There are significant and persistent disparities in access to kidney transplantation and as a result most patients with end stage renal disease receive hemodialysis (HD). HD is unique as it is a treatment performed in a group setting which lends itself to forming social networks. The goal of this research study is to identify and characterize the social networks of HD patients and measure the influence of HD social networks on knowledge, attitudes, and behaviors towards transplantation using repeated surveys and a network targeted educational intervention.

Kidney transplantation is the treatment of choice for end stage renal disease (ESRD). Unfortunately, there are significant and persistent gender and racial disparities in access to transplantation and the successful navigation of transplant evaluation process. The process includes understanding suitability/eligibility for transplantation, obtaining a referral to a transplant center for medical evaluation, completion of medical testing for transplant candidacy, and asking family and friends to donate. Women and ethnic minorities are less likely to be informed of the option of transplantation, referred to a transplant center, and complete the medical evaluation needed to be placed on the kidney transplant waiting list. Modifiable barriers to completing these steps include: lack of knowledge, lack of social support, and inability to discuss transplantation and living donation. As a result chronic hemodialysis (HD) becomes the default treatment. HD clinics are unique social environments in that patients receive their treatments in a group setting which facilitates social networks and the formation of unique micro-communities based on their assigned shift. The robust influences of social networks are well-described in other chronic conditions like obesity, diabetes, and smoking, however, there is scant information on the impact of social networks in HD clinics. Differences in social network structure by gender and race may create critical barriers in the kidney transplant process. White men tend to have diffuse networks which are better for communicating novel information. Women and minorities tend to have clustered/dense networks, which tend to reinforce attitudes and behaviors. It has been speculated that social networks negatively affect transplantation through the clustering of inaccurate information, negative attitudes, and exposure to failed transplants. Our previous observational research, however, suggests that HD patient social network structure can facilitate the completion of pre-transplant steps by providing patients with needed information and support. Further insight into HD social networks has the potential to guide measures to reduce transplant disparities. Our overarching hypothesis is that hemodialysis social networks facilitate the spread of information and behaviors, and that using a targeted social network intervention will improve access to transplantation especially among women and minorities. To address this hypothesis, we will pursue the following specific aims:

Aim 1: Characterize hemodialysis patient social networks by gender and race. Using our previously validated survey, we will model social networks of hemodialysis patients in two independent clinics.

Hypothesis 1: Female hemodialysis patients of all race and ethnicity have dense (higher clustering coefficient) social networks when compared to male HD patients.

Aim 2: Assess the relationship between network structure and knowledge, attitudes, and behaviors towards kidney transplantation. Using the network information from Aim 1, we will analyze the clustering of information, attitudes, and behaviors towards kidney transplantation.

Hypothesis 2: Patients in dense HD clinic social networks have similar attitudes (both positive and negative) about transplantation and behaviors (initiating transplant discussion and completing steps in the kidney transplant process) as compared to patients in diffuse networks.

Aim 3: Compare and contrast the diffusion of knowledge, attitudes, and behaviors regarding transplantation through HD social networks by targeting either the most clustered (as measured by clustering coefficient) or the most central members (as measured by betweenness centrality) of the network to disseminate a transplant education intervention. We will assign every HD clinic shift to one of the two targeting strategies and we will measure the spread of information, attitudes, and behaviors by comparing the targeted patients to the other patients on their shift.

Hypothesis 3: Targeting patients with high clustering coefficient will be the most effective method to spread information and favorable attitudes toward transplantation, as well as completion of steps toward transplantation.

Inclusion Criteria:

• All patients 18 years of age or older, able to consent to participation and respond to the validated survey in English, and who are present in the HD clinics during the initial survey periods, July 2018 to June 2022 are eligible to participate in the baseline network survey.
• Transplant eligible patients will be eligible to participate. All patients who participate will be evaluated via chart review by a study transplant nephrologist for transplant candidacy and, if eligible, the primary nephrologist will be asked whether the patient is an appropriate candidate and reasons they have not been referred. We will defer the transplant referral to the attending nephrologist. Our criteria for transplant ineligibility are nursing home resident, >79 years of age, active or recent malignancy, inoperable coronary artery disease, or comorbid illness that will make it unlikely that the patient will survive >3 years after transplant.

Exclusion Criteria:

• Patients who are unable or unwilling to provide consent or complete the survey in English .

Once the final analysis is completed. De-identified data will be available on request from the investigator once proper institutional review board application has been confirmed.

7/2023

Proposal from researcher

Hemodialysis staff will be unaware as to whether the patient was targeted based on their network centrality or clustering. Survey administrators will be unaware of which patient received the living donor intervention.