873 research outputs found
Problematic Interactions between AI and Health Privacy
No full textProblematic Interactions Between AI and Health Privacy Nicholson Price, University of Michigan Law SchoolFollow Abstract The interaction of artificial intelligence (AI) and health privacy is a two-way street. Both directions are problematic. This Essay makes two main points. First, the advent of artificial intelligence weakens the legal protections for health privacy by rendering deidentification less reliable and by inferring health information from unprotected data sources. Second, the legal rules that protect health privacy nonetheless detrimentally impact the development of AI used in the health system by introducing multiple sources of bias: collection and sharing of data by a small set of entities, the process of data collection while following privacy rules, and the use of non-health data to infer health information. The result is an unfortunate anti-synergy: privacy protections are weak and illusory, but rules meant to protect privacy hinder other socially valuable goals. The state of affairs creates biases in health AI, privileges commercial research over academic research, and is ill-suited to either improve health care or protect patients. The health system deeply needs a new bargain between patients and the health system about the uses of patient data
Evening with the Reverend and Mrs. Herbert V. Nicholson
No full textProgram for a talk by Reverend Nicholson sponsored by several Los Angeles area Japanese community organizations. Introduction to "Valient Odyssey, Herbert Nicholson in and out of America's concentration camps" by Michi Weglyn and Betty E. Mitson is featured.The Japanese American Relocation Collection is composed of ephemera related to the relocation program during World War II. Items include the official government report of Manzanar Relocation Center, a photo album, post-war activism materials related to preserving and remembering the camps, various clippings, and documents. The strength of this collection is found in its many perspectives on the controversial relocation program and how it has been presented since World War II
Are Trade Secrets Delaying Biosimilars?
No full textOn 6 March 2015, the United States Food and Drug Administration (FDA) approved, under the Biologics Price Competition and Innovation Act (BPCIA), a biosimilar of filgrastim (Neupogen), for treating chemotherapy-caused neutropenia (1). Although this action represents a step toward cheaper medical treatments, it masks systemic problems. Not only has it taken 5 years since the BPCIA\u27s passage (2), but economists estimate that even by 2020, biosimilar competition will reduce consumer prices only modestly (3). Why will price competition be so lacking? One key reason is the barrier to competitive entry created by trade secrecy in biologics manufacturing
Am I My Son? Human Clones and the Modern Family
No full textAs increasingly complex assisted reproductive technologies (ART) become available, legal and social conceptions of family become ambiguous and sometimes misaligned. The as-yet unrealized technology of cloning provides the clearest example of this confusion: is the legal parent of a clone the individual cloned, or are that individual‘s parents also the parents of the clone? These issues have been generally obscured by the debates around the deployment of ART, especially cloning; far less consideration has been given to the way these new technologies impact the way we think about and develop law on the relationships between genetic, social, gestational, and legal parenthood. This article considers these issues in depth, looking at competing common-law frameworks for determining parentage, the statutory framework of parentage, and deeper theoretical concerns underlying the area. The article concludes that an intent-based framework, with at least some external limitations, most accurately matches law to social views of parents using new forms of ART
Problematic Interactions Between AI and Health Privacy
No full textThe interaction of artificial intelligence (AI) and health privacy is a two-way street. Both directions are problematic. This Essay makes two main points. First, the advent of artificial intelligence weakens the legal protections for health privacy by rendering deidentification less reliable and by inferring health information from unprotected data sources. Second, the legal rules that protect health privacy nonetheless detrimentally impact the development of AI used in the health system by introducing multiple sources of bias: collection and sharing of data by a small set of entities, the process of data collection while following privacy rules, and the use of non-health data to infer health information. The result is an unfortunate anti-synergy: privacy protections are weak and illusory, but rules meant to protect privacy hinder other socially valuable goals. The state of affairs creates biases in health AI, privileges commercial research over academic research, and is ill-suited to either improve health care or protect patients. The health system deeply needs a new bargain between patients and the health system about the uses of patient data
The Social Cost-of-Living: Welfare Foundations and Estimation
No full textWe present a new class of social cost-of-living indices and a nonparametric framework for estimating these and other social cost-of- living indices. Common social cost-of-living indices can be understood as aggregator functions of approximations of individual cost-of-living indices. The Consumer Price Index (CPI) is the expenditure-weighted average of first-order approximations of each individual’s cost-of-living index. This is troubling for three reasons. First, it has not been shown to have a welfare economic foundation for the case where agents are heterogeneous (as they clearly are.) Second, it uses an expenditure-weighted average which downweights the experience of poor households relative to rich households. Finally, it uses only first-order approximations of each individual’s cost-of-living index, and thus ignores substitution effects. We propose a “common-scaling” social cost-of-living index, which is defined as the single scaling to everyone’s expenditure which holds social welfare constant across a price change. Our approach has an explicit social welfare foundation and allows us to choose the weights on the costs of rich and poor households. We also give a unique solution for the welfare function for the case where the weights are independent of household expenditure. A first order approximation of our social cost-of- living index nests as special cases commonly used indices such as the CPI. We also provide a nonparametric method for estimating second- order approximations (which account for substitution effects).Inflation, Social cost-of-living, Demand, Average Derivatives
