There is a widespread assumption among health policy experts that the key problem with runaway health care costs is unnecessary care driven by the incentive to over-treat that is inherent in fee-for-service payment of doctors. Therefore, the argument goes, we need to improve financial incentives for care coordination and reorganize doctors into "Accountable Care Organizations," forcing primary care, specialist physicians, and hospitals into shared financial arrangements that shift at least some insurance risks onto providers, countering the fee-for-service incentive to over-treat.
While there are certainly some doctors providing unnecessary procedures due to fee-for-service financial incentives, it is extremely unlikely that this is the root of our health care cost problem. The argument that fee-for-service incentives are the driver of excess health care cost is based on a fundamental misdiagnosis of the reasons for unsustainable cost escalation in U.S. health care.
If one attempts to quantify the sources of excess U.S. health spending by looking at actual evidence, it is apparent that exorbitant and unnecessary administrative costs are the biggest driver (around 20-25% of National Health Expenditures1,2), followed by unnecessary care due to over-treatment3,4 (perhaps 10% of NHE, of which only a fraction is attributable to fee-for-service incentives), and expensive complications of under-treatment due to lack of access (perhaps 5-10% of NHE, plus a lot of suffering and death that does not show up in health spending figures). About half of over-treatment is due to unreasonable demands for care by patients, most of which is actually driven by providers (direct-to-consumer advertising for drugs, ads by hospitals, and by the recommendations of doctors.) Malpractice costs and defensive medicine are only a few percent at most.5
There is a problem with lack of coordination of care for certain patients, but the far bigger problem is inadequate access to necessary care. There is a nation-wide shortage of doctors in primary care and also in many specialties. This is compounded by the problems of un-insurance and under-insurance, and the refusal of many doctors to accept patients with insurance plans that are onerous, pay low fees, or both. Care coordination is meaningless without access.
According to a recent CBO report6, all 34 pilot care coordination projects funded by CMS either failed to save any Medicare spending at all, or if they did save on health care spending, they cost more in administrative expenses than they saved, for a net increase in total cost for all of them. Three of four payment reform demonstration projects that relied on pay-for-quality incentives failed to save money, and the only successful one negotiated a discounted, bundled fee for coronary bypass surgeries and did not use pay-for-performance incentives. After three years, the PROMETHEUS project on bundled payments for episodes of care has failed to implement any actual contracts due to the complexities of defining a "bundle."7 Just about all the "cost saving" initiatives in the Affordable Care Act (ACA) are along the same lines and will fail for the same reasons.
There are only a few U.S. health reform programs that have actually achieved significant cost savings without relying on "cherry picking" healthier populations and avoiding sicker ones. Major examples are Community Care of North Carolina8 and Rocky Mountain Health Plans in Colorado9. The common denominator is not elimination of fee-for-service, which both still employ; it is physician leadership, high levels of physician participation and buy-in, significantly improved access to outpatient care for sicker high-risk patients, and a shared commitment to quality improvement.
Part of the problem is indeed the imbalance in pay between certain specialties and primary care, rooted in the flawed Medicare SGR physician fee schedule, and we do need to re-think how the money is distributed between "cognitive services" and procedures. Increased payment for primary care and care-coordination is part of the solution, but does not require shifting insurance risk onto doctors via HMO's or ACO's.