Feasibility and clinical associations of remote monitoring in continuous ambulatory peritoneal dialysis: a prospective single-centre observational pilot study

Background Remote patient monitoring (RPM) is increasingly used in automated peritoneal dialysis, whereas evidence in continuous ambulatory peritoneal dialysis (CAPD) remains limited. This pilot study evaluated a CAPD-adapted RPM workflow and explored clinical and healthcare-utilization outcomes. Methods This prospective, uncontrolled, single-centre observational study followed 15 adults receiving stable CAPD during July–November 2025, with January–June 2025 as the historical reference. Daily clinical and treatment data were transmitted through MyPD and the Sharesource CAPD portal. Primary outcomes were feasibility, the daily-report transmission rate, and safety. Secondary exploratory outcomes included blood pressure (BP), mean absolute deviation from target weight, time to outpatient assessment, healthcare utilization, complication management, and satisfaction. Clinical comparisons were descriptive; healthcare-event rates were compared using exact Poisson methods. Clinical imaging was exploratory. Results All 15 participants completed follow-up. The daily-report transmission rate was 97% according to platform reports; non-receipt reflected technical problems or omitted patient submissions.No clinical harm attributed to RPM was reported; no technique failures or transfers to haemodialysis occurred. All participants completed the satisfaction assessment; the reported mean score was 4.6 ± 0.4/5. Historical cohort-level home BP was 138 ± 10/82 ± 7 mmHg, compared with 131 ± 7/77 ± 5 mmHg during the final observation month; only aggregate data were used, precluding paired analysis. Fluid-related events, early peritonitis, and constipation-related drainage problems were reported in seven, two, and four participants, respectively, comprising 13 distinct patients. Unscheduled visits numbered 25/90 patient-months historically and 15/75 during RPM (incidence-rate ratio 0.72; 95% confidence interval 0.35–1.42; p = 0.344). Hospitalizations numbered four/90 and one/75 patient-months (0.30; 0.01–3.03; p = 0.385). Conclusions The workflow was feasible, with a high reported daily-report transmission rate and favourable satisfaction ratings in this small, digitally selected cohort. The absence of reported RPM-related clinical harm does not establish safety. Historical clinical comparisons remain exploratory, and sparse healthcare events produced imprecise estimates. The uncontrolled design precludes causal inference. Larger controlled multicentre studies are required.