Enhanced Advocacy and Health Systems Training through Patient Navigation Increases Access to Living Donor Kidney Transplantation

Abstract

Background:

To date, no living donation program has simultaneously addressed the needs of both transplant candidates and living donors by separating the advocacy role from the candidate and improving potential donor comfort with the evaluation process. We hypothesized that development of a novel program designed to promote both advocacy and systems training among transplant candidates and their potential living kidney donors would result in sustained increases in living kidney donor transplantation (LDKT). To this end, we developed and implemented a Living Donor Navigator (LDN) Program at the University of Alabama at Birmingham.

Methods:

We included adult patients awaiting kidney-only transplant in a retrospective cohort analysis. Using time varying Cox proportional hazards regression, we explored likelihood of living donor screening and approval by participation in the LDN program.

Results:

There were 56 LDN participants and 1,948 non-participants (standard of care). LDN was associated with a 9-fold increased likelihood of living donor screenings (adjusted hazard ratio (aHR): 9.27, 95%CI: 5.97–14.41, p<0.001) and a 7-fold increased likelihood of having an approved living donor (aHR: 7.74, 95%CI: 3.54–16.93, p<0.001) compared to standard of care. Analyses by participant race demonstrated higher likelihood of screened donors and a similar likelihood of having an approved donor among African Americans compared to Caucasians.

Conclusion:

These data suggest that both advocacy and systems training are needed to increase actual LDKT rates, and that LDN programs may mitigate existing racial disparities in access to LDKT.

INTRODUCTION

Kidney transplantation is associated with improved long-term outcomes and is considered the gold standard for treatment of end stage renal disease (ESRD).^1–3 Living donor kidney transplantation (LDKT) is associated with a significant survival benefit over deceased donor transplantation.^3–5 Despite these proven benefits, LDKT has declined in the United States since 2004.⁶ The rationale for this decline appears to be multifactorial.^6–8 A recent survey by the National Kidney Foundation found that 1 in 4 Americans would consider living kidney donation if they knew someone who needed a kidney,⁹ and a survey from the Mayo Clinic demonstrated that 84% of individuals surveyed would donate to a friend or family member and 49% to a complete stranger.¹⁰ Taken as a whole, these results would suggest that many potential donors do not know they are needed, which proves particularly problematic given that lack of knowledge about how to ask someone to donate was the most prevalent barrier to achieving LDKT identified by transplant candidates.¹¹

Various programs have been designed to separate this advocacy role from the transplant candidate, including the Johns Hopkins Live Donor Champion Program and Smartphone app,¹² the Boston-based House Calls Program,^13,14 and the Northwestern culturally targeted website Infórmate.^15,16 While these programs have demonstrated increases in donor inquiries, gains in actual approved donors and subsequent LDKTs have been modest. These results are not surprising given that these programs were designed to overcome transplant candidate concerns and none systematically addressed potential living donor concerns. Formative research among potential African American living kidney donors identified lack of knowledge / understanding of the donation evaluation process to be the most important barrier to living donation.¹⁷ These findings highlight the inherent complexity of the living donor selection process which begins with potential donor inquiry, screening of potential donors for absolute contraindications, evaluation and ultimately approval.¹⁸ The process is time-consuming, overwhelming, and involves complex and frequent interactions with the healthcare system, and as such, many potential living donors withdraw from the process prior to approval and donation.^19–21 Moreover, this formative research identified a previously unmet need of comfort with the donor evaluation process among African American potential living donors, which is significant given that living donation rates and LDKT are lower among African Americans compared to their Caucasian counterparts.^3,4,22,23

