Journal of Medical Ethics最新文献

Introduction: The integration of artificial intelligence (AI) into healthcare introduces innovative possibilities but raises ethical, legal and professional concerns. Assessing the performance of AI in core components of the United States Medical Licensing Examination (USMLE), such as communication skills, ethics, empathy and professionalism, is crucial. This study evaluates how well ChatGPT versions 3.5 and 4.0 handle complex medical scenarios using USMLE-Rx, AMBOSS and UWorld question banks, aiming to understand its ability to navigate patient interactions according to medical ethics and standards.

Methods: We compiled 273 questions from AMBOSS, USMLE-Rx and UWorld, focusing on communication, social sciences, healthcare policy and ethics. GPT-3.5 and GPT-4 were tasked with answering and justifying their choices in new chat sessions to minimise model interference. Responses were compared against question bank rationales and average student performance to evaluate AI effectiveness in medical ethical decision-making.

Results: GPT-3.5 answered 38.9% correctly in AMBOSS, 54.1% in USMLE-Rx and 57.4% in UWorld, with rationale accuracy rates of 83.3%, 90.0% and 87.0%, respectively. GPT-4 answered 75.9% correctly in AMBOSS, 64.9% in USMLE-Rx and 79.6% in UWorld, with rationale accuracy rates of 85.4%, 88.9%, and 98.8%, respectively. Both versions generally scored below average student performance, except GPT-4 in UWorld.

Conclusion: ChatGPT, particularly version 4.0, shows potential in navigating ethical and interpersonal medical scenarios. However, human reasoning currently surpasses AI in average performance. Continued development and training of AI systems can enhance proficiency in these critical healthcare aspects.

Hamas militants have raped and mutilated the bodies of dozens of men and women in Israel during their attack and captivity in Gaza. The exact extent of these atrocities, however, is unknown. For reasons of this sort and others, prosecuting sexual abusers during armed conflicts is notoriously difficult. In an attempt to make a legal case against Hamas militants, the Israeli authorities have recently ordered civilian hospitals to breach medical confidentiality and report unidentified data of patients who have suffered bodily mutilation and sexual abuse during the Hamas attack or during captivity in Gaza. Patients were not informed about this policy. This article first establishes a case for why medical confidentiality is such an important existential and normative value. In this light, it then argues that the Israeli policy is morally unjustified.

The development of vaccines against SARS-CoV-2 (COVID-19) presented a unique set of challenges. There was a global need for safe, effective vaccines against a new virus. In response to the development of vaccines for COVID-19 (some of which used novel technologies), there was a proliferation of no-fault compensation schemes (NFCS) for COVID-19 vaccine injuries. We identified 28 national vaccine injury NFCS operating in December 2019. Just 2 years later, over 130 countries had some NFCS coverage for COVID-19 vaccines. This rapid expansion was primarily driven by the creation of three multinational schemes. The COVID-19 Vaccines Global Access (COVAX) scheme covers vaccines given under the COVAX framework in 92 low and middle-income countries across the globe. The African Vaccines Acquisition Trust (AVAT) scheme covers vaccines administered under the AVAT framework in 36 African and Caribbean countries. The UNICEF scheme covers vaccines administered by UNICEF in 18 Asian countries.Because of the sudden expansion of no-fault compensation for vaccine injury, especially in developing economies, more research on the foundations, procedures and outcomes of NFCS is needed. In this article, we examine how these NFCS meet the needs of individual claimants and society more widely. To do so, we first review the rationales offered to support the creation of vaccine injury NFCS. We then argue that, in order to achieve their function as compensation mechanisms, NFCS should be accessible and offer substantive and procedural justice to claimants. Finally, we focus on transparency and accountability as necessary requirements to allow scrutiny over existing NFCS and their wider impacts.

