A Dickensian duo of contrasting imperatives underlies the current management of atherosclerotic heart disease: to treat the anatomic stenosis causing myocardial ischemia and to prevent the plaque disruption causing myocardial infarction. The former leads us to do stress tests and angiograms to identify flow-limiting lesions and then to crush them by direct surgical intervention, while the latter leads us to identify risk factors, such as cholesterol, and prescribe drugs, such as statins, to mitigate those risks.

A host of clinical trials support each approach. In general, treatment is better suited to unstable coronary syndromes, and prevention is better suited to stable coronary disease. The recent COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial, for example, showed that percutaneous coronary intervention (PCI) did not improve outcomes when added to optimal medical therapy in patients with stable disease (2). It is uncertain, however, if these findings are applicable to older subjects, given their under-representation in the pivotal trials upon which current guidelines are based.

In this issue of the Journal, Afilalo et al. (3) shed new light on the effectiveness of statins in “elderly” patients with known coronary disease. In their meta-analysis of 9 trials enrolling patients between 65 and 82 years of age, statins reduced all-cause mortality by 22% over 5 years of follow-up. Rates of nonfatal infarction, stroke, and revascularization were reduced even more. Importantly, the magnitude of benefit was substantially larger than previously reported and increased with age, a 50% mortality reduction being observed in those over 80 years of age. Although the authors were unable to perform a formal assessment of safety, a recent review concluded that statins are “remarkably safe,” even in the elderly (4).

A limitation of this meta-analysis is the unavailability of patient-level data, which would have permitted an assessment according to pertinent clinical and demographic subgroups. Such details can be important, because patients in clinical trials differ from those in routine clinical practice. In addition, the time to onset of benefit and the effect of low-density lipoprotein (LDL) cholesterol (at baseline and on treatment) were not evaluated.

Nonetheless, using a previously published empirical survival model (5), the study has substantial practical relevance (Table 1). As suggested by Afilalo et al. (3), an 85-year-old man with known disease has a greater baseline risk (18.8% vs. 1.3%), and therefore obtains a greater benefit (8.9% vs. 0.4%) from statin treatment compared with an otherwise similar 55-year-old man. Thus, although the absolute gain in survival is lower in the 85-year-old (2.1 vs. 4.3 years), it is proportionately greater (75% vs. 23%) relative to baseline expectations of survival (2.8 vs. 18.8 years). All things being equal, then, the older we become, the more value we should ascribe to each additional year of life (and to the treatment that provides it).

Two practical problems, however, continue to plague the effectiveness of statin therapy in clinical practice. First, long-term adherence to therapy remains poor, averaging only 62% over 2 years (6- 10). Second, the so-called “treatment gap” (the proportion of statin candidates not receiving statin treatment) remains large, especially among the elderly. In one study, only 57% of “high-risk” patients attained LDL <100 mg/dl, and 18% attained LDL <70 mg/dl (11). Preferred drug lists and tiered formularies further compound this misutilization (12).

Meanwhile, the use of PCI continues to increase despite the lack of equivalent evidence of outcomes benefit. Current reimbursement policy actually encourages such misuse. Once drugs and devices are approved for marketing, physicians often use them in unapproved ways, and payers reimburse such “off-label” use to the same degree as “on-label” use.

Fine-tuning these financial incentives might help to close the treatment gap and increase adherence to statin therapy (13). Suppose the Center for Medicare and Medicaid Services (CMS) discounted a drug’s price, not by some fixed amount as it does now, but in direct proportion to its proven therapeutic benefit, so that one with more benefit is awarded a higher discount than another with less benefit. For example, imagine that one statin costing $2 per day (call it Positor) has a discount of 90% because it is proven to improve survival, whereas a competitive statin also costing $2 per day (call it Negator) has a discount of 10% because it only improves some surrogate marker. As a result, patients pay $6/month for Positor versus $54/month for Negator. Which of the two would you and your patient prefer?

Because the total price of the prescription does not change (with CMS picking up the difference), a high discount will not erode the manufacturer’s profits. In fact, profits are more likely to soar as a result of facilitated access to the less-costly drug and favorable shifts in market share. The CMS might even exploit this windfall to negotiate volume discounts with the manufacturer, thereby offsetting some of the cost of the plan. Although such negotiations are currently prohibited by existing law, the issue is under Congressional review. Thus, as a result of these incentives, the manufacturer is rewarded with competitive advantages for documenting the benefits of its drugs, and the patient is rewarded with better access to these benefits at more affordable prices.

Critics might argue that this plan requires a huge new bureaucracy. On the contrary, the needed infrastructure is largely in place. The Food and Drug Administration already calls on advisory panels to help decide if a drug meets standards of efficacy and safety. All it need do is empower these panels with the additional authority to set the discount for the drug based on the scientific evidence they are already reviewing. As new data become available, manufacturers could request a new review, just as they now petition for labeling changes.

Innovative strategies such as this are already being piloted in the private sector. Over the past decade, health plans have sought to save money by charging higher patient copayments for brand-name drugs to encourage the use of lower-cost generics. Now they are lowering copayments for drugs with proven preventive benefit (14). If patients no longer have to pay the cost of a statin, so goes the reasoning, they will more likely adhere to treatment, thereby forestalling higher-cost eventualities (stroke, reinfarction, revascularization). Aetna will conduct a formal evaluation of this new strategy later this year, by waiving copayments for statins (along with a few other agents) in post-infarction patients and tracking subsequent outcome and cost. From the patients’ perspective, this might be the closest thing there is to getting something for nothing.

The CMS could attach similar financial incentives to the performance of PCI by reimbursing approved “on-label” use at a higher level than for unapproved “off-label” use. Simply put, if payment drives utilization, and utilization drives outcome, payment can drive outcome. Unlike “pay-for-performance,” for which there is little proof of effectiveness (15), evidence-based reimbursement would: 1) target the expectation of benefit rather than actual outcome; 2) rely on empirical data rather than consensus opinion; 3) apply to individuals rather than groups; and 4) be large in size and immediate in impact.

If we are to pay for preventive care in proportion to its clinical value, consumers will need to be informed about these values. Currently, manufacturers are faced with substantial regulatory hurdles in attempting to provide truthful information to physicians and patients about the use of their products. Insurers, patient advocacy groups, and the Surgeon General’s office are not so constrained, and are free to inform the public truthfully about the value even of “off-label” uses. They should be encouraged to do so.

Although these are revolutionary proposals, so too was the original Medicare law. Its passage instantly assured the elderly of unfettered access to health care, but it also engendered a deep disconnect between what the providers of that care should do (according to the evidence), and what they are paid to do (according to reimbursement policies). The statin treatment gap is but one example.

This situation will not change unless and until we realign the financial and scientific incentives and begin rewarding caregivers, not for the prodigal provision of products and services, but for the enlightened provision of therapeutic benefit. Evidence-based reimbursement can be the bridge to this “far, far better thing” (1).