Peer Support for Hemodialysis Patients

In 2022, Medicare fee-for-service hospitalizations for patients with kidney failure cost over $10 billion, according to the United Stated Renal Data System.¹ Among the estimated 815,000 individuals living with kidney failure, approximately 58% receive maintenance hemodialysis. The hemodialysis population is particularly vulnerable to high rates of acute care utilization due to factors such as advanced age, multiple comorbidities, dialysis-related complications, and greater susceptibility to infections,¹ factors often compounded by low self-efficacy, poor self-management, and limited social support. Peer mentor interventions, delivered by lay individuals with shared lived experience, have emerged as a promising strategy to reduce acute care use by improving self-efficacy, quality of life, depression, and physical outcomes such as interdialytic weight gain.² Regional and national peer networks, including the National Kidney Foundation's peer mentorship program,³ are already established. However, their effectiveness in reducing acute care utilization among patients receiving maintenance hemodialysis is unknown, highlighting a critical gap in kidney care.

In this issue of JASN, Golestaneh et al. present findings from PEER-HD, a pragmatic, randomized trial evaluating whether a telephone-based peer mentor intervention could reduce acute care utilization among high-risk patients receiving in-center hemodialysis in the Bronx, New York, and Nashville, Tennessee.⁴ The intervention included weekly phone calls between peer mentors and mentees over 3 months, followed by 9–15 months of monitoring for hospitalizations and emergency department (ED) visits. Owing to the coronavirus disease 2019 pandemic, follow-up was shortened to 9 months for participants enrolled after April 2020.

In the primary intention-to-treat analysis, the adjusted incidence rate ratio for hospitalizations and ED visits showed no significant difference between the intervention and usual care groups (adjusted incidence rate ratio, 0.85; 95% confidence interval, 0.64 to 1.15). In addition, participants in the peer mentorship group did not show improvements in dialysis adherence, interdialytic weight gain, or serum albumin levels compared with those receiving usual care. Similarly, no significant differences were found in dialysis knowledge, self-efficacy, communication satisfaction, self-care, coping, social support, or quality of life. However, in a prespecified as-treated analysis, among those who actually received peer support, hospitalizations and ED visits dropped by 40%—a promising finding.

The investigators highlight a potential dose–response effect that may help explain the intervention's effect in the as-treated analyses. Only 55 of the 100 participants assigned to the intervention received more than 20 minutes of peer mentorship. The reasons for limited engagement are unclear but may include mentor-related challenges (e.g., inconsistent training uptake, time constraints), participants' readiness for change, and disruptions caused by the coronavirus disease 2019 pandemic, which likely contributed to higher dropout rates.

Although the trial was not powered to detect differences in as-treated or subgroup analyses, some findings are noteworthy. Subgroup analyses suggest that Black participants benefited from the intervention, a significant observation given their disproportionate burden of kidney failure and highest rates of ED encounters across all kidney replacement modalities.¹ The intervention was especially effective in the Bronx, where participants' more complex sociodemographic and clinical profiles may have increased their receptivity to peer support and amplified the intervention's effect. Notably, 64% of Bronx participants received at least 20 minutes of mentorship, compared with 30% in Nashville. Mentor training was conducted entirely in person at the Bronx, while Nashville used a hybrid model due to pandemic constraints. Furthermore, mentee-reported ratings of mentors were higher in the Bronx.

Although the peer mentorship intervention did not significantly affect the primary outcome, the results trended in a favorable, protective direction. This may reflect limited statistical power, partly due to a higher-than-expected dropout rate. The authors cautiously note a potential cost benefit: Among participants who received at least 20 minutes of mentorship, acute care utilization dropped by 40%, which, if scaled, could translate to an estimated $4 billion in annual savings. To support the broader implementation of similar peer mentorship programs for patients on hemodialysis, the authors have previously detailed the mentor training components, behavioral framework, and evaluation methods.^5,6 However, a few uncertainties remain that, if further explored, could help strengthen the design of future peer mentorship interventions. One key question is whether there was sufficient evidence to hypothesize that peer mentorship alone can reduce acute care utilization. Although previous studies have shown benefits for short-term outcomes in patients with kidney disease,² it may be overly optimistic to expect peer mentorship by itself to significantly reduce acute care utilization. There may be underlying barriers, beyond those discussed in the article, that are not addressed by a program focused primarily on enhancing self-management behaviors. For example, patients may lack reliable transportation to dialysis, medical appointments, or pharmacies. Other unmet health-related social needs, such as food insecurity or housing instability, could also hinder their ability to manage their health effectively. These are challenges that peer mentors may not be equipped to address, as they fall outside the typical scope of their training and role. Thus, interventions that combine self-management support with strategies to address unmet social needs may be more effective. Nonetheless, the authors deserve recognition for developing and testing a comprehensive intervention for a highly vulnerable population.

