Remote Treatment Monitoring on Hospitalization and Technique Failure Rates in Peritoneal Dialysis Patients

Abstract

Background 

An integrated kidney disease healthcare company implemented a peritoneal dialysis (PD) remote treatment monitoring (RTM) application in 2016. We assessed if RTM utilization associates with hospitalization and technique failure rates.

Methods 

We used data from adult (age ≥18 years) patients on PD treated from October 2016 through May 2019 who registered online for the RTM. Patients were classified by RTM use during a 30-day baseline after registration. Groups were: nonusers (never entered data), moderate users (entered one to 15 treatments), and frequent users (entered >15 treatments). We compared hospital admission/day and sustained technique failure (required >6 consecutive weeks of hemodialysis) rates over 3, 6, 9, and 12 months of follow-up using Poisson and Cox models adjusted for patient/clinical characteristics.

Results 

Among 6343 patients, 65% were nonusers, 11% were moderate users, and 25% were frequent users. Incidence rate of hospital admission was 22% (incidence rate ratio [IRR]=0.78; P=0.002), 24% (IRR=0.76; P<0.001), 23% (IRR=0.77; P≤0.001), and 26% (IRR=0.74; P≤0.001) lower in frequent users after 3, 6, 9, and 12 months, respectively, versus nonusers. Incidence rate of hospital days was 38% (IRR=0.62; P=0.013), 35% (IRR=0.65; P=0.001), 34% (IRR=0.66; P≤0.001), and 32% (IRR=0.68; P<0.001) lower in frequent users after 3, 6, 9, and 12 months, respectively, versus nonusers. Sustained technique failure risk at 3, 6, 9, and 12 months was 33% (hazard ratio [HR]=0.67; P=0.020), 31% (HR=0.69; P=0.003), 31% (HR=0.69; P=0.001), and 27% (HR=0.73; P=0.001) lower, respectively, in frequent users versus nonusers. Among a subgroup of survivors of the 12-month follow-up, sustained technique failure risk was 26% (HR=0.74; P=0.023) and 21% (HR=0.79; P=0.054) lower after 9 and 12 months, respectively, in frequent users versus nonusers.

Conclusions 

Our findings suggest frequent use of an RTM application associates with less hospital admissions, shorter hospital length of stay, and lower technique failure rates. Adoption of RTM applications may have the potential to improve timely identification/intervention of complications.

Visual Abstract

Introduction

The modality of peritoneal dialysis (PD) is suggested to associate with favorable outcomes compared to in-center hemodialysis (HD); however, technique failure is common and adjusted rates for the hospital length of stay tends to be longer in PD (^12–3). In the United States, patients on home dialysis typically have monthly clinic visits where their clinical status, assessment of treatment quality, and adherence is assessed based on written treatment records and self-reported complications (⁴,⁵). This monitoring process poses limits to the clinician’s view of complications and clinical needs, which can impede timely medical decisions.

Remote monitoring systems may improve the care team’s ability to actively identify urgent concerns in patients on PD and react in a timely manner with diagnostic examinations, interventions, or patient education (⁶). Remote monitoring systems broadly include an array of technologies and processes in the areas of telemedicine/telehealth such as telephonic/video assessments, connected health sensors, and health record portals (⁶). Prior reports of various types of remote monitoring systems tested in the PD population suggest they associate with improvements in patient satisfaction, quality of life, nutrition, exit-site infection rates, peritonitis rates, and hospitalization rates (^{789101112–13}). However, patient groups included in the evaluations of remote monitoring systems to date are small and the findings may not be generalizable.

A large integrated kidney disease healthcare company started using a remote treatment monitoring (RTM) application throughout its dialysis organization in the United States since October 2016. The provider constructed and integrated the RTM system into its clinical systems as a quality improvement process. The RTM is an online portal–based treatment record application, whereby patients can create an online account and record the details of individual PD treatments, associated vital signs, and complications. We aimed to assess the associations between the level of utilization of the RTM application and hospitalization and technique failure rates.

Materials and Methods

General Design

We performed a retrospective analysis to assess the longitudinal associations between the frequency of use of the RTM in patients on PD and outcomes after deployment in a dialysis organization in the United States. This analysis was performed under a protocol that was reviewed by New England Independent Review Board who determined it was an exempt assessment of existing patient data from a quality improvement process, which was anonymized and did not require informed consent per title 45 of the United States Code of Federal Regulations part 46.102 (1-9652-1; New England Independent Review Board, Needham Heights, MA). The analysis was conducted in adherence with the Declaration of Helsinki.