In this article, our aim is to show why increasing the effectiveness of detecting doping fraud in sport by the use of artificial intelligence (AI) may be morally wrong. The first argument in favour of this conclusion is that using AI to make a non-ideal antidoping policy even more effective can be morally wrong. Whether the increased effectiveness is morally wrong depends on whether you believe that the current antidoping system administrated by the World Anti-Doping Agency is already morally wrong. The second argument is based on the possibility of scenarios in which a more effective AI system may be morally worse than a less effective but non-AI system. We cannot, of course, conclude that the increased effectiveness of doping detection is always morally wrong. But our point is that whether the introduction of AI to increase detection of doping fraud is a moral improvement depends on the moral plausibility of the current system and the distribution of harm that will follow from false positive and false negative errors.

We respond to David Wilkinson's arguments against our view of the ethicality of doctors' strikes and our claim that the 2023-2024 UK doctors' strikes are morally permissible and arguably supererogatory.Wilkinson proposes that in specialist outpatient settings, striking doctors should help arrange their own cover to prevent disproportionate harm to patients and to abide by the principles of non-maleficence and fiduciary duty. This hasn't happened during the 2023-2024 UK doctors' strikes; therefore, in his view, these strikes are morally impermissible. We reject Wilkinson's proposal on the grounds that the risk of disproportionate harm is adequately mitigated by existing arrangements and his interpretations of non-maleficence and fiduciary duty are overly demanding.We agree with Wilkinson that strikes put particularly high pressure on covering doctors in chronically under resourced specialisms. But this doesn't justify calling off or depowering doctors' strikes because, without effective strikes, under-resourcing is likely to continue and, ultimately, cause even more harm.Wilkinson argues that doctors cannot justifiably strike in the interests of public health because they don't have a broad duty to public health. We think they do have such a duty; however, we argue that doctors can justifiably strike in the interest of public health whether they have such a duty or not.Finally, we defend our claim that doctors' strikes can be supererogatory from Wilkinson's objections that there may be no such thing as supererogatory action and that our view absurdly entails that strikes can be supererogatory despite placing unfair demands on others.

Those giving birth within modern maternity systems are recognised as facing a number of barriers to person-centred care. In this paper, I argue that in order to best facilitate the conditions for positive change, work needs to be done to provide a more granular articulation of the specific barriers. I then offer a nuanced and contextually aware articulation of one key component of the overall failure to ensure person-centred care: medical authority and the expectation of conformity. Articulating these barriers with increased specificity is valuable, as it creates a stronger foundation from which to challenge existing problems which serve to constrain the autonomy of birthing individuals. The analysis offered in this paper also underscores the need for change at an institutional, rather than individual, level.

Background: Institutional review boards (IRBs) have been criticised for delays in approvals for research proposals due to inadequate or inexperienced IRB staff. Artificial intelligence (AI), particularly large language models (LLMs), has significant potential to assist IRB members in a prompt and efficient reviewing process.

Methods: Four LLMs were evaluated on whether they could identify potential ethical issues in seven validated case studies. The LLMs were prompted with queries related to the proposed eligibility criteria of the study participants, vulnerability issues, information to be disclosed in the informed consent document (ICD), risk-benefit assessment and justification of the use of a placebo. Another query was issued to the LLMs to generate ICDs for these case scenarios.

Results: All four LLMs were able to provide answers to the queries related to all seven cases. In general, the responses were homogeneous with respect to most elements. LLMs performed suboptimally in identifying the suitability of the placebo arm, risk mitigation strategies and potential risks to study participants in certain case studies with a single prompt. However, multiple prompts led to better outputs in all of these domains. Each of the LLMs included all of the fundamental elements of the ICD for all case scenarios. Use of jargon, understatement of benefits and failure to state potential risks were the key observations in the AI-generated ICD.

Conclusion: It is likely that LLMs can enhance the identification of potential ethical issues in clinical research, and they can be used as an adjunct tool to prescreen research proposals and enhance the efficiency of an IRB.