Mednotes

L3 Ethics Ethics of Resource Allocation

Properties2
tagsethics, lecture
aliasesL3 Ethics Ethics of Resource Allocation

16 min read

Overview

This lecture asks how health care resources should be rationed justly, given that some rationing is unavoidable. It works through the QALY as the standard health-economic tool for prioritisation, the main objections to using QALYs alone, the “rule of rescue” as an alternative pull on decisions, and then broadens out to justice itself: Rawls’ original position as a way of generating fair principles, and two live NZ/international cases (ethnicity-based surgical prioritisation, and race and the SOFA score in COVID-19 ventilator triage) where “colourblind” clinical criteria can still produce unjust outcomes.

The scope and inevitability of rationing

  • Justice in health care is broader than rationing alone; the areas covered are: rationing, equity in how services are structured and delivered, research participation, global health inequity, public vs private funding of health care, research priorities, and payment for organ donation.
  • Rationing is inevitable, so the operative question is “how to ration justly?”, not whether to ration.
  • Light (1997)‘s warning: treating rationing as simply inevitable and only asking “how to ration reasonably” is like medicine deciding death is inevitable and therefore only focusing on dying reasonably - death is inevitable, but that conclusion denies the whole purpose of medicine. The parallel point: accepting rationing as a given risks quietly abandoning the pursuit of justice in how it is done.
  • Possible criteria that have been proposed for rationing: age, responsibility for disease, prognosis, length of time waiting, benefiting others, desert, ability to pay, urgency of need.
  • Case example - Leslie Burke: did not want artificial nutrition and hydration (ANH) withdrawn from him; relevant to the GMC’s “Withholding and Withdrawing Life-Prolonging Treatments” guidance; the impact on resources was raised as a major concern in his case.

QALYs as a tool for maximising benefit

  • PHARMAC: created in 1993; manages the pharmaceutical schedule of government-subsidised community pharmaceuticals and the funding of medicines used in NZ public hospitals; aims to secure the best possible health outcomes from the money government spends on medicines; uses several criteria, including cost-effectiveness as measured by tools such as the QALY.
  • QALY (Quality Adjusted Life Year) definition: a year of healthy life expectancy is worth 1; a year of unhealthy life is worth less than 1, with the precise value falling the worse the person’s quality of life. Because being dead is worth zero, a QALY can in principle be negative, i.e. some health states are considered worse than death (Williams 1985).
  • Efficiency and priority: a beneficial health activity generates a positive number of QALYs; an efficient activity is one where cost per QALY is as low as possible; high-priority activities have low cost-per-QALY and low-priority activities have high cost-per-QALY (Williams 1985).
  • Cost per QALY gained for selected interventions (Williams 1985), lowest (most efficient) to highest:
    • GP advice to stop smoking: 260
    • Hip replacement: 1140
    • CABG for severe angina, left main disease: 1590
    • GP control of total serum cholesterol: 2600
    • Breast cancer screening: 5340
    • CABG for mild angina, 2-vessel disease: 19250
    • Hospital haemodialysis: 21500

Objections to QALYs

  • Quality of the underlying data: most QALY utilities are “ex ante” preferences (how people imagine an illness would affect them) rather than “ex post” (how people who actually have the illness rate their own quality of life) - raising the question of whether you can really know what an illness is like without having it.
  • Ageism objection: the elderly have lower remaining life expectancy, so the same treatment yields fewer QALYs for them; this also raises the separate question of how QALYs handle care for chronic illness in the young. QALYs don’t “directly” discriminate on the basis of age. The standard response to the ageism worry is the “fair innings” argument.
  • Double jeopardy objection (Harris 1987): patient X was paralysed in an accident and has quality of life (QoL) 0.5; patient Y has QoL 0.7. A treatment T would raise X’s QoL by 0.2 and Y’s QoL by 0.3. Because T produces more QALYs for Y, a QALY-based allocation favours Y - so X is penalised twice: once by the original misfortune (lower baseline QoL), and again by being deprioritised for the treatment that could help her.

The rule of rescue

  • Definition: providing aid to identified victims of illness or accident.
  • The public is more sympathetic to the needs of a named person who is dying than to statistical lives, so the rule of rescue (RoR) exerts a real pull on rationing decisions.
  • RoR is not captured by QALY calculations, and is itself highly arguable as a principle for allocating resources.
  • Illustrative case - Child B (Jaymee Bowen): non-Hodgkin’s lymphoma in 1990, acute myeloid leukaemia in 1993, relapsed in 1995 with 6-8 weeks to live. A second opinion estimated a 20% chance of remission. Cambridge and Huntingdon health authority declined to fund treatment, with chemotherapy and transplant costed at approximately £70,000. An anonymous donor ultimately funded the treatment (Ham 1999).

