Technology Diffusion under New Payment and Delivery Models

Specific Aims

By 2023, the Congressional Budget Office (CBO) projects total health care spending by the federal government will reach above 7% of GDP. Former CBO and Office of Management and Budget Director Peter Orszag testified in 2008 that “No other single factor will exert as much influence over the federal government’s long term fiscal balance as the future growth rate of costs in the health sector.” In addition, health care spending growth, which economists believe is driven by advances in medical technology, threatens the viability of the private health care financing system, with the potential to generate a chain of actions that reduce access to care and undermine health outcomes. However, spending growth in a sector per se is neither good nor bad: If the value of the clinical benefits associated with a new technology exceeds its costs, greater spending enhances societal welfare. Conversely, if the clinical benefits fall short of costs, welfare will decline as spending rises. In classical economic models, rational actors pay for a new technology only when benefits exceed costs. Yet well-known market imperfections in health care due to, among other things, insurance and
information asymmetries imply that new technology may be adopted even if the benefits do not exceed the costs. Furthermore, one blanket judgment about costs and benefits of technology is inappropriate. Benefits exceed costs for some technologies but not for others. The ability of physicians and the provider organizations within which they practice to discriminate between beneficial and wasteful new technology is critical in determining the ultimate effects of technological adoption on spending growth and the value of expenditures.

New strategies to contain spending growth focus on delivery system and payment reform. The Affordable Care Act encourages new payment models for Medicare, such as risk-based contracts with Accountable Care Organizations (ACOs), to slow spending growth while improving quality. Over 250 provider groups have contracted with Medicare as ACOs, and many similar payment arrangements have been reached with commercial insurers. Under such arrangements, provider organizations have incentives to adopt and use new and existing technologies more judiciously than providers paid on a fee-for-service (FFS) basis. These activities can slow spending growth if they affect the types of technologies that diffuse and the rate of diffusion. They could be detrimental if high value technologies are not adopted or adopted slowly. The potential impact of these models will likely vary across organizations. The structure of the existing delivery system is very heterogeneous, ranging from solo practitioners to integrated delivery systems with significant infrastructure for management of quality and spending. Even large organizations vary significantly along dimensions such as
physician specialty mix, relationships with hospitals, and care management infrastructure. These traits may have important ramifications for spending and quality.

We propose to examine the relationship between organization traits and diffusion of medical technology using a wealth of diverse data from IMS Health, Medicare, and clinical registries for the period 2005-2015. We will study the diffusion of selected new technologies (drugs, devices, and biologics), including a subset that can be described as of higher or lower value, in 4 disease categories – cancer, depression, cardiac, and hip degeneration – as a function of organization characteristics. We will also conduct quasi-experimental analyses of the effects of ACO contracts on Medicare spending for a comprehensive set of new technologies and for selected services of lower or higher value. Our Specific Aims are:

AIM 1: To identify organizational characteristics, adjusting for physician and patient factors, associated with the diffusion of selected new technologies. We will estimate models that produce organization-specific diffusion parameters for each technology, and determine if decisions to use new technologies are correlated among different technology types, or within and between disease conditions.

AIM 2: To identify organizational characteristics, adjusting for physician and patient factors, associated with use of lower and higher value services. Using measures promulgated by professional societies, technology characteristics, and FDA alerts, we will distinguish higher from lower value services, identify organizational factors predictive of their use, and determine if decisions to use higher value and lower value services are correlated.

AIM 3: To estimate the effects of ACO contracts in Medicare on spending for new technologies and for lower vs. higher value services. Using a quasi-experimental design, we will compare organizations that participate in the Medicare ACO demonstration program to organizations that do not to assess if risk bearing by organizations influences spending on new technologies and on selected services of higher or lower value.

Given the increasing focus on provider organizations as the locus of accountability in health care, our research will provide critical information about the potential of one prominent strategy– risk-based contracting with large integrated provider organizations -- for influencing technological diffusion and controlling expenditures while maximizing value. By examining diffusion across a broad range of technologies having different financial incentives for use by organizations under current payment approaches, our findings also will inform the development of other organization-level approaches to achieving these goals.


Schwartz AL, Zaslavsky AM, Landon BE, Chernew ME, McWilliams JM. Low-Value Service Use in Provider Organizations. Health Serv Res. 2018 Feb;53(1):87-119. doi: 10.1111/1475-6773.12597. Epub 2016 Nov 10. PubMed PMID: 27861838; PubMed Central PMCID: PMC5785325.


Schwartz AL, Chernew ME, Landon BE, McWilliams JM. Changes in Low-Value Services in Year 1 of the Medicare Pioneer Accountable Care Organization Program. JAMA Intern Med. 2015 Nov;175(11):1815-25. doi: 10.1001/jamainternmed.2015.4525. PubMed PMID: 26390323; PubMed Central PMCID: PMC4928485.

Gilstrap LG, DeFilippis EM, Stevenson LW. An Unintended Consequence of the 21st-Century Cures Act for Patients With Heart Failure. Circulation. 2017 Jul 11;136(2):123-125. doi: 10.1161/CIRCULATIONAHA.117.028747. PubMed PMID: 28696263; PubMed Central PMCID: PMC5512586.

Keating NL, Huskamp HA, Schrag D, McWilliams JM, McNeil BJ, Landon BE, Chernew ME, Normand ST. Diffusion of Bevacizumab Across Oncology Practices: An Observational Study. Med Care. 2018 Jan;56(1):69-77. doi: 10.1097/MLR.0000000000000840. PubMed PMID: 29135615; PubMed Central PMCID: PMC5726588.