Variation in Kidney Transplant Referral: How Much More Evidence Do We Need To Justify Data Collection on Early Transplant Steps?

Referral to a transplant center for a medical eligibility evaluation is the initial necessary step to receiving a kidney transplant among patients with ESKD. The data reporting this critical step are lacking because patient referral information is not captured in national surveillance data. Prior work has shown that barriers to early steps in the kidney transplant process may be different from other downstream steps. For example, barriers to transplant referral may not align with the barriers that referred patients face in accessing the waiting list.1 This suggests that the efficacy of specific interventions in improving referral outcomes will depend on the issues they address. For example, if a major barrier to starting the transplant evaluation is access to transportation, providing more education about transplant may not have any effect on increasing transplant access. Before developing or implementing any interventions to improve transplant access, it is vital to conduct epidemiologic analyses on distinct transplant steps to ensure a thorough understanding of the barriers and facilitators of each step. The study by Kim et al.2 in this issue of JASN reported for the first time on nationwide data of kidney transplant referral information among Canadian provinces, finding that only 17% of incident adult patients with ESKD were referred for kidney transplantation within the first year of initiating chronic dialysis. It is not possible to determine the proportion of patients with ESKD who are eligible for kidney transplantation because this would require that all patients with ESKD undergo the medical eligibility evaluation at a transplant center. Prior data suggest that at least half of patients with ESKD are eligible for transplant,3 making 17% a dismally low proportion of patients who have been referred in a relatively timely manner after the initiation of chronic dialysis. Similar to a prior United States study examining variation in referral among patients with ESKD in Georgia, patient characteristics associated with a lower likelihood of referral among Canadian patients included older age, female sex, and a greater number of comorbid conditions.1,2 However, in contrast to the socioeconomic disparities in access to kidney transplant documented in prior United States studies,1,4,5 there did not appear to be as much of an effect of socioeconomic barriers to referral within the Canadian patient population. Patients residing in a neighborhood with a median annual income of <$50,000 did have lower referral, but the analysis may not have had enough power to detect this difference (in main analyses: hazard ratio 0.88; 95% CI, 0.76 to 1.02; in analyses restricted to younger patients with no comorbidities: hazard ratio 0.76; 95% CI, 0.57 to 1.00). It should also be noted that researchers did not find a lower likelihood of referral among nonwhite Canadian patients. However, approximately 7% of Canadian patients with ESKD were pre-emptively referred for transplant, and whether these patients were more likely to be wealthy or white was not described. Given data from the United States indicating that patients pre-emptively referred for transplant are more likely to be non-Hispanic white with substantially better access to healthcare,6 there may also be differences in the characteristics of pre-emptively referred patients in Canada. Kim et al. also report a threefold variation in transplant referral across the Canadian provinces. As indicated in the study, some of this variation is driven by patient-level factors, such as age, sex, and comorbidities, but it is unlikely that factors at the patient level explain the observed variation in full. Given the complexity of the transplant process, it is likely that cross-level variation may be contributing to the geographic variation in referral rates. For example, ESKD provider-level factors could contribute to geographic variation and were not examined. Unmeasured provider-level factors such as the content or quality of patient transplant education; the time a provider spends educating patients about their treatment options; and factors more specific to the provider such as level of education or training, knowledge about transplantation, and unconscious or conscious bias could contribute to this variation in transplant referral across regions. Additional interesting questions remain. The steps in the transplant process after referral—start and completion of the transplant evaluation at a transplant center and wait listing—were unmeasured in this study. The unexpected association of higher referral in provinces with lower deceased donor transplant rates, and the lack of association with median waiting time, could have been explained by variation in who started and/or completed the transplant evaluation and who was placed on the deceased donor wait list. Like referral, the start of the transplant evaluation is not collected in national surveillance data, in either the United States or Canada. Prior studies have identified patient-reported barriers to completing the transplant evaluation once referred, including communication between patients and providers,7 sociocultural factors,7 and financial concerns.8 Researchers also found that lower patient-perceived general knowledge about transplant and concerns about finding a living donor are associated with delays in starting the transplant evaluation process.8 Moreover, it is likely that, across the 12 centers examined in this study, there are differences in transplant education practices, eligibility criteria for starting evaluation, and specific evaluation-related variables such as medical tests required before wait listing which may drive delays in providers’ decision to refer a patient for transplant evaluation. Where do we go from here? In the United States, variation in transplant access across dialysis facilities has led to broader policy changes in the proposed Calendar Year 2019 ESRD Prospective Payment System, including actions for dialysis facilities to improve patient referrals. This includes quality improvement activities that encourage dialysis facilities to work with transplant centers to improve patient care coordination and the creation of a new wait-listing metric to incentivize transplant access.9 However, it does not appear that this new wait-listing metric will be tied to pay for performance. In addition, there is substantial concern among both dialysis facilities and transplant centers about this measure.10 Variation in transplant rates across dialysis facilities has also been reported in some provinces in Canada, such as Ontario.11 Canada could propose similar monitoring of nephrologist providers, i.e., through benchmarking and notifying providers when they are lower than the national average in the number of patients they refer for transplant. Outside of policy, large-scale, pragmatic trials to improve transplant referrals have shown promise12 and other dialysis facility-based provider educational interventions have also begun in some areas of Canada.13 Before intervention and quality measure implementation, a regular, systematic data collection of key transplant steps—including transplant referral and start of the transplant evaluation process—must be collected nationally. Ideally, this data collection would be funded in part by Health Resources & Services Administration/Organ Procurement and Transplantation Network (e.g., United Network for Organ Sharing) or National Institutes of Health (e.g., United States Renal Data System) in the United States, and continued to be collected by the Canadian Organ Replacement Register through the Canadian Institute for Health Information in Canada. We certainly have enough evidence now that this substantial variation in access to key early steps in transplant necessitates ongoing data collection to help inform future interventions to improve access to kidney transplantation at every step in the process. Disclosures None.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.