Justice, equity and ethnicity in NZ health policy

  • Used on their own, QALYs are considered “unfair” or “unjust” - which raises the underlying question of what justice actually is.
  • Two forms distinguished: retributive justice, and distributive justice - the allocation of liberties and resources, and also how and what gets allocated.
  • He Korowai Oranga guides the NZ government and the health and disability sector to achieve the best health outcomes for Māori. Key threads:
    • Rangatiratanga: Māori having control over their own health and wellbeing, and over the direction and shape of their own institutions, communities and development.
    • Building on the gains: whānau ora remains a priority for the next decade, and building on existing gains is central to achieving pae ora.
    • Equity: better health for all New Zealanders, with achieving health equity as a minimum standard.
    • Equity means recognising that people with different levels of advantage need different approaches and resources to reach equitable health outcomes, because in Aotearoa NZ health differences exist that are not only avoidable but unfair and unjust.
  • Policy example: from February 2023, Auckland surgeons were required to consider a patient’s ethnicity alongside other factors when prioritising patients for operations; some surgeons were unhappy about this. Te Whatu Ora’s Equity Adjustor Score uses an algorithm that prioritises patients on five factors: clinical priority, time spent on the waitlist, geographic location (isolated areas), ethnicity, and deprivation level, giving priority weight to Māori and Pacific patients on the grounds of historically unequal access to healthcare.
  • McMillan’s response (opinion piece): prioritisation decisions are “clinical” judgments but are also fundamentally ethical judgments; given the likelihood of reasonable disagreement about who should be treated first and why, fairness requires that decisions be made consistently, using transparent criteria.
  • Supporting NZ ethnic-disparity data presented in the lecture:
    • Breast cancer mortality (1995-2020): consistently and substantially lower in non-Māori non-Pasifika women than in Māori or Pasifika women throughout the period, despite an overall downward trend in all three groups.
    • Breast screening (BSA) coverage: exceeded 70% in Pasifika and non-Māori non-Pasifika women by the two-year period ending 2020, then fell sharply to minimum levels by the period ending 2022; Māori coverage remained the lowest of the three groups throughout.
    • Cervical cancer mortality (2010-2018): generally highest in Māori and especially Pacific women compared with Asian, Other and All-women groups, though confidence intervals are wide (particularly for Pacific) and overlap with other groups in several years.
    • NCSP cervical screening coverage (to March 2022): declined across all ethnic groups over the six years; “Other” had the highest coverage throughout, followed by Pacific, with Māori and Asian lowest and similar to each other.
  • HPV self-testing trial (Sykes et al., BJOG, 2025): Aotearoa NZ is presented as the first country to adopt HPV self-testing as the primary cervical screening tool; framed as improving accessibility for under-screened people and addressing screening inequity, and, combined with HPV vaccination uptake, as a potential route to effectively eliminating cervical cancer.

Rawls’ theory of justice

  • John Rawls (1921-2002), Harvard professor, author of A Theory of Justice (1971), described as having reshaped the intellectual political landscape.
  • Key quote: “Justice is the first virtue of social institutions, as truth is of systems of thought. A theory however elegant and economical must be rejected or revised if it is untrue; likewise laws and institutions no matter how efficient and well-arranged must be reformed or abolished if they are unjust.”
  • The original position: a hypothetical contracting situation in which citizens are ignorant of their place in society, class or social status (social goods), and of their own intelligence, strength and health (natural goods). Any person could turn out to be any member of society, and no one knows their own conception of the good or the particulars of their rational life plan. Under this “veil of ignorance,” the principles of justice selected would treat everyone with equal concern and respect, so that no one is advantaged or disadvantaged by natural chance or the contingency of their social circumstances.
  • Triage under scarcity is used (illustrated with a WWI-era aid-station photograph) as a practical setting in which Rawlsian, justice-based thinking about fair allocation applies.

