Geographic Disparities in Lung Transplant Rates

Martin Kosztowski; Sheng Zhou; Errol Bush; Robert S. Higgins; Dorry L. Segev; Sommer E. Gentry


American Journal of Transplantation. 2019;19(5):1491-1497. 

In This Article


In this national study of lung transplant recipients, we found that, even after adjusting for LAS and blood type, lung transplant rates varied considerably depending on a candidate's DSA. The IRR ranged as high as 21.73 meaning that there existed a pair of DSAs that had a 21.73-fold difference in lung transplant rate after adjusting for LAS and blood type. Candidates with the same LAS and blood type in 2 different DSAs were expected to have a 2.05-fold difference (MIRR=2.05) in lung transplant rate. An increase in LAS from 38–42 to 42–50 was associated with a 1.54-fold increase in LT rate. This effect is less than the MIRR, so moving from one of these LAS categories to the next would have had less impact on LT rate than moving from one DSA to another.

A number of other studies explored geographic disparities in lung transplantation.[13] For each lung transplant, there are on average 5.96 candidates in each UNOS region with a higher LAS who are bypassed for a transplant when the local DSA is used as the first unit of allocation.[14] Low local lung availability, defined as the ratio of donor lungs to waitlist candidates in a DSA, was associated with lower lung transplant rates and higher mortality/removal from the waitlist.[15] This study uses a superior metric, the median incidence rate ratio (or MIRR) of LT across DSAs, to quantify geographic disparity. MIRR has been applied previously to quantify geographic disparity in kidney and liver transplantation (M.G. Bowring, E.H. Chow, X. Luo, et al., unpublished data, 2018),[6] and has a straightforward interpretation as the median factor by which LT rate would change if the candidate were listed in a different DSA. Because the MIRR is an IRR, it is directly comparable to the IRR of changes in LAS category and blood type. No other metric allows us to make such comparisons. Knowing the degree of geographic disparity in access to transplant helps patients and physicians decide if they want to list in their local DSA or improve their chances by listing in multiple DSAs.

In New York City, the adjusted LT rate for LAS 50–100 was 3.20 LT per person-year, whereas the LT rate in the neighboring DSA in New Jersey was 12.49. This is an almost fourfold difference in transplant rate for patients in the highest LAS category (50–100). Miriam Holman had an LAS of over 90,[4] so if she had been listed in her neighboring DSA in New Jersey rather than in New York City, her lung transplant rate would have been significantly higher.

The OPTN was recently challenged in court over its use of DSA as the first unit of allocation. Plaintiffs argued that using DSA as the first unit of allocation was inconsistent with the Final Rule, and that this policy disadvantaged higher priority candidates outside a donor's DSA, who are often in desperate need of a transplant. During an emergency meeting, the UNOS Executive Committee decided to use a 250-nautical mile circle from the donor hospital rather than the DSA as the first unit of allocation. A recent study of liver transplant rates before and after Share 35 found that broader sharing did not reduce geographic disparities, and that a candidate's DSA continued to be a major determinant of access to a liver transplant (M.G. Bowring, E.H. Chow, X. Luo, et al., unpublished data, 2018). Another simulation study showed that broader sharing alone would not reduce geographic disparities in liver transplant, but that combining DSAs into novel regions using mathematical redistricting optimization would.[16] It is not known whether the new 250-nautical mile circle will actually reduce the impact of geography on LT rate, so we intend to study this question as soon as data are available.

There are several limitations to our study. The transplant rate in a DSA is affected by offer acceptance decisions. In the case of 2 DSAs with the same organ availability, the LT rate would be higher in the DSA that is more willing to accept marginal or extended criteria donors. In addition, not all patients have equal access to lung transplant centers. There are 20 states that do not have a single LT center.[17] Patients who have to travel long distances to transplant centers have been shown to have lower transplant rates and may never even be evaluated for a transplant.[18] Our measure of geographic disparity captures only those patients who are evaluated and listed for a transplant. Finally, our study asked whether candidates with similar LAS are being transplanted at different rates across the country, without analyzing outcomes after transplant. An organ allocation policy that reduced geographic disparities would likely decrease the number of patients transplanted at low LAS and increase the number of patients transplanted at high LAS. Because high LAS patients are associated with increased morbidity and mortality following transplantation,[19,20] some would argue that transplanting more high LAS patients is undesirable. The LAS is a weighted score that combines pretransplant mortality and posttransplant survival, and weighting posttransplant survival more heavily could be accomplished with a straightforward change in the calculation of LAS. If geographic disparity is desirable because it helps candidates with low allocation priority get transplanted in some places, then surely the best response is not to preserve disparity but to reorder allocation priority so that the candidates who should be transplanted have a higher rate of transplantation no matter where they list.

In summary, we found that DSA of listing was a major determinant of LT rate for candidates across the country before the emergency lung allocation change. Changing a candidate's DSA would have had a greater effect on the candidate's LT rate than changing LAS categories from 38–42 to 42–50 (Table 2). Using the 250-nautical mile radius as the first unit of allocation might or might not reduce geographic disparities, so continued vigilance is warranted.