmaceutical industry (who are happy to have an increased demand as long as price controls or competitive bidding are not involved). An insurance problem logically leads to an insurance solution. The policy debate about solutions included whether Medicare Part D should offer catastrophic insurance for a few or provide some level of coverage to most Medicare recipients (it ended up doing both, with a sizeable gap between when the general coverage stops and the catastrophic coverage starts). The facts relevant here were those that would answer the questions: What proportion of older people would be covered, and how much would the program would cost the government (both important political issues)? Data on how the program could be structured to best improve access, equity, or health outcomes did not enter into the political debate. And data seemed irrelevant to the ideology underlying the policy debate about whether the solution should use private insurers in a lightly regulated market or a public agency more akin to Medicare Part A. Few voices contested the definitions of the problem as one of insurance.
There are at least three alternative frameworks that have been marginalized that would also address the problem of the costs of prescription drugs—reducing drug costs, increasing elders' incomes, and reducing the need for drugs.
Reducing drug costs is the most frequently discussed competing frame. A modest amount of research documents the rapidly rising costs of prescription medications, but less research offers solutions. One solution that has generated a lot of political debate is reimporting prescription drugs from Canada. Drugs cost less in Canada because their federal government maintains price controls over medications, a regulatory strategy that has not received any serious debate in the United States. Avorn mentions drug reimportation and price controls but saves most of his critique of drug prices for the way that companies produce brand-name drugs based on patented formulations, which often are different in only minor ways from off-patent drugs, and sell at exorbitant prices. He argues that drug company marketing, combined with limited independent information and pressures on physicians, leads to unnecessarily high costs for treating many common chronic conditions. His solution is to disseminate information on the cost effectiveness of all drugs and rely on the voluntary efforts of beneficent doctors to select the most cost-effective treatments. Avorn avoids suggesting mandatory regulations that new medications be cost effective in addition to the current requirement of safety and clinical effectiveness. He also shies away from proposals to let Medicare bargain over prices with drug manufacturers in the way the Veterans Administration does. Putting the buying power of 43 million Americans into a single pool would result in better prices, a situation that is the pharmaceutical industry's worst nightmare.
Two competing frameworks that are not commonly discussed in the prescription drug debates are increasing the incomes of older persons so that they can afford their medications and/or reducing the need for prescription drugs. The group most severely affected by prescription drug costs is older persons who have too much money to qualify for Medicaid and not enough to pay for their minimum daily living needs. Those with incomes between 100% and 199% of the federal poverty level comprise more than one quarter of the total elderly population and account for almost three times as many elders as those in poverty (Federal Interagency Forum on Aging Related Statistics, 2004). One of the recommendations from the 2005 White House Conference on Aging focused on this issue, suggesting that eligibility for Supplemental Security Income (SSI) benefits be raised above the poverty level to as much as 175% poverty. Given the blind eye that the media (and, ironically, the White House) turned towards this latest White House Conference, it is unlikely that this recommendation for improving the economic situation of older Americans will receive much policy attention, regardless of how much scientific information we muster to support the recommended solution.
Finally, the incentives and training in contemporary medicine are primarily focused on drugs and surgery. Avorn complains that there are few comparisons of the effectiveness of new drugs in direct comparison to existing medications, but there are even fewer studies comparing drug treatment to other treatments. If we reduce the need for prescription drugs, especially those expensive drugs taken for many years for conditions such as hypertension and diabetes, then insurance coverage becomes less of a problem. But pharmaceuticals are a convenient and relatively quick treatment to offer and one that physicians have exclusive rights to supply. Nonetheless, changes in physical activity, diet, and other interventions can be as effective, or more effective, than drugs for some common chronic diseases (e.g., in relation to diabetes, see Knowler et al., 2002). In addition, focusing on entire populations and changing the environmental conditions that those populations face (e.g., improving the availability of nutritious food, designing communities to encourage walking, etc.) can improve the risk factors for large numbers of elders without recourse to medicating all of them (Wallace, 2005). But no company profits enough from promoting walking among older persons to generate the type of promotional activity that we see from drug companies who are promoting the latest expensive antihypertensive medication. It would be interesting, however, to conduct a cost analysis of creating safe and attractive walking spaces in urban areas and compare that to paying for years of hypertension and diabetes treatment. Would this scientific information change the debate? Not by itself, but politicians are attuned to how policies impact budgets, and having cost data available in any debate strengthens the argument.
Reorienting Initiatives
Knowing that data can be an important tool in the effort to frame problems and put them on the policy agenda, promote adequate solutions, and monitor implementation, what can we do to increase the prospects that scientific evidence is used to inform policy and practice? First, we must assure that a steady stream of well-conducted scientific findings and data continue to be produced. The Waite volume offers a roadmap for continuing to develop policy relevant demographic research, and the Avorn book suggests future directions in pharmaceutical epidemiology. It is also crucial to work against reductions in data collection, such as proposals to prohibit collecting data by race and ethnicity, that would make it impossible to marshal the facts on an issue or solution.
Second, data do not speak for themselves, so they need credible and forceful advocates. Researchers in universities and practitioners in well-regarded practice locations bring important credibility when discussing scientific information, and they have a responsibility to communicate their knowledge not only to their students and clients but also to policy makers and the general public. Scientific organizations like The Gerontological Society of America should to be at Congressional hearings to present authoritative information that helps inject some facts into policy debates.
Third, because facts do not necessarily carry an argument in the face of powerful political and economic interests, it is also necessary to restructure the incentives in the system to give scientific evidence a leading role in policy and practice decisions. If relying on evidence-based policy and practice pays (financially or for electoral purposes), politicians and businesses will be more likely to follow the evidence and encourage others to do the same. Changing these incentives are essential to improve the chances that the state of the art in demography and pharmaceutical epidemiology will have an impact on policy and practice.
References
| ||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
