Whose Values Matter? Ethical Pluralism and Public Attention in Healthcare AI

PURPOSE

This study examines how different individuals prioritize ethical values when these values come into conflict in the context of AI-powered healthcare and surgery. From participant responses, we identify recurring patterns of value prioritization, referred to as ethical personas, to show where judgments converge, where they diverge, and which perspectives account for those differences.

PROCEDURES

You will read eight short scenarios, each presenting a situation where two ethical values come into conflict. For each scenario, you will indicate where you stand between the two values, showing which value you believe should take priority and to what degree, on a scale from 1 to 5.

RISKS AND BENEFITS

This study presents minimal risk. Some individuals may experience mild discomfort when engaging with speculative health scenarios. You may skip any question and withdraw at any time without penalty. There is no direct personal benefit; the benefit lies in contributing to academic understanding of how people prioritize competing ethical values in healthcare AI.

CONFIDENTIALITY

I understand that I am free to withdraw my consent at any time without negative consequences.
I understand that my participation is CONFIDENTIAL — responses are anonymous, de-identified, and no personal identifiers will be collected.
I understand that my responses will be combined with other participants’ responses to identify common patterns of value prioritization (ethical personas), and that demographic data will only be reported in aggregate.
I understand that results may appear in an academic publication.
I understand that this study is conducted by Nadine El-Mufti and Mohamad Salim under the supervision of Marta Kersten-Oertel, Department of Computer Science and Software Engineering, Concordia University (marta.kersten@concordia.ca).

CONDITIONS OF PARTICIPATION

Participation is entirely voluntary. You may withdraw at any time. All participants must be 18 years of age or older.

Scenario: A patient undergoes a pre-surgical scan with an advanced AI. The AI not only plans the surgery but also detects a high genetic predisposition for a severe, unrelated neurological condition that will likely manifest in 10-15 years. The hospital's policy is to not disclose such incidental, non-actionable findings to prevent "psychological harm," but the patient's surgeon feels they have a right to know. Value Conflict: Autonomy (A patient's right to know all information about their own health) vs. Safety (The medical principle of "do no harm," including protecting the patient from psychological distress).

Scenario: A new AI surgical robot has a 5% higher success rate than any human surgeon for a complex procedure. However, its decision-making is a "black box," meaning its developers cannot fully explain why it makes certain micro-adjustments during surgery. Surgeons must trust its proven outcomes without understanding its internal logic. Value Conflict: Transparency (The need for surgeons and regulators to understand how a tool works) vs. Innovation (Adopting a new technology that demonstrates superior results, even if its process is opaque).

Scenario: An AI is designed to optimize the national organ transplant waiting list. To maximize "successful life-years" gained, the algorithm consistently prioritizes younger, healthier recipients who live in areas with better post-operative care. This systematically de-prioritizes patients in remote or low-income communities with fewer healthcare resources. Value Conflict: Equity (Ensuring fair and unbiased access to organs for all, regardless of background) vs. Efficiency (Optimizing the limited resource of donor organs to achieve the maximum number of successful, long-term outcomes).

Scenario: During a pre-operative assessment, an AI screening tool analyzes a patient's DNA (with their consent) to check for anesthesia risks. The AI also flags a genetic marker for high-risk addictive behaviors. The patient's insurance company, which partly funded the AI's development, wants this secondary data to adjust the patient's future coverage and premiums. Value Conflict: Privacy (Protecting the patient's sensitive genetic data from being used for non-clinical, financial purposes) vs. Efficiency (Streamlining health-risk data between providers and insurers to manage system-wide costs).

Scenario: A new remote-surgery system allows top surgeons to operate on patients in rural areas. The government has a budget to roll this out. They can either install the high-cost, high-tech system in 10 mid-sized towns (providing excellent care to a few thousand people) or use the same funds to build 50 traditional, less-advanced clinics in many more remote areas (providing basic care to ten thousand people). Value Conflict: Innovation (Deploying cutting-edge, high-quality surgical care) vs. Accessibility (Ensuring the broadest possible reach of basic health services to the largest number of people).

Scenario: To build a powerful AI for detecting rare cancers, a research institute wants to use all "anonymized" surgical biopsy samples from a national hospital network. To get enough data, they propose an "opt-out" system where all patients automatically contribute their data unless they actively sign a form to decline. This will massively accelerate the research. Value Conflict: Social Impact (The immense potential public good from large-scale medical research) vs. Autonomy (A patient's right to give explicit, informed consent before their data or tissue is used, even if anonymized).

Scenario: A new AI-assisted surgical tool has a 10% higher success rate for critical heart surgeries. However, the tool relies on single-use, high-energy components that are non-recyclable and create a significant amount of toxic medical waste. The hospital must choose between offering the best patient outcome and adhering to its environmental commitments. Value Conflict: Safety (Maximizing the chance of survival and a successful outcome for the individual patient) vs. Sustainability (The long-term environmental impact and viability of medical practices).

Scenario: A private tech company develops a revolutionary AI algorithm that dramatically improves surgical outcomes. They offer it to hospitals at a very high price, arguing it's necessary to recoup their massive R&D costs. Public hospitals argue that since the AI was trained using (anonymized) public patient data, the algorithm should be open-source and freely accessible to all. Value Conflict: Innovation (Allowing private companies to profit, thus encouraging them to invest in R&D) vs. Equity (Ensuring that breakthroughs trained on public data are available to all, not just those who can afford the high price).