Case Study 1: The Doctor, the Dose, and the Five

The Situation

Dr. Amara Osei is a physician working the overnight shift in a regional hospital's emergency department when a mass casualty incident occurs: a building collapse sends twenty-three patients in rapid succession. The ED is overwhelmed.

Among the patients is one in critical condition who will die within the hour without an immediate, resource-intensive intervention — a procedure that will require Dr. Osei's full attention and the hospital's only functional trauma bay for at least ninety minutes. While Dr. Osei and the team are occupied with this patient, five other patients with severe but not immediately critical injuries will deteriorate significantly. Two of them will likely die. The other three will survive, but with substantially worse outcomes than if treated promptly.

The hospital's senior administrator has told Dr. Osei: "Statistically, we save more life-years by treating the five. The algorithm says so." Dr. Osei has sixty seconds to decide.

What Each Framework Says

The Consequentialist Analysis

A consequentialist framing focuses on expected outcomes. The administrator's algorithm is doing exactly what consequentialism recommends: quantifying expected welfare and directing resources to maximize it.

The calculation here might look like: treating the critical patient saves one life, with perhaps 30–40 life-years remaining. Treating the five saves approximately two lives (preventing the likely deaths) and significantly improves outcomes for three others. In expected life-years, the aggregate benefit of treating the five is higher.

But a sophisticated consequentialist wouldn't stop at the first-order calculation. What are the second-order effects? If physicians can be algorithmically instructed to abandon critical patients because the numbers favor other patients, several things happen:

• Trust in the doctor-patient relationship erodes. Patients in critical condition who reach the ED may wonder: "Will I be treated, or will I be calculated away?"
• Physicians who routinely triage by algorithm rather than clinical judgment may develop habits of detachment that affect the quality of care more broadly.
• The most vulnerable patients — those who are critical and have no other options — face the highest risk of being deprioritized.

A consequentialist might conclude that a rule of "treat critical patients as priority" produces better welfare outcomes over the long run than a rule of "optimize per-incident expected life-years," because of these systemic effects on trust and physician culture.

This illustrates the distinction between act utilitarianism (assess this specific act in this specific situation) and rule utilitarianism (follow the rule that, if generally followed, produces the best outcomes). The two forms of consequentialism often diverge — and this case is a good example of why.

The Kantian Analysis

Kant would resist the administrator's framing at the conceptual level. The critical patient is not a unit in an optimization function. She is a person — a rational agent with dignity — who presented to this hospital in need of care. The physician-patient relationship creates a duty of care to this specific person.

The Formula of Humanity is directly relevant: using the critical patient's life as an acceptable cost to optimize outcomes for others treats her as a means to the welfare of the five, not as an end in herself. The five patients also matter enormously — but their welfare cannot be secured by treating another patient's life as disposable input.

A Kantian analysis also examines the universalizability test. Consider the maxim: "When treating one patient would prevent better outcomes for more patients, deprioritize the one." Could you universalize this? If it became a universal law that critical patients are deprioritized when the numbers favor others, what happens? Patients who are critical have no assurance that presenting to an ED means they'll receive care — the system becomes one in which your treatment depends on who else shows up at the same time. No rational agent could will this as a universal principle they might themselves be subject to.

The Kantian conclusion: Dr. Osei has a duty to the critical patient that isn't simply overridable by aggregate calculations. This doesn't necessarily mean she never makes triage decisions — the Kantian framework allows for triage under conditions of genuine scarcity. But triage must respect each patient's dignity, not reduce them to welfare units.

The Virtue Ethics Analysis

Aristotle's framework asks a different question: what would a physician of excellent character — a physician with phronesis, practical wisdom — do in this situation?

A person of practical wisdom would not be running an algorithm in their head. They would be drawing on years of clinical judgment, a deep understanding of what medicine is for, and the virtues that define excellent doctoring: compassion (genuine concern for the suffering person before them), courage (the willingness to make hard decisions without retreating into bureaucratic cover), justice (giving each patient what they're due as a patient), and practical wisdom itself (the ability to integrate all of these in a specific situation).

Virtue ethics observes that the administrator's framing — "the algorithm says so" — is itself a virtue problem. It outsources moral responsibility to a calculation. A physician of good character doesn't abdicate judgment to an algorithm; they use judgment, including moral judgment, as a core professional competency. The excellent physician feels the weight of the decision — they are not indifferent to the five patients — but they do not hide behind a number.

What would the practically wise physician do? They might: - Begin with the critical patient, as their duty of care requires - Simultaneously mobilize every available resource for the five — calling for additional staff, implementing parallel triage protocols, contacting other facilities - Accept that they cannot save everyone, but ensure that no patient is treated as disposable

The virtue framework also attends to who Dr. Osei is becoming through this choice. Physicians who routinely treat patients as optimization targets tend to lose something essential to excellent medicine. The patient in front of you — critical, terrified, dying — is not an inconvenience to the algorithm.

The Hard Part: Where the Frameworks Diverge

This case shows real disagreement between the frameworks, not just different emphases.

A strict act-utilitarian might genuinely accept the administrator's reasoning. If the numbers favor treating the five, treat the five. The discomfort is a moral intuition shaped by evolutionary psychology for small-group contexts — not a reliable guide to aggregate welfare.

A Kantian would resist this strongly. The critical patient's dignity is not tradeable. There's a categorical difference between triage under genuine scarcity (choosing among patients when you cannot serve all) and using a patient's prioritization as an input to an optimization function.

The virtue ethicist focuses less on the binary choice and more on the quality of decision-making — the habits of character that produce excellent medical judgment over a career.

Discussion Questions

1. Is there a morally relevant difference between triage (prioritizing among patients under genuine scarcity) and optimization (deprioritizing a patient because the numbers favor other patients)? How would each framework characterize that difference?

2. The administrator says "the algorithm says so." Does this change the moral situation? Does it shift moral responsibility, or does it just add a layer of distance?

3. Suppose Dr. Osei learns after the fact that her decision — whichever way she chose — led to worse outcomes than the alternative would have. Does this change the moral quality of her decision? What does each framework say?

4. What obligations, if any, does the hospital system have to design its policies so that individual physicians are never put in this position?

5. Are there cases where you think the administrator's consequentialist framing is clearly correct? Where the line between acceptable triage and unacceptable instrumentalization lies?