Leaders in healthcare organizations struggle to match continuously expanding medical services and technologies with limited resources to pay for them. Whether one is involved with paying for care through insurance benefits or delivering care as a provider, we make these difficult choices every day. I have seen them made in a variety of ways. Most commonly, one continues what was done the year before, adding or subtracting at the margin according to some combination of new evidence, new technologies, patient or consumer demands, inertia, and internal politics. Occasionally, the next increment of care is subjected to rigorous analysis within the limits of the data. Rarely (in my experience, never) is the full complement of practices, current and potential, compared against a standard metric to determine which should continue. Most organizations, inside and outside health care, have difficulty stepping back from what they do in order to subject all their work to rigorous comparative analyses to determine what stays and what goes. Lack of good data and common agreement about how to compare services, especially in health care, makes this analysis challenging. Even more problematic are the challenges of organizational disruption when longstanding services, the people who provide them, and the people who get them face the cutting room floor because a new technology or a new service has been found to be more suited to the organization’s purposes. This situation is especially true in health care, in which staff and patients form alliances that create formidable barriers to reprioritization, regardless of how rational and well documented the grounds for doing so might be. Consider, for example, how seldom there have been successful consolidations of hospitals in spite of numerous efforts to do so across the country during the past decade. But let us assume, for the moment, that a healthcare organization has the will and the skills required to start from scratch to determine what benefits or what service will be provided going forward. Several barriers remain. First, our evidence base for population effectiveness and cost benefit is limited. We are making headway as we develop the computing power required to gather and massage large databases and to track individuals longitudinally. But we have a long way to go before we can establish airtight cases for one service over another across the spectrum of healthcare interventions. Second, we do not have agreement about what we are trying to measure as benefit. We lack a common language for comparing alternatives. Clinicians operate within the patient-centric Hippocratic tradition in which the overriding measure is whether or not an intervention helps the individual patient compared with alternative interventions for the same problem. The concern is not how it compares with another intervention for another segment of the population with a different potential disease burden. To compare alternative interventions for a population requires us to develop a language that reflects a communitarian or population-based ethic. Without the language—the framework imbedded in our decisionmaking processes—we have nothing to guide us toward sound, ethical decisions that are in the best interests of the most people within a community or population. Even if we could operate with a population-based focus, we struggle with how to define the “greatest good.” Is it reduction in mortality? Morbidity? Is it appropriateness of resource use to achieve the effect? Does it matter whether or not we are talking about people at the end of their productive lives or at the beginning? There is no agreement in our society, of course. So it usually is easier to default to the Hippocratic framework in which the ethical obligation is to do as much as you can for those in front of you. This framework relegates us to trying to fit as many goods and services as we can into our budgets, relying on tradition and inertia and “next increment” analysis to make these decisions. In my experience, preventive services usually receive the harshest next increment inspection, at least in general healthcare organizations. The burden of proof for these services appears far greater than for more traditional curative services additions. (Imagine who usually wins when a cardiac surgeon and a chief of From the Kaiser Foundation Health Plan Inc., Oakland, California Address correspondence and reprint requests to: David M. Lawrence, MD, Kaiser Foundation Health Plan, Inc., One Kaiser Plaza, Oakland, CA 94612. E-mail: david.lawrence@kp.org. The full text of this article is available via AJPM Online at www.elsevier.com/locate/ajpmonline.