Consumer Financial Incentives: A Decision Guide for Purchasers
There has been some research into the characteristics that make a consumer likely to respond to the information provided in an incentive plan (Table 5). When the aim is to encourage the selection of high performance health plans or providers, an important issue is the extent to which the information about the performance of providers or plans offered through the incentive program is considered new information.5
Studies have shown that sometimes consumers already have an informal sense of health plans' or providers' performance, and that the addition of a report card merely confirms those impressions. In such a case, a new report card may have little impact on consumers' behavior.4-5 One implication of this finding is that consumers who are new to a market, and have no prior information about providers or plans, may be particularly likely to respond to information—and also to incentives.4,8,115
For all the approaches to establishing incentives discussed in this Guide, a major factor determining whether people become engaged in, and effective managers of, their health care decisions is the extent to which they are "activated consumers." Hibbard and colleagues116 have developed a tool to measure consumer activation—the general concept is that a consumer needs to have the confidence and knowledge to step into the decisionmaking role.
Patients who are female, younger, and better educated are more likely to be activated consumers, but these readily measured variables account for relatively little of the variation in activation.117 Rather, personal behaviors like asking questions and reading medication labels are better markers of an activated consumer.
Some patients may not want to be activated consumers. In fact, a national survey found that some people—although they are happy to discuss options with their physicians—actually do not want to be decision makers in their own health care.118
Patients who are elderly or very sick—those who have the highest health care costs—are less likely to want to make their own decisions. In terms of using information for reducing disparities in health care delivery, it also was found that African American and Hispanic patients were less likely than others to prefer an active role in decisionmaking.118 Consequently, consumer responses to incentives are likely to vary widely, both among people within a group and among groups. Special efforts will be needed to ensure that all patients benefit from performance measurement and incentive programs.
Question 18. What special accommodations, if any, should be made for lower income, underserved, or sicker consumers?
Socioeconomic factors and health status clearly have an impact on consumers' responses to incentives. This has been most carefully studied with respect to consumer incentives to use fewer or less expensive drugs. It has been shown that even among patients with chronic diseases, patients with a lower income and those who are sicker are more likely to stop medications because of cost.57
The impact of income and health status on the response to incentives can be mitigated through the design of the incentive program. If, as in the example of Aetna's HealthFund program, first-dollar coverage for certain necessary care is part of the benefit design—for example, people with diabetes in HealthFund get their diabetes medications at no cost—so then there is much less reason to be concerned about the negative impact of cost-sharing.26
Whatever incentive approach is used, it will be important to consider the possible impact of limited health literacy. Descriptions of the program should be sufficiently simple that consumers can understand them, even if their educational level is low. If this is not achieved, large segments of the population may fail to understand how the incentive functions and therefore cannot be expected to respond as desired.
A relatively novel approach to assisting patients with their health care and health care decisionmaking is the use of "patient navigators"—health care professionals who help patients navigate the complex health care system and the myriad potential barriers to accessing high quality care. Such barriers can range from a patient's mistrust of providers, to a patient's lack of child care, to cultural barriers.119
An example of this approach is the National Cancer Institute's program to use patient navigators to address disparities in care affecting underserved populations.120 The results of the studies funded under this program should be available soon and should help us understand when and how navigators are needed and what navigation assistance is most important.
Question 19. Is there a role for consumer financial incentives in an overarching disparities-reduction strategy?
There are at least two ways in which introducing incentive programs also could reduce disparities in health care delivery. The first is the simple act of disseminating information about the quality of providers' health care performance to populations negatively affected by disparities. Research has shown that, before providers' report cards are released, minority groups seem to have less access to information that defines which providers perform better based on the available, measurable quality indicators.
Before the institution of public reporting of surgeon-specific rates of mortality from bypass surgery in New York State, there was no relationship between surgeons' mortality rates and the probability that an African American patient would choose a particular surgeon. Among white patients, however, those from ZIP codes associated with a high education level and a high or middle income level were more likely to choose a surgeon whose surgical mortality rate was low. After 1 year of public reporting, African American patients from ZIP codes with a high education and a high or middle income level were also more likely to choose a surgeon who had a low mortality rate.5
The second way incentive programs could reduce disparities is that information about disparities can be included in the set of measurements collected and reported.68 For example, patient experience (patient satisfaction) scores for hospitals could be reported by race and ethnicity. This would give providers an incentive to reduce disparities and provide culturally sensitive care, while helping minority-group consumers identify hospitals in which they would be most comfortable receiving care.