Patient navigator programs have been postulated to help overcome the burden of navigating complex medical systems and improve the systems training of participants. Within the cancer literature, the theoretical basis for the success of patient navigation programs has been described as focusing on relationship-building between patients and navigators, to augment social support and reinforce patient-clinician relationships.²⁴ Multiple studies of cancer patients have demonstrated the feasibility and efficacy of navigators in improving outcomes including more efficient use of healthcare systems, cost savings, and improved patient satisfaction with medical processes.^25–27 Studies of navigators for other chronic diseases demonstrate improved processes of care but have failed to examine patient experiences, outcomes, and costs.²⁸ Most recently, the Medical University of South Carolina implemented a patient navigator program to enhance LDKT. Their patient navigators interacted directly with transplant candidates alone to enhance knowledge about benefits of LDKT as well as enhance systems training around healthcare utilization. The program demonstrated an increased likelihood of donor inquiry and screening, but failed to show an increase in total approved donors or actual LDKTs.²⁹

While the results from various living donor programs have been promising, existing programs have primarily focused on addressing transplant-candidate factors by separating the advocacy role from the candidates. To date, no living donation program has simultaneously addressed the needs of both transplant candidates and living donors by separating the advocacy role from the candidate and improving potential donor comfort with the evaluation process. We hypothesized that development of a novel program designed to promote both advocacy and systems training among transplant candidates and their potential living kidney donors would result in sustained increases in LDKT. To this end, utilizing the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework,^30,31 we developed and implemented the University of Alabama at Birmingham (UAB) Living Donor Navigator (LDN) Program, and herein, we report our initial results with regard to reach and effectiveness.

MATERIALS AND METHODS

We conducted a retrospective cohort study to evaluate the impact of the LDN program at our institution. We utilized the STROBE guidelines for reporting an observational study.

Living Donor Navigator Program

The LDN program combines advocacy training adapted from the Johns Hopkins developed Live Donor Champion program¹² with the systems training of the UAB developed Patient Navigator Program,³² and in so doing, simultaneously addresses both transplant candidate (e.g. separating the candidate from the advocacy role) and potential donor related factors (e.g. helping potential donors navigate the donation/evaluation process) implicated in declining living donation rates. The UAB LDN Program utilizes lay navigators from the local community to deliver both the advocacy and systems training. The term “systems training” is intended to encompass the process by which lay citizen navigators educate living kidney donor candidates on the “why, how, and when” associated with necessary testing and physician visits required for completion of the donor evaluation process. The program was first offered to all patients evaluated for kidney transplantation beginning in February of 2017.

Description of Intervention Components

Advocacy Training

A 2-month advocacy training program, modeled after the Johns Hopkins Live Donor Champion program,¹² combines education, advocacy, and instrumental support by pairing candidates with a live donor advocate. The live donor advocate is a friend, family member, or community member trained to spread awareness of ESRD, LDKT, donation, and the candidate’s story throughout their social network. The four educational sessions offered 2 weeks apart consist of didactic and hands-on (e.g. role playing) interactive lessons taught by living donor navigators who have been trained on a wide range of topics including LDKT, donation, anti-coercion training, and social networking (Table 1). They are provided with educational resources regarding LDKT and donation, as well as conversation starters.

Table 1.

Summary of Material Covered in Living Donor Navigator Sessions

Session	Summary of Material Covered
1	Introduction to Kidney Transplantation and Live Donation Administer baseline knowledge assessment Examine topics of renal failure, kidney transplantation, and living donation (including donor risks, benefits, eligibility, costs, and compatibility)
2	Initiating a Conversation with Potential Live Donor Candidates Discuss common concerns associated with asking about donation Discuss effective communication strategies Engage in conversation examples and role-playing Reiterate importance of avoiding coercion
3	Identifying a Social Network Brainstorm social networks Share examples of success stories Assist participants with writing their story Assist participants with creating a Facebook page or writing a letter
4	Living Donor and Recipient Panel/Surgeon and Nephrologist Panel/Recap Panel of prior donors and recipients to share their story Potential donors encouraged to attend Q&A with a transplant surgeon and nephrologist Administer post-class knowledge assessment and program evaluation