A second key question is how to improve implementation of the peer mentorship intervention.

This can be more effectively achieved by using implementation science frameworks. There are three categories of frameworks: process frameworks, which outline the steps needed to implement a new intervention; determinant frameworks, which help identify barriers and facilitators to a new intervention's success; and evaluation frameworks, which help measure how successful a new intervention is and if it was implemented as intended (Table 1).^7,8 For implementation outcomes, the authors provide data on two relevant outcomes: feasibility and fidelity. The program was feasible—84% of participants completed it. However, fidelity, or the extent to which the program was delivered as intended,⁹ was low. Only 55% of participants actually received the intended dose of support. This limited engagement highlights implementation challenges that might have been anticipated with earlier application of an implementation framework.

Table 1.

Examples of implementation science frameworks and outcomes

Implementation Frameworks
Framework	Purpose	Examples in the Literature
Process	Specify steps for implementation	Knowledge to action
Determinant	Identify barriers and facilitators that may influence implementation success	CFIR
Evaluation	Specify implementation outcomes to measure	RE-AIM

Implementation Outcomes
Outcome	Description	Available Measurement
Acceptability	Satisfaction with various aspects of the innovation	Survey Interviews Administrative data
Adoption	Uptake, initial implementation	Administrative data Observation Interviews Survey
Appropriateness	Perceived fit, relevance	Survey Interviews Focus groups
Feasibility	Actual fit or utility	Survey Administrative data
Fidelity	Delivered as intended	Observation Checklists Self-report
Implementation cost	Cost-effectiveness	Administrative data
Sustainability	Maintenance, continuation	Case audit Semi-structured interviews Questionnaires Checklists

Table adapted from Damschroder et al.⁷ and Proctor et al.⁹ CFIR, Consolidated Framework for Implementation Research; RE-AIM, Reach, Effectiveness, Adoption, Implementation, Maintenance.

To strengthen future interventions, implementation science frameworks should be purposefully integrated into both the design and execution phases. In addition, a broader evaluation of implementation outcomes, such as acceptability, appropriateness, fidelity, implementation cost, penetration, and sustainability, will be essential to inform intervention adaptation and scale-up (Table 1).⁹ To accelerate kidney health research, future pragmatic trials should thoughtfully incorporate implementation science frameworks when integrating interventions into routine clinical care.

In summary, although the intention-to-treat analysis of the PEER-HD intervention did not demonstrate a significant benefit, the as-treated analysis and certain subgroups, specifically Black participants and those residing in the Bronx, demonstrated reductions in acute care utilization. These findings suggest that peer mentorship may offer meaningful benefits in specific contexts. Further research is needed to clarify the mechanisms through which peer support exerts its effects and improve its implementation. Applying implementation science frameworks can help guide this process by identifying barriers and facilitators, outlining implementation steps, and assessing appropriate implementation outcomes. In the meantime, peer support remains a valuable—and underused—tool in dialysis care.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/F366.

Author Contributions

Conceptualization: Flor Alvarado, Rasheeda K. Hall.

Writing – original draft: Flor Alvarado, Rasheeda K. Hall.

Writing – review & editing: Flor Alvarado, Rasheeda K. Hall.

Funding

R. Hall: National Institute on Aging (R01AG076918). F. Alvarado: National Institute of Arthritis and Musculoskeletal and Skin Diseases (K12AR084224).