Age, “Life-Cycles,” and the Allocation of Scarce Medical Resources

In Italy, the shortage of ventilators grew so urgent at peak surge that those over the age of 65 years—as well as younger patients with significant comorbidities—were being denied access to them.¹ That this exclusion also included younger patients with comorbidities suggests that it was not age itself that served as a criterion, but likelihood of survival. This justification would be consistent with Italian guidelines issued March 16, which state, “An age limit for the admission to the ICU may ultimately need to be set. The underlying principle would be to save limited resources which may become extremely scarce for those who have a much greater probability for survival and life expectancy, to maximize benefits for the largest number of people.”² By referring to both survival and life expectancy, however, this justification leaves unresolved the ethical issue of whether—and, if so, exactly how—age might appropriately be used as an allocation criterion.

One approach to this issue that has been gaining popularity recently is articulated by Douglas White and Bernard Lo,³ who argue that “saving the most lives” is “ethically insufficient” for rationing ventilators during the coronavirus disease 2019 (COVID-19) pandemic, and should be supplemented with “life years” and, more specifically, “life-cycle” considerations, based on the idea that everyone deserves equal opportunity to live through various stages of life. Such supplemental principles are adopted by White in the Pittsburgh Protocol for allocation of ventilators and other scarce resources during the COVID-19 surge, where life-cycles serve as a “tie-breaker” for patients with similar need and prognoses.⁴

White and Lo appeal to our moral intuitions by pointing to the idea that “many people” would support prioritizing a patient who would lose 40 or more years of expected life over another. This contrast is commonly cited by defenders of the life-cycle approach. However, our intuitions can be skewed by focusing on this type of example: for instance, an otherwise healthy 15-year-old vs a 55-year-old, who both need access to ventilator support. While an interesting hypothetical, the reality in the current COVID-19 pandemic points us to a much different likely scenario, given what is known about the risk of severe complications among the young and the old. Based on all available evidence concerning COVID-19 to date, a 15-year-old has an extremely small chance of needing ICU resources, let alone ventilator support. Indeed, despite the fact that those younger than 18 years make up 22% of the US population, they represent only 1.7% of the confirmed cases of COVID-19, and hospitalization data suggest that only 0.58% of pediatric COVID-19 patients were even admitted to an ICU, let alone required a ventilator. This is consistent with global data.⁵ Thus, a tie-breaker between a 55-year-old and a 15-year-old is very unlikely to be the actual scenario faced by ICU professionals, because the 15-year-old will almost always fall into a different triage category.

Likewise, saving 40 life-years among those with a more plausible risk of needing ventilator support—say, between a 40-year-old and an 80-year-old—is equally unlikely to pose a dilemma requiring nonmedical criteria. What is the likelihood of a 40-year-old and an 80-year-old having the same prognosis/lack of comorbidities that would place them in a different triage category? The percentage of people with at least one comorbidity associated with increased mortality for COVID-19 is more than double in those aged 80 years and older than it is for those aged 40 to 49 years (80.7% vs 38.1%). This percentage drops even further for those aged 30 to 39 years (26.8%) and below 20% for those younger than 30 years (Table 1 in Adams et al⁶). The relevance of these comorbidities associated with age are clearly recognized by the Swiss Academy of Medicine.⁷

A tie-breaker scenario is far more likely to be a choice between a 58-year-old and a 63-year-old. It is these patients that are likely to have similar risk of need for ventilatory support (as well as comorbidities) under the demographic risk stratifications we have seen with COVID-19. They are, however, categorized as falling into different “life-cycle” categories by the Pittsburgh Protocol.⁴ In short, the only cases in which life-cycle considerations are likely to be useful as a tie-breaker are the very ones that create the most discomfort with their application.

With this more likely scenario of fewer life-cycles to distinguish patients of similar prognosis in mind, the challenges posed by life-cycle considerations become clear. Philosopher John Taurek outlines the foundational moral problems of inter-personal (as opposed to intra-personal) life-year trade-offs in his classic paper “Should the numbers count?”⁸ In short, from each individual’s perspective, each loses the maximum they can: the rest of their life. More fundamentally, not everyone will ascribe value to sheer number of life-cycles experienced, valuing instead quality within a particular life-cycle over sheer number of life-cycles. We believe it is precisely this idea that led Ezekiel Emanuel and Alan Wertheimer⁹ to supplement their original articulation of why life-cycles might be justified in terms that included an “investment refinement,”⁹ the idea being that there might well be greater loss in being denied the “payoffs” of certain life investments than there is in the loss of sheer numbers of life-cycles. Interestingly, Emanuel and Wertheimer’s original life-cycle argument favors many adults over younger children for this reason, despite the fact that the children have had the opportunity to live through fewer life-cycles. At the least, priority based on number of life-cycles (rather than, for example, life investments) requires an argument for why someone with fewer life-cycles is more deserving of the resource.

Perhaps the most important feature of allocation criteria is that these be justifiable to the public. We believe the use of age is best justified as employed in the Swiss Academy of Medical Sciences Recommendations,⁷ which state:

Age in itself is not to be applied as a criterion, as this would be to accord less value to older than to younger people, thus infringing the constitutional prohibition on discrimination. Age is, however, indirectly taken into account under the main criterion “short-term prognosis,” since older people more frequently suffer from comorbidity. In connection with COVID-19, age is a risk factor for mortality and must therefore be taken into account.

In short, we suggest restricting rationing by medical professionals to their area of expertise, likelihood of medical benefit, rather than placing them in a situation of being asked to assess who is “more deserving” of being saved based on the number of life-cycles they have, or have not, experienced.