Systems Training

In addition to teaching advocacy training to candidates and their advocates, the living donor navigators also engage with candidates’ potential donors. The goal of systems training is to educate potential donors on the “why, how, and when” associated with necessary testing and physician visits required for completion of the donor evaluation process. Navigators first make contact with potential donors after the donors have contacted our transplant center to complete their medical screening. Potential donors who pass screening and are scheduled for evaluation are then contacted by our navigators. Navigators provide these potential donors with additional resources, including social support, knowledge of local resources for lab draws, radiology, etc., as well as answer questions regarding what to expect during the evaluation process. Navigators make frequent contact with potential donors (e.g. both navigator and potential donor initiated conversations) via telephone and email, including reminders about physician and testing appointments. Moreover, the navigators greet potential donors upon arrival on their physician-visit day and guide them through our large academic medical system, ensuring potential donors receive a concierge style service (Figure 1).

Figure 1.

Process map of evaluation for living kidney donation

Data Sources

All data were obtained from TransChart (UAB’s transplant database), transplant administration records, and the electronic medical record. These records track the number, date, and demographics of individuals who are referred for transplantation at UAB and those who contact UAB to become living kidney donors, whether they are screened eligible for donation and/or are approved to donate.

Study Population

We utilized a retrospective cohort of adult (≥18 years), kidney-only patients who were evaluated for kidney transplant at UAB between January 1, 2016-March 1, 2018 (N=2,099). Patients were classified by whether they participated in the LDN program. Data were gathered on living donor outcomes (whether the donor was screened and was approved for donation) and patient outcomes (had a donor screened, had a donor approved for donation, died while waiting, received deceased donor transplant). Patients were excluded if they were listed for or received a non-kidney or multi-organ transplant (N=41), if they initiated the LDN classes prior to their kidney transplant evaluation (N=2), or if they ultimately received a living donor kidney transplant from a donor who was screened prior to the patient’s evaluation date (N=52). Figure 2 presents a study population flow diagram. Donor records were excluded if the donor was screened prior to the patient’s evaluation date. Institutional Review Board approval (protocol 300000266) was obtained for this study.

Figure 2.

Study population flow diagram

Data Analyses

Continuous characteristics of LDN participants vs. standard of care (non-participants) were compared using t-tests or non-parametric equivalents as appropriate, and categorical characteristics were compared using chi-square or Fisher’s exact tests. To measure LDN program effectiveness in advocacy training, the likelihood of having a donor screened (e.g. potential living donor inquiries regarding initiating evaluation process and completes an initial screening tool regarding health and social history) was assessed: patients were followed until the earliest of first donor screening date, death, receipt of a deceased donor kidney transplant, or administrative end of study (September 30, 2018). To measure LDN program effectiveness in systems training, the likelihood of having a donor approved (e.g. donor completes evaluation process and was cleared by the clinical team for donation, either for the intended recipient or in the Kidney Paired Donation program) was assessed: patients were followed until the earliest of donor approval date, death, receipt of a deceased donor kidney transplant, or administrative end of study. Cox proportional hazards regression with a time varying component (to allow LDN participants to contribute follow-up time prior to initiating the class) was utilized to explore the likelihood of donor screening and approval by LDN participation. Variables significant on univariate analysis at p < 0.1 were considered for multivariable analysis, with the most parsimonious model selected by minimizing Akaike’s Information Criterion. We also stratified both models by patient race and explored a 4-level variable for patient race and LDN participation. All analyses were performed using SAS 9.4.

Sensitivity analyses

We limited models to patients who were ultimately added to the waiting list, as patients who were not yet listed may have been less motivated to pursue living donor transplantation. To address potential sources of bias, including selection bias, we also generated models comparing 1) participants to non-participants among those who were offered the program and were interested in participating, 2) patients evaluated before February 1, 2017 when the program was first introduced and those evaluated February 1, 2017 and after, and 3) among those evaluated since the program was first introduced, a) comparing participants to candidates interested in LDN but who were unable to participate, b) comparing participants to candidates not interested in LDN, and c) comparing candidates interested in LDN but who were unable to participate to candidates not interested in LDN (Supplemental Table S1). Inferences were confirmed, and thus models from the primary analyses are presented.

