Optimizing Kidney Allocation: LYFT for LIFE

Under the current system, kidneys are allocated to patients primarily based on the time that the patient has been on the waiting list and the quality of the match.  If we evaluate these criteria "locally" there's nothing obviously wrong but if we step back and think globally, that is think about what the ultimate goal of the transplant system should be, then the current system is deeply misguided.  Suppose that we want the transplant system to maximize total life expectancy or, as it is known in the literature, to maximize the life-years from transplant (LYFT).

The current system does not maximize life expectancy.  In the current system, a 60 year old patient can be given a 20 year old kidney–that's a waste because the life expectancy of the kidney is longer than that of the patient; it's like putting a new clutch in a car that is rusting away.  If we had 20 year-old kidneys to spare, this wouldn't be a big problem.  But we don't have 20-year old kidneys to spare, so we also give 20-year old patients 60-year old kidneys which means the kidney is likely to die early taking the patient along with it.  If we want to maximize total life expectancy, younger people should get younger kidneys.

Here is a simple example to illustrate the principle.  Suppose that the life expectancy of both patients and kidneys is 75 years of age so everyone dies when they are 75 or when their kidney is 75, whichever comes first.  Thus, if we allocate the 20 year old kidney to the 60 year old patient and vice-versa we gain a total of 30 years of life expectancy.

Kidney
Age
Patient Age Life Years
20 60 15
60 20 15
30 years Total

But if we allocate the 60 year old kidney to the 60 year old patient and the 20 year old kidney to the 20 year old patient we more than double life expectancy to 70 years in total.

Kidney
Age
Patient Age Life Years
60 60 15
20 20 55
70 years Total

It's not just age that matters, it turns out that the longer a patient has been on dialysis the less is their life expectancy after transplant (dialysis stresses the body so the sooner we get someone a transplant the better).  Although it may seem unfair, if we want to maximize total life expectancy we are doing the wrong thing by giving more points to patients who have been on the list longer.  

An optimized allocation system that took into account these considerations would increase total life
expectancy (modestly but significantly, about 11,500 extra life years) but it wouldn't benefit every individual.  Maximizing life expectancy would shift organs away from older people and people who have been on the waiting list a long time towards younger people.  As a result, some patients have argued that LYFT is unfair.  The Office of Civil Rights is even asking whether LYFT might violate age discrimination laws.  

But consider, would the older patients have objected to LYFT when they were younger?  If not, shouldn't their objections be discounted?  More formally, consider how people would vote behind a veil of ignorance.  By definition a LYFT approach maximizes total life expectancy, so without knowing the specifics of who you are or when you might need a transplant it's likely that behind a veil of ignorance just about everyone would favor LYFT.  Thus, in my view LYFT is a fair and ethical system. 

Here are previous MR posts on kidney transplant policy.

Comments

Comments for this post are closed