Triage in practice: the Richard Smith case

  • Case: Richard has been allocated the last ICU bed and is reasonably stable. A new patient arrives who urgently needs ICU admission, and moving Richard would be much riskier for him than leaving him in place. Should Richard be moved?
  • Intensive Care Society guidelines (2011): when a transfer is needed because no ICU beds are available, it may be appropriate to consider moving a new, potentially unstable patient rather than a patient already on the unit who is more stable and less likely to deteriorate. In general, no patient should be subjected to an intervention that is not in their own best interest, so transferring a patient purely to make room for another could be considered unethical - though it may nonetheless be the most pragmatic approach.
  • A 2021 framework (developed post-pandemic) placed greater emphasis on justice and duty of care in decisions of this kind.

Race-conscious rationing: COVID-19 ventilator allocation

  • Debate (Schmidt, Roberts & Eneanya, “Rationing, racism and justice”): the SOFA (Sequential Organ Failure Assessment) score used to predict short-term ICU survival incorporates creatinine, which simultaneously reflects social disadvantage. Creatinine tends to be higher in Black communities because of higher rates of chronic kidney disease, which are themselves a consequence of higher rates of diabetes and hypertension driven by health inequities and structural racism, not biology alone.
  • Multi-principle strategy for allocating critical care during a public health emergency, using a points system (1-4) across two principles:
    • “Save lives” (SOFA score, short-term survival): 1 point if SOFA <6; 2 points if SOFA 6-8; 3 points if SOFA 9-11; 4 points if SOFA ≥12.
    • “Save life-years” (assessed prospects for survival after hospital discharge): 2 points if life expectancy <5 years despite successful treatment of the acute condition; 4 points if death is likely within 1 year despite successful treatment.
    • The combined priority score maps to a colour-coded priority category: RED, highest priority, score 1-3; ORANGE, intermediate priority (reassess as needed), score 4-5; YELLOW, lowest priority (reassess as needed), score 6-8.
  • Worked example: an ICU with 12 ventilators, 11 already in use, and 3 equally critically ill COVID-19 patients in respiratory failure competing for the last one:
    • John: 55, white accountant, comprehensive employer-sponsored health insurance, from a well-off suburb, average life expectancy 85, generally healthy but alcohol-dependent. Creatinine 1.18, SOFA 6, life expectancy >5 years - high priority score.
    • James: 55, unemployed Black sales assistant, poorly controlled hypertension, mild chronic kidney disease, asthma, from a neighbourhood where people die around 25 years earlier than in John’s. Creatinine 1.4, SOFA 7, life expectancy <5 years - intermediate priority score.
    • Martin: 55, recently furloughed Black hotel worker, from a neighbourhood with average life expectancy 70, severe diabetes, advanced chronic kidney disease. Creatinine 4.1, SOFA 12, life expectancy >5 years - intermediate priority score.
  • Proposed responses to the racial-bias problem raised by the article: using triage officers with equity expertise (though it is unclear whether they could depart from a clinical test); adjusting creatinine by race (rejected as creatinine is not the main variable and the adjustment would not be accurate); dropping creatinine altogether (problematic because it is integral to SOFA); and adding equity weights to the scoring system.

Self-test

  1. Define the QALY, and explain how a QALY value can end up negative.
  2. Using Williams (1985), explain what makes a health care activity “efficient” and what determines whether it is high or low priority.
  3. Distinguish “ex ante” from “ex post” utilities, and explain why this distinction is a problem for QALY data.
  4. Describe the ageism objection to QALYs, and name the standard response to it.
  5. Describe the “double jeopardy” objection to QALYs using Harris’s example of patients X and Y.
  6. Define the rule of rescue, and explain why it is not captured by QALY-based allocation.
  7. Describe the Child B case and explain what it illustrates about rationing decisions.
  8. Distinguish retributive justice from distributive justice.
  9. Describe the key features of Rawls’ original position, and explain how it is meant to produce fair principles of justice.
  10. In the Richard Smith case, what does the 2011 Intensive Care Society guidance say about transferring a stable patient to admit a new, unstable one?
  11. Explain how NZ’s Equity Adjustor Score prioritises elective surgery patients, and what concern this policy raised among some surgeons.
  12. Explain how the SOFA score’s use of creatinine can introduce racial bias into ventilator triage, according to Schmidt, Roberts and Eneanya.
  13. In the multi-principle ventilator allocation strategy, distinguish the “save lives” principle from the “save life-years” principle.
  14. A patient has SOFA 9 and life expectancy <5 years despite successful treatment. What priority score and colour category would they receive under the multi-principle strategy?
  15. Using Rawls’ original position, explain why an ethnicity-based rationing criterion (such as an Equity Adjustor Score) might be defended, or criticised, from behind the veil of ignorance.

Answers

Graph View

Backlinks