RESULTS

Program Reach

The study population included 2,004 UAB adult (≥18years) patients evaluated for kidney-only transplant. All patients evaluated beginning February 2017 were approached about participating in the LDN program, with 111 completing an interest form and 56 of those patients choosing to participate. Among those who were interested but declined, 100% (n=55) cited distance to our transplant center as the reason. Additional reasons provided included current health state, conflicts with dialysis sessions, and lack of transportation. Participants and non-participants were similar in age and gender. The proportion of African Americans in the participant group was greater than the non-participant group (80.4% vs. 63.9%, respectively, p=0.06). Participants had a shorter dialysis vintage (Median: 0.73years vs. 1.27years, respectively, p=0.006) and lived closer (22miles vs. 85miles, respectively, p<0.001) to UAB compared to non-participants (Table 2).

Table 2.

Demographics by participation

	LDN participants (N=56)	Non-participants (N=1,948)	p
Age at evaluation in years, median (IQR)	56.7 (46.7–61.9)	54.5 (45.0–61.9)	0.69
Male sex, N(%)	26 (46.4)	1,107 (56.8)	0.12
Race/ethnicity, N(%)
Caucasian	10 (17.9)	633 (32.5)	0.06
African American	45 (80.4)	1,245 (63.9)
Hispanic	0 (0.0)	39 (2.0)
Other/unknown	1 (1.8)	31 (1.6)
Time on dialysis at evaluation in years, median (IQR)	0.73 (0.00–1.56)	1.27 (0.36–4.08)	0.006
Distance to UAB in miles, median (IQR)	22.0 (9.8–78.1)	85.0 (38.0–142.5)	< 0.001
Married/partnered, N(%)	30 (53.6)	930 (47.7)	0.39

Program Effectiveness of Advocacy Training – Likelihood of Donor Screening

Among 56 LDN Program participants, 113 donor screenings occurred for a rate of 2.02 screened donors per participant; in comparison, among 1,948 non-participants, 955 donor screenings occurred for a rate of 0.49 screened donors per participant. On unadjusted analyses, only participation in the LDN program (hazard ratio (HR): 7.39, 95%CI: 4.87–11.21, p<0.001) and being married (HR: 1.75, 95%CI: 1.44–2.13, p<0.001) were associated with increased likelihood of donor screening (Table 3). After adjusting for multiple factors, participation in the LDN program remained the strongest predictor of having a living donor screened, and in fact, participation in the program increased this likelihood more than 9-fold compared to standard of care (adjusted HR (aHR): 9.27, 95%CI: 5.97–14.41, p<0.001). This finding persisted independent of race, such that African American participants were 8-fold more likely to have a donor screened than African American non-participants (aHR: 8.47, 95%CI: 5.05–14.20, p<0.001) and 3-fold more likely than Caucasian non-participants (aHR: 3.20, 95%CI: 1.90–5.38, p<0.001) (Table 4).

Table 3.

