Key Points

Peer mentorship did not reduce hospitalizations and emergency department visits in patients receiving hemodialysis overall but was effective in as-treated analysis.Peer mentorship may have improved dialysis knowledge, self-assessment of health, and possibly depressive symptoms.There was potential benefit from peer mentorship among self-identified Black patients and among those enrolled in Bronx, New York.

Background

Patients receiving maintenance hemodialysis are hospitalized frequently, leading to disproportionate cost of care and contributing to high morbidity and mortality.

Methods

To test the effectiveness of peer mentorship to reduce hospitalization rates among patients receiving hemodialysis, we performed a multicenter, pragmatic, randomized, controlled trial. Two hundred patient participants receiving hemodialysis at high risk for hospitalization were enrolled, 140 in Bronx, New York, and 60 in Nashville, Tennessee. Of these, 101 were randomized to the peer mentor intervention, and 99 were randomized to usual care. The intervention consisted of trained mentors, patients receiving hemodialysis, placing weekly telephone calls to their assigned patient participant mentees over a 3-month period. During telephone calls, mentors listened, provided emotional and informational support, and promoted self-management. During the intervention period and up to 15 months later, patient participants were observed for incidence of the primary outcome: monthly counts of unplanned hospitalizations and emergency department visits.

Results

The mean age of all patient participants was 54 (SD 13) years, 100 (50%) were female, 110 (57%) were Black, and 70 (39%) were Hispanic. The adjusted incidence rate ratio (aIRR) of primary outcome was not different between intervention versus usual care in intention-to-treat analysis (aIRR, 0.85; 95% confidence interval, 0.64 to 1.15). Prespecified as-treated analyses, with as treated defined by >20 minutes of telephone contact between mentor/mentee pairs, suggested effectiveness of peer mentoring (aIRR, 0.60; 95% confidence interval, 0.47 to 0.76), and exploratory post hoc analyses demonstrated differences in effectiveness by site and self-identified race of mentees.

Conclusions

Peer mentorship did not significantly reduce the rate of hospitalization or emergency department visits overall, but there may have been effectiveness among the following subgroups: those enrolled in Bronx, New York, and those who self-identified as Black.

Clinical Trial registry name and registration number:

ClinicalTrials.gov 2017-8531; NCT03595748.

Podcast

This article contains a podcast at https://dts.podtrac.com/redirect.mp3/www.asn-online.org/media/podcast/JA...