In early 2008, the UNOS requested comment on a proposal to improve the way kidneys are allocated to recipients. Specifically, it wanted to add an estimate of how long a recipient is likely to live after a transplant, called life years from transplant (LYFT), when assigning organs to patients. LYFT compares predicted survival times after a transplant to remaining on dialysis. The primary effect would be to allocate more kidneys to younger recipients and fewer to older patients.

The current UNOS rules for allocating kidneys from deceased donors is beyond the scope of this blog post. It is sufficient to say the rules are rather complex and not particularly equitable. The important fact is that since there are many more patients waiting for transplants than there are organs available, the allocation method is actually a way to ration organs, ensuring some patients are more likely to get a kidney than others. Patients who don’t get transplants will die while on dialysis.

Under the current allocation rules, if the patient is an adult, the age of the patient and the age of the donor are not considered when making a match. (Patients under the age of 18 are considered juveniles, and get priority to kidneys and their donors must be under 35 years of age.) This means a kidney from a 70-year-old donor may be given to a 25-year-old patient, or the kidney from a 25-year-old donor may be given to a 70-year-old patient. Since kidneys from young donors survive longer than those from older donors, this allocation does not maximize LYFT.

To increase expected LYFT, donor age and patient age should be more closely matched. This will improve outcomes for younger patients with little to no harm to older patients. However, donors tend to be young (most are accident victims) while patients tend to be old (kidney disease is often a chronic condition related to Type 2 diabetes or hypertension). Since there is a shortage of kidneys, this means that matching donors with patients by age will increase the number of young patients getting transplants and reduce the number of older patients getting one.

In response to this proposal, several researchers provided analyses comparing the current allocation scheme to the proposed scheme. For instance see Cur. Opinion Nephr. Hypertens. Nov 2007 (subscription required) and Amer. J. Transpl. Aug 2008. Some groups were quite adamant that more research was needed prior to acceptance, see Amer. Soc. Transpl. Surgeons comment and the UNOS response. These researchers spend a lot of time arguing over the predictive power of certain demographic and medical variables and in comparing outcomes for cohort groups.

However, the most interesting comment came from Segev, Gentry, and Montgomery in Amer. J. Transpl. Oct 2007. They say that changing the allocation method for deceased donor kidneys would result in behavioral changes among patients on the waiting list that would reduce the expected benefit of the change. Specifically, if their wait times shorten, younger adults would be less likely to look for a live donor. Conversely, older adults would be more likely. Since transplants from live donors provide much better outcomes than from deceased donors, the final lifetime benefits could be much lower than what is originally expected. This is a clever insight that looks past all the statistical data on outcomes to see the human aspect of the problem. (Segev and Gentry were mentioned in an earlier blog post as the original researchers that worked to improve the efficiency of kidney exchanges.)

As of Nov 2009, the UNOS is still developing its new allocation policy. Meanwhile, it has run into another unintended consequence problem to its current policy. Patients with high levels of antibodies for human leukocyte antigens (HLA) as measured via the calculated panel reactive antigen (PRA) have a harder time finding a matching donor than other patients. To compensate they are given extra points to put them higher on the UNOS waiting list. Some transplant programs have started using a desensitization protocol to prepare the patient for surgery (more about this in a future blog post) to help them avoid organ rejection. However, once the treatment is complete, these patients are no longer eligible for the points. Without those points they are no longer at the top of the list, and so can’t immediately get a kidney. Desensitization therapy is very expensive ($15,000 a day for 5 days) and must be done just prior to surgery or else it wears off. The allocation rules need to account for these patients.

Note that in the short run, changing the organ allocation policy does nothing to shorten the waiting list or the median wait time. In the long run, it may reduce the total number of transplants a patient needs during a lifetime, which will indirectly shorten the wait list. However, if the effect described by Segev, et al. occurs, the overall effect is unknown. It would be better if there was a way to eliminate the waiting list. But that’s a long road that involves better health care for patients with a predisposition to diabetes and hypertension, better organ recovery from deceased donors, more live donors, and stem cell research for artificial kidneys.

[Update1: Added a paragraph explaining why the new rules will shift more transplants to younger patients.]

[Update2: Corrected last paragraph. Changing the allocation policy can change wait times.]