Results of Cox proportional hazards regression

	Unadjusted			Adjusted
Donor screening	HR	95% CI	p	aHR	95% CI	p
LDN participation	7.39	4.87–11.21	< 0.001	9.27	5.97–14.41	< 0.001
Age at evaluation, per 10 year increase	0.86	0.79–0.93	< 0.001	0.75	0.68–0.82	< 0.001
Male sex	0.84	0.70–1.01	0.06	0.75	0.62–0.92	0.005
Race/ethnicity
Caucasian	REF			REF
African American	0.36	0.30–0.43	< 0.001	0.37	0.30–0.47	< 0.001
Hispanic	0.99	0.58–1.69	0.96	0.93	0.53–1.66	0.82
Other/unknown	0.31	0.16–0.60	< 0.001	0.28	0.15–0.52	< 0.001
Time on dialysis, per 1 year increase	0.81	0.76–0.87	< 0.001	0.85	0.80–0.91	< 0.001
Distance to UAB, per 25-mile increase	1.00	0.98–1.03	0.84
Married/partnered	1.75	1.44–2.13	< 0.001	1.75	1.41–2.18	< 0.001
Donor approval	HR	95% CI	p	aHR	95% CI	p
LDN participation	4.62	2.33–9.26	< 0.001	7.74	3.54–16.93	< 0.001
Age at evaluation, per 10 year increase	0.87	0.73–1.03	0.10	0.73	0.61–0.86	< 0.001
Male sex	1.81	1.19–2.74	0.005	1.61	1.07–2.44	0.02
Race/ethnicity
Caucasian	REF			REF
African American	0.19	0.12–0.30	< 0.001	0.23	0.14–0.38	< 0.001
Hispanic	1.34	0.60–2.99	0.48	1.29	0.58–2.87	0.54
Other/unknown	0.28	0.04–2.03	0.21	0.19	0.02–1.59	0.13
Time on dialysis, per 1 year increase	0.72	0.54–0.95	0.02	0.79	0.61–1.02	0.07
Distance to UAB, per 25-mile increase	1.03	1.00–1.06	0.05	1.05	1.01–1.09	0.01
Married/partnered	2.88	1.88–4.42	< 0.001	2.74	1.74–4.31	< 0.001

Table 4.

Cox regression with 4-level variable for participation and race (among Caucasians and African Americans)

	aHR	95% CI	p
Donor screening
AA non-participant vs. CA non-participant	0.38	0.30–0.47	< 0.001
AA participant vs. CA non-participant	3.20	1.90–5.38	< 0.001
AA participant vs. CA participant	0.31	0.11–0.83	0.02
AA participant vs. AA non-participant	8.47	5.05–14.20	< 0.001
Donor approval
AA non-participant vs. CA non-participant	0.22	0.13–0.38	< 0.001
AA participant vs. CA non-participant	1.83	0.74–4.53	0.20
AA participant vs. CA participant	0.31	0.06–1.46	0.14
AA participant vs. AA non-participant	8.24	3.05–22.27	< 0.001

Program Effectiveness of Systems Training – Likelihood of Donor Approval

Nine living donors were approved to donate among 56 LDN Program participants for an approval rate of 16.1%; in contrast, 100 living donors were approved among 1,948 non-participants for an approval rate of 5.1%. On unadjusted analyses, participation in the LDN program (HR: 4.62, 95%CI: 2.33–9.26, p<0.001), male gender (HR: 1.81, 95%CI: 1.19–2.74, p=0.005), and being married (HR: 2.88, 95%CI: 1.88–4.42, p<0.001) were associated with increased likelihood of donor approval; whereas increased dialysis vintage (HR: 0.72, 95%CI: 0.54–0.95, p=0.02) was associated with decreased likelihood of donor approval (Table 3). After adjusting for multiple factors, participation in the LDN program remained the strongest predictor of having a living donor approved, such that participation in the program increased this likelihood more than 7-fold compared to standard of care (aHR: 7.74, 95%CI: 3.54–16.93, p<0.001). This finding also persisted independent of race, such that African American participants were 8-fold more likely to have a donor approved than African American non-participants (aHR: 8.24, 95%CI: 3.05–22.27, p<0.001) (Table 4).

DISCUSSION

In this study examining the reach and effectiveness of the UAB LDN Program, we demonstrated that implementation of an LDN Program was associated with a 9-fold increased likelihood of living donor screenings and a 7-fold increased likelihood of having an approved living kidney donor among program participants compared to standard of care, suggesting that both advocacy and systems training are needed to increase actual LDKT rates. Importantly, the LDN Program was also demonstrated to have exceptional reach and effectiveness among African Americans, a population that historically has faced disparities in access to LDKT. In fact, African Americans represented 80.4% of program participants. Furthermore, compared to African American non-participants (standard of care patients), African American LDN Program participants were 8-fold more likely to have a donor screened and 7-fold more likely to have a donor approved. Moreover, these analyses by participant race demonstrated higher likelihood of screened donors and a similar likelihood of having an approved donor among African Americans compared to Caucasians. Taken as a whole, these data suggest that the UAB LDN Program, a first-of-its-kind, has the potential to increase LDKT and importantly mitigate existing disparities in access to this life sustaining therapy.

