Abstract: FR-PO505
Cluster Randomized Controlled Trial (CRCT) of Remote Monitoring (RM) of Patients on Automated Peritoneal Dialysis (APD): Impact on Restricted Mean Survival Time (RMST)
Session Information
- Peritoneal Dialysis: Current Topics
November 04, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 702 Dialysis: Home Dialysis and Peritoneal Dialysis
Authors
- Paniagua, Ramón, Instituto Mexicano del Seguro Social, Ciudad de Mexico, DF, Mexico
- Qureshi, Abdul Rashid Tony, Karolinska Institutet, Stockholm, Stockholm, Sweden
- Avila, Marcela, Instituto Mexicano del Seguro Social, Ciudad de Mexico, DF, Mexico
- Ramos, Alfonso, Macrotech Pharmaceutical, Ciudad de Mexico, DF, Mexico
- Lindholm, Bengt, Karolinska Institutet, Stockholm, Stockholm, Sweden
Background
RM of patients (pts) on APD provides new opportunities to prevent complications and improve treatment quality. We analyzed the effects of RM-APD with RMST, a robust and clinically interpretable measure which quantifies the survival curve without assumptions of distribution of events.
Methods
In a CRCT, 21 hospitals were randomly assigned to use either RM-APD (n=10) or conventional APD (n=11) for the treatment of pts with recent onset of peritoneal dialysis. 403 pts in RM-APD and 398 in APD were followed up for at least one year. Primary outcomes were time to first event of Composite Index 1 (all-cause mortality or adverse event (AdEv) or hospitalization), and Composite Index 2 (cardiovascular mortality, or first AdEv or hospitalization related to cardiovascular disease, fluid overload and insufficient dialysis efficiency). We used and compared RMST and Cox proportional regression analysis for statistical analysis.
Results
Overall RM-APD associated with longer RMST resulting in shorter restricted mean time lost. Over 16 months, RM-APD as compared to APD can delay the first occurrence of outcomes by on average: for Composite Index 1 by 0.78 months (p=0.09), and its components mortality index 1 by 1.13 months (p=0.001), AdVe index 1 by 0.82 months (p=0.08) and hospitalization index 1 by 0.68 months (p=0.09), and for Composite Index 2 by 1.12 months (p=0.01), and its components mortality index 2 by 1.09 months (p=0.001), AdVe index 2 by 1.16 months (p=0.007) and hospitalization index 2 by 0.94 months (p=0.005). Cox regression analysis showed broadly similar results. All-cause and cardiovascular deaths, shifts to hemodialysis and unscheduled visits were more frequent among APD group.
Conclusion
Whereas both RMST and Cox analysis show that RM-APD may delay first occurrence of deaths, AdVe and hospitalizations, RMST is more robust as it is free from assumptions of proportional hazards demanded by Cox models.
Funding
- Commercial Support –