In addition to the hoped-for effects of an incentive program, purchasers will need to monitor, and try to minimize, unplanned negative consequences. Earlier, we described potential, unintended responses from consumers, especially skipping or delaying important treatment to avoid out-of-pocket costs. There also may be important unintended consequences in terms of providers' responses to tiering, which by definition includes public reporting of quality ratings, and the concern those reports may raise. Three unintended consequences to look for are: providers' selection of patients, "cherry picking" the healthiest patients; diversion of attention away from important aspects of care that are not measured in quality ratings; and widening gaps in performance among providers.
- Selection of patients. Providers may avoid treating sicker patients in the belief that adjustments made for severity of illness in quality ratings are not adequate and that caring for such patients will reduce their measured performance. Surveys done after New York instituted public reporting for coronary bypass surgery showed that two thirds of cardiac surgeons admitted to avoiding referrals of the most severely ill patients.121 One approach that might reduce the probability that an incentive program would experience this problem would be to include, among the performance measures, some structural or process measures of quality that apply equally to all patients, regardless of their severity of illness. Risk adjustment of outcome measures like mortality rates will also minimize selection incentives, as long as providers believe the risk adjustment is adequate. In addition, including explicit reporting of case-mix data that show which providers are avoiding or accepting the more difficult cases—or providing differential rewards for meeting performance goals with more difficult patients—might increase providers' willingness to take on those cases. Another possibility would be to collect and report information about patients who change from one provider to another. A provider who is avoiding sicker patients would be identified by the high case-mix scores of patients leaving his practice.69
- Diverting attention from aspects of care not included in quality ratings. Incentive programs may focus providers' attention on the aspects of care for which there are quality performance measurements, to the detriment of performance in other areas.122 This potential problem highlights the importance of selecting measures judiciously and of paying attention to interrelationships among targeted and untargeted domains of performance. Using some broader measures of outcome, such as patients' experiences or decubitus ulcer (bed sore) rates and pain scores in hospitals, may mitigate this problem as well.
- Widening performance gaps among providers. This problem is most likely to occur if a program is designed to reward only providers that meet a high standard of performance or that are the highest ranked among peers. If this approach takes substantial resources away from other providers, their performance may actually get worse. The problem is of particular concern if it has an impact on safety-net providers and/or if there are not enough alternative options for those patients who receive care from providers with poor performance. If these adverse consequences are anticipated or noted, purchasers can adopt auxiliary programs to help safety-net providers improve their performance.
Assessing the impact of a consumer incentive program is challenging because so many other factors simultaneously affect the quality and cost of patient care. Ideally, purchasers would implement the incentive program in one market or submarket and track the same performance measures on a set of comparison providers in another area. Some large employers, the Federal Centers for Medicare & Medicaid Services, and state Medicaid programs may be in a position to pilot consumer incentives in this way, but most purchasers cannot set up their programs as controlled trials. Therefore, special effort is needed to disentangle the effects of the program from other trends.
At a minimum, purchasers using strategies that target provider or plan choice should collect baseline data on the targeted performance measures before the program begins. This will be a critical part of program implementation because consumers and plans or providers need to learn about the measures and current level of performance. Baseline data about provider or plan market share also must be obtained. As the program is implemented, its effects can be evaluated in terms of the change in performance and market share for high and low performance providers, preferably relative either to a comparable but unaffected population or to the trend in performance and market share existing before the program's implementation.
For programs targeting selection of treatment options or reducing health risks, the key baseline data relate primarily to consumers' choices among treatment options or health risk behaviors. Assessment of the program's impact, then, involves re-measuring the baseline variables to determine the magnitude of change achieved.
To understand the impact of any type of consumer financial incentive program, consumers and providers can be surveyed for feedback about unexpected problems with the measures used, including difficulties with access to care. Similarly, purchasers can track a set of performance indicators that are outside of the incentive program to better understand both negative and positive spillover effects from the program on untargeted clinical domains. Evaluation of the program also can include assessing not just average performance but also the effects of the program on different parts of the delivery system, including patients from low and high income levels and providers with high and low baseline performance ratings.
The Agency for Healthcare Research and Quality (AHRQ) and the Commonwealth Fund have recently collaborated on establishing priorities for research into the impact of consumer-oriented programs on clinical outcomes.
Purchasers have to decide how rigorous an evaluation needs to be in order to ascertain whether a program is working and how to improve it. To adhere strictly to scientific standards of evidence may be too costly and may produce evidence too late to be useful for decisionmaking—but erroneous conclusions that may be drawn from anecdotal or incomplete information may have substantial costs as well.