ODYSSEEODYSSEE Kidney Health Trial

INTRODUCTION

Severe CKD is defined as a risk of greater than 10% for progressing to RRT \[home hemodialysis (HHD), home peritoneal dialysis (HPD), and transplantation\] within 2 years. There is a need to improve access to CKD self-care counselling and RRT education for patients with severe CKD. Trials of CKD self-care education have achieved therapeutic benefits with moderate to high patient-provider contact. There is high potential for a trial of digital counselling for CKD self-care and RRT education to enhance patient health and quality of life.

HYPOTHESES

The primary hypothesis is that ODYSSEE-KH versus usual care (UC) will significantly increase the incidence of home RRT, measured by a composite index of HHD, HPD, and preemptive kidney transplant at trial completion (median = 19 months; range: 12 to 27 months).

The secondary hypothesis is that ODYSSEE-KH for CKD self-care and RRT education improves Home RRT, RRT preparation, annual hospitalization rate, engagement with CKD self-care resources at months 6 and 12 and trial completion and scores on outcome measures.

RECRUITMENT

Patients diagnosed with CKD who are 18 years of age or older were recruited from University Health Network (UHN), Sunnybrook Hospital, Scarborough Health Network, and The Ottawa Hospital.

DESIGN

ODYSSEE-KH is a double-arm, parallel-group, randomized controlled trial that has assessments at baseline, months 6 and 12, and trial completion (median = 19 months; range: 12 to 27 months). This is a single-blind design with research personnel masked.

ODYSSEE-KH combines automated digital counselling of CKD self-care with renal replacement therapy (RRT) education. UC enhances the standard of usual care by providing patients with conventional digital CKD education. Over 27 months, patients will be emailed on a weekly basis with a digital link to log on to their respective program using a password-protected, personal account.

ANALYSIS

Separate GLMs will evaluate if Digital Counselling versus UC is independently associated with outcomes at months 6 and 12 and trial completion (median = 19; range: 12 to 27 months). Dependent variables include the KDQOL-SF, SF-36, EUROIA, PHQ-9, GAD-7, MIDLS, ESSI, PWB, BMPN, AI, as well as a modified SEMCD-6. Multivariable models will adjust for baseline assessments of each outcome and potential baseline covariates (noted above). In all GLMs, significant interactions will be followed by subgroup analyses with Bonferroni post hoc tests.