LDKT rates have been linked to population health and minority prevalence such that regions with a greater prevalence of minorities have corresponding worse population health and as a result face significant disparities in access to LDKT.³³ These data highlight that the greater comorbid disease burden among African American populations, namely diseases such as diabetes and hypertension that are known risk factors for ESRD, negatively impacts LDKT rates and may in part explain disparities in LDKT among African Americans compared to their Caucasian counterparts.^34–38 This is not surprising given that living donor-recipient race is almost 95% concordant as most individuals turn to their social networks to identify living kidney donors.³⁹ Taken in context it is likely that African American transplant candidates will need a larger pool of potential living kidney donors to be screened in order to identify a medically suitable donor that can be approved for living donation.^40,41 Results from the UAB LDN Program suggest that the program provides a platform for African Americans to do just that. Analyses by race indicated higher likelihood of donor screenings and similar likelihood of donor approval among African Americans compared to Caucasians. In other words, the UAB LDN Program has demonstrated effectiveness in not only improving LDKT, but in particular may serve to mitigate racial disparities in access to LDKT.

In addition to assessing program effectiveness, assessment of program reach is essential to ensuring more widespread adoption. Importantly, the UAB LDN Program had a significant impact among African American patients awaiting kidney transplant, such that more than 80% of program participants were African American. Programmatic investment by the African American community without question enhanced the efficacy of the program in this historically disadvantaged patient population. While the LDN reach helped to mitigate disparities in access to LDKT among African Americans, the program did introduce an unintended geographic disparity in access, such that on average program participants lived 63 miles closer to UAB than non-participants. Notably 100% of patients approached to participate in the program expressed interest; however, 44% declined citing distance as the reason. Future work to expand both reach and adoption is necessary to overcome this inadvertent geographic disparity. Utilization of telehealth services has previously been applied in both medical and public health settings to address geographic disparities, with notable success.^42–46 As such, ongoing work to expand the UAB LDN Program via telehealth modalities is underway.

While the UAB LDN Program, a first-of-its-kind, has demonstrated excellent reach and effectiveness and in particular is the first program to show enhanced effectiveness among African Americans, the program does have several limitations. The program was developed at a single center in the rural, deep South, and as such generalizability to centers in other regions of the country may be limited. However, utilizing the RE-AIM framework it is likely that the existing program can be tailored to meet the needs of other diverse areas in the US.³¹ In addition, the program in its present form is geographically limited in its delivery, such that participants are required to travel to the transplant center to participate in the advocacy training. Telehealth delivery is a plausible solution to this limitation and is actively being pursued at UAB. Finally, we did not plan the study prospectively prior to program roll out, and as such, did not collect certain data points that may have further informed program fidelity and results. For example, while we know the navigators interacted with each living donor at specified time points (see Figure 1), we do not have data on frequency of interactions beyond those listed or information on interest in living donation among the participants’ social networks at the time of evaluation. As such, there is a strong possibility of residual confounding in our retrospective cohort study. However, the inferences reported were robust to multiple sensitivity analyses, and thusly, we believe data from the LDN program provide new insight into an innovative strategy for increasing living donation and transplantation.

In conclusion, the UAB LDN Program is highly effective and has tremendous reach. It is the first program designed to promote LDKT that has demonstrated sustained increases in both donor screenings and approvals, and importantly, has proven effective among African Americans. This first-of-its-kind program has tremendous promise for mitigating disparities in access to LDKT among African Americans. Future work designed to overcome geographic limitations in participation is needed, but in the setting of modern technology, geography is likely easily overcome with the advent of a telehealth approach to the UAB LDN model.

ABBREVIATIONS

aHR

adjusted hazard ratio

ESRD

end stage renal disease

LDKT

living kidney donor transplantation

LDN

Living Donor Navigator

RE-AIM

reach, effectiveness, adoption, implementation, maintenance

UAB

University of Alabama at Birmingham