Drug Approval in a Learning Health System
No full textThe current system of FDA approval seems to make few happy. Some argue FDA approves drugs too slowly; others too quickly. Many agree that FDA—and the health system generally—should gather information after drugs are approved to learn how well they work and how safe they are. This is hard to do. FDA has its own surveillance systems, but those systems face substantial limitations in practical use. Drug companies can also conduct their own studies, but have little incentive to do so, and often fail to fulfil study commitments made to FDA. Proposals to improve this dynamic often suggest gathering more information after approval in various ways and incorporating that information into FDA’s decision-making process, making the information/access tradeoff more nuanced than a sharp binary at approval. The drug approval regime has already begun to move in this direction
Distributed Governance of Medical AI
No full textArtificial intelligence (AI) promises to bring substantial benefits to medicine. In addition to pushing the frontiers of what is humanly possible, like predicting kidney failure or sepsis before any human can notice, it can democratize expertise beyond the circle of highly specialized practitioners, like letting generalists diagnose diabetic degeneration of the retina. But AI doesn’t always work, and it doesn’t always work for everyone, and it doesn’t always work in every context. AI is likely to behave differently in well-resourced hospitals where it is developed than in poorly resourced frontline health environments where it might well make the biggest difference for patient care. To make the situation even more complicated, AI is unlikely to go through the centralized review and validation process that other medical technologies undergo, like drugs and most medical devices. Even if it did go through those centralized processes, ensuring high-quality performance across a wide variety of settings, including poorly resourced settings, is especially challenging for such centralized mechanisms. What are policymakers to do? This short Essay argues that the diffusion of medical AI, with its many potential benefits, will require policy support for a process of distributed governance, where quality evaluation and oversight take place in the settings of application—but with policy assistance in developing capacities and making that oversight more straightforward to undertake. Getting governance right will not be easy (it never is), but ignoring the issue is likely to leave benefits on the table and patients at risk
Risk and Resilience in Health Data Infrastructure
No full textToday’s health system runs on data. However, for a system that generates and requires so much data, the health care system is surprisingly bad at maintaining, connecting, and using those data. In the easy cases of coordinated care and stationary patients, the system works — sometimes. But when care is fragmented, fragmented data often result. Fragmented data create risks both to individual patients and to the system. For patients, fragmentation creates risks in care based on incomplete or incorrect information, and may also lead to privacy risks from a patched-together system. For the system, data fragmentation hinders efforts to improve efficiency and quality, and to drive health innovation based on collected data. Efforts to combat data fragmentation would benefit by considering the idea of health data infrastructure. Most obviously, that would be infrastructure for health data — that is, infrastructure on which health data can be stored and transmitted. But it should also be an infrastructure of health data — that is, a platform of shared data on which to base further efforts to increase the efficiency or quality of care
The Cost of Novelty
No full textPatent law tries to spur the development of new and better innovative technology. But it focuses much more on “new” than “better”—and it turns out that “new” carries real social costs. I argue that patent law promotes innovation that diverges from existing technology, either a little (what I call “differentiating innovation”) or a lot (“exploring innovation”), at the expense of innovation that tells us more about existing technology (“deepening innovation”). Patent law’s focus on newness is unsurprising, and fits within a well-told narrative of innovative diversity accompanied by market selection of the best technologies. Unfortunately, innovative diversity brings not only the potential benefits of technological advances but also the costs: incompatibility between different technologies; a spread-out, shallow pool of knowledge; and the underlying costs of developing parallel technologies that aren’t actually better. These costs matter. Biomedical innovation illustrates the high costs of divergence. Although pharmaceuticals are touted as a poster child for patents, the world is rife with me-too drugs that drive up costs with little to show for it. Biomedical innovation often suffers from a particular trap: Patent incentives push innovators toward “new,” but incentives from Food and Drug Administration approval and insurer reimbursement push innovators toward “not too new.” In this space, artificially constricted markets do a poor job of selecting better technologies. The result is a proliferation of technologies that are “new for the sake of new,” giving us the costs of divergence without much in the way of benefits. This Essay presents an original spectrum of innovative divergence, illuminates how various patent doctrines drive divergence, and lays out the substantial costs of divergence through biomedical examples. It analyzes the complex interactions between three different incentives for biomedical innovation and presents policy prescriptions to help avoid the trap of “new for the sake of new.” In the process, it lays out how innovation scholars and policymakers alike should take into account the cost of novelty
- …
