In “Is the US Health Care System Wasteful and Inefficient? A Review of the Evidence,” Sherry Glied and Adam Sacarny examine the common wisdom that the American health system is inefficient. Throughout the authors are precise in their definition of inefficiency. Specifically, following the framework of economics, inefficiency relates to excess provision of care relative to what would be needed to generate the same amount of health. High prices, while a prime determinant of high spending in the United States, are not necessarily evidence of inefficiency. Similar, greater care provided because of poor American health behaviors, and thus greater need, would also contribute to higher spending in the United States but would not indicate greater inefficiency.
There are a few definitional issues worth noting. Inefficiency may reflect using too many resources (or an unnecessarily expensive mix of inputs) to produce specific services, or it could reflect delivering too many services. If one conceptualizes the “product” as an episode of care, as opposed to delivery of a specific service, these two definitions converge because the specific services are simply inputs to treatment of the episode of care. Thus, overprovision of services is akin to using an unnecessarily expensive mix of inputs to produce health. This argument may seem purely semantic, and to an extent it is, but it emphasizes that health is the output and services are the inputs. A national health accounts perspective that treats the output of the health care sector as physician visits or hospital stays obscures the key point that efficiency requires generating health with fewer of these services as opposed to more.
As Glied and Sacarny note, the geographic variations literature provides some of the strongest evidence suggesting inefficiency in the system. They provide a thoughtful discussion of the evidence, raising methodologic issues such as how coding may affect the conclusions, and conceptual issues, such as whether inefficiency arises because some areas have inefficient production functions versus the possibility that inefficient areas produce at a different point of the production function. Moreover, they note that understanding the geographic variation is more complex than simply identifying high- and low-efficiency areas. The high-use areas for Medicare populations are not necessarily the high-use areas for commercial populations, and areas that are high users of some services may not be high users of other services. Thus, policy prescriptions are complex. Forcing high-spending areas to reduce use may not lead to outcomes comparable to those in low-use areas.
In fact, one of the strengths of this work is the discussion of strategies to reduce waste. My favorite passage is from the conclusion: “Recognizing the presence of inefficiencies, however, is not the same thing as addressing them.” Importantly, we must be aware of the production function associated with reducing waste. Reforms that require additional resources (such as case managers, data analysts, and targeted programs) entail a real resource cost. Many supply-side interventions require such resource use. Benefit design changes, including, but not limited to value-based insurance designs (V-BIDs), may be less expensive to maintain, though the authors' point about the importance of simplicity is well taken.
It is clear that there is no magic bullet and that no option is perfect. For example, ample evidence suggests that higher patient cost sharing can reduce utilization possibly without adversely affecting health, on average, but that the response to higher cost sharing is imperfect. Patients cut back on high- and low-value services in similar proportion. Maybe we can do better through more targeted cost sharing with programs such as V-BIDs, but the imperfections in patient response to cost sharing do not negate the basic finding that higher cost sharing can reduce waste. Of course, higher cost sharing increases risk, and the decline in financial protection suggests that the reduction in waste in the delivery of care, on average, does not imply that higher cost sharing would produce a more efficient health care system.
One way or another, the United States must slow the rate of health care spending growth. Doing so likely requires better understanding of where savings opportunities exist, and the precise discussion of the related concepts and evidence, as provided by Glied and Sacarny, is a useful contribution to that debate.