Early Start Meets Remote Monitoring: A Synergistic Strategy to Reduce Hospitalization and Peritonitis Rates in Automated Peritoneal Dialysis

Technical Summary: Early Start Meets Remote Monitoring in Automated Peritoneal Dialysis

Problem Statement
Peritoneal dialysis (PD) is a standard renal replacement therapy, but traditional protocols mandate a 4–6 week catheter maturation period before initiation. This delay often necessitates emergent hemodialysis for patients requiring urgent support. Early-start peritoneal dialysis (ESPD), defined as initiating dialysis within 14 days of catheter implantation, addresses this barrier but raises concerns regarding catheter complications (e.g., leaks, hernias) and suboptimal dialysis adequacy during the immediate post-implantation period. Furthermore, while Automated Peritoneal Dialysis (APD) improves quality of life, managing it remotely to prevent complications like peritonitis and unplanned hospitalizations remains a challenge. There is a lack of evidence regarding the interaction between early-start timing and the use of Remote Patient Monitoring (RPM) technology on patient and catheter outcomes.

Methodology
This study employed a single-center, retrospective cohort design involving 218 adult patients who commenced APD between June 2008 and November 2025 at Taipei Medical University. The cohort was stratified into four groups based on two binary variables:

1. Start Timing: Early Start (ESPD, ≤14 days post-implantation) vs. Non-Early Start (conventional start, >14 days).
2. Monitoring Modality: Remote Patient Monitoring (RPM, introduced October 1, 2020, using the Sharesource cloud-based platform) vs. No RPM (conventional outpatient/telemedicine follow-up).

The four resulting groups were: ES+RPM ($n=54$), ES+noRPM ($n=81$), noES+RPM ($n=41$), and noES+noRPM ($n=42$).

Primary outcomes included exit-site infections (ESI), peritonitis, and catheter survival. Secondary outcomes included tunnel infections, mechanical complications (leaks, migration, hernia), all-cause mortality, and hospitalization rates (catheter-related and all-cause) expressed per 100 patient-years. Statistical analysis utilized logistic regression for binary outcomes and Poisson regression for count data (hospitalizations), adjusting for age, sex, and dialysis vintage. Survival analyses employed Aalen-Johansen estimators for catheter survival (treating death/transplant as competing events) and Cox proportional-hazards models.

Key Contributions

• Synergistic Analysis: This is the first study to specifically examine the interaction between early-start timing and remote APD monitoring, creating a four-group stratification to isolate the effects of each variable and their combination.
• Safety Validation of ESPD in APD: The study provides evidence that ESPD is safe even in patients with lower baseline serum albumin (a marker of acute uremia), showing no compromise in infectious or mechanical complication rates compared to conventional starts.
• Quantification of RPM Impact: The research isolates the independent contribution of RPM, demonstrating its association with reduced hospitalizations and mortality across both early-start and conventional-start populations.

Results

• Baseline Characteristics: Groups were comparable in age, sex, BMI, and comorbidities. ESPD patients had significantly lower serum albumin (3.40 vs. 3.78 g/dL, $p<0.001$), reflecting their more acute clinical presentation.
• Infectious Complications: Peritonitis incidence differed significantly across groups ($p=0.001$), with the lowest rates in RPM groups (ES+RPM: 31.5%; noES+RPM: 29.3%) compared to non-RPM groups (ES+noRPM: 51.9%; noES+noRPM: 64.3%). While unadjusted peritonitis rates were lower with RPM, adjusted logistic regression did not show statistically significant differences, likely due to limited statistical power.
• Hospitalization: RPM was strongly associated with reduced hospitalization.
  • All-cause hospitalization: Rates were markedly lower in RPM groups (ES+RPM: 14.69/100 PY; noES+RPM: 12.93/100 PY) compared to non-RPM groups (ES+noRPM: 111.56/100 PY; noES+noRPM: 135.35/100 PY; $p<0.001$).
  • Adjusted Analysis: Compared to the noES+noRPM reference, ES+RPM had a significantly lower incidence rate ratio (IRR) for all-cause hospitalization (0.141; 95% CI 0.106–0.187) and catheter-related hospitalization (0.221; 95% CI 0.121–0.403).
• Mortality: Adjusted mortality was significantly lower in both RPM groups compared to the noES+noRPM reference (ES+RPM: OR 0.165; noES+RPM: OR 0.110; both $p<0.001$).
• Catheter Survival: Unadjusted log-rank testing showed no significant difference in catheter survival across groups ($p=0.989$). However, adjusted Cox regression revealed that the ES+RPM group had a significantly lower hazard of catheter removal compared to noES+noRPM (HR 0.353; 95% CI 0.129–0.969; $p=0.043$).
• Mechanical Complications: No significant differences were found in catheter leaks, migration, or hernias across groups.

Significance and Claims
The authors conclude that ESPD on APD is a safe strategy that does not compromise patient survival or increase complication rates, even in nutritionally vulnerable patients with lower albumin. The study posits that RPM is independently associated with substantially lower hospitalization rates and reduced peritonitis incidence.

The paper claims that integrating RPM with early-start programs should be considered a standard care model for incident APD patients. The data suggests that the combination of early initiation and remote monitoring (ES+RPM) offers a protective effect on catheter longevity and significantly reduces the burden of hospitalization and mortality. The authors emphasize that while the study is retrospective and single-center, the magnitude of the reduction in hospitalization and the safety profile of early starts support the adoption of this synergistic strategy. They acknowledge limitations regarding potential unmeasured confounding and secular trends but assert that the findings warrant the integration of these technologies into standard PD protocols.