Patient Population

We used data from adult (age ≥18 years) patients on PD treated anytime during October 1, 2016–May 31, 2019 at the dialysis organization (Fresenius Kidney Care, Waltham, MA) of a large integrated kidney disease healthcare company (Fresenius Medical Care, Bad Homburg, Germany). We included data from all patients on PD who: (1) registered online and created an RTM account on or before May 31, 2018; (2) were treated continuously with PD for at least 30 days after registration; and (3) were not hospitalized within 30 days after registration. We excluded patients with: (1) a body mass index >65 kg/m², or (2) missing data for any covariates used for the adjustment of the analysis (refer to the Variables section of the Materials and Methods).

RTM Application

The RTM named the “PatientHub” is a health record portal technology available to all patients on PD treated by the dialysis organization in the United States who have internet access. The RTM can be used through either a secure personal website portal or through a mobile device application. After the creation of a personal RTM account and profile, patients can: (1) view their dialysis orders, laboratory results, concomitant medications, and supply orders; and (2) document their daily PD treatment data, vital signs, and complications. A schematic of the prior monitoring process and RTM process is shown in Figure 1.

Figure 1.: Schematic of paper and electronic charting workflows. (A) Schematic of how the paper flowsheets were submitted manually for review by clinicians once a month. (B) Schematic of how electronic flowsheets are submitted via PatientHub RTM for daily review by clinicians.

The daily PD treatment data documented in the RTM include the treatment date, total ultrafiltration from cycler, dialysate type, bag size and number of bags used, medications added to the dialysate, as well as details on any manual exchanges performed. The daily clinical data documented include weight, BP, pulse, temperature, blood glucose, and confirmation of routine exit-site care. The data on complications documented daily include drain/fill, PD fluid, or exit-site issues. Once submitted by the patient, the data are displayed in the electronic medical record for the care team to review. The care team is instructed to review patient RTM entries on at least a daily basis on business days, but the timing of the daily review is at the discretion of the clinician. A preview of the PatientHub RTM application is shown in Figure 2.

Figure 2.: Preview of the PatientHub RTM application. PD, peritoneal dialysis.

Variables

The dependent variables were hospital admission counts per patient year (PPY), hospital days PPY, and sustained technique failure counts (PPY) from 30 days after RTM registration to 3, 6, 9, and 12 months of follow-up. Sustained technique failure was defined as PD complications that required patients to receive >6 consecutive weeks of treatment with HD.

The independent variables were the frequency of RTM use during the baseline period 30 days after online registration. We defined nonusers as patients who never documented any treatment record in the RTM within 30 days of registration, moderate users as patients who documented one to 15 treatment records in the RTM within 30 days of registration, and frequent users as patients who documented >15 treatment records in the RTM within 30 days of registration. These cut points for classification of the frequency of baseline RTM use were chosen based on the distribution of data; most patients tended to use the RTM >15 times or not at all within 30 days of registration.

Covariates used for the description of the baseline patient characteristics and for adjustment of statistical models included age, sex, race, ethnicity, dialysis vintage, alcohol use, urbanicity of residence, education level, congestive heart failure, diabetes, ischemic heart disease, albumin, residual kidney function, and weekly Kt/V. The most recent categoric variables to the date of registration were recorded. The mean value of continuous variables during 30 days from RTM registration was computed and recorded, with exception of residual kidney function which was determined from the most recent value.

Statistical Analyses

Adjusted Poisson regression models were constructed to assess the associations in hospital admission and day rates at the 3, 6, 9, and 12 months of follow-up periods for nonusers (reference) versus moderate users and frequent users. The number of days patients were actively receiving dialysis during the follow-up period (patient exposure days) was used to calculate the hospital admission and day rates PPY.

An adjusted Cox regression model was constructed to assess the associations in sustained technique failure rates at 3, 6, 9, and 12 months of follow-up periods for nonusers versus moderate users and frequent users. A Kaplan–Meier curve with log-rank tests was used to assess the associations in the time on PD modality over the 12 month follow-up for nonusers versus moderate users and frequent users.

For hospitalization and technique failure rates, we censored data on patients who died, received a transplant, or where discharged from the provider’s clinic network at end of each respective 3-month period of the 12-month follow-up.

Similar to the previous analysis, adjusted Cox regression methods were used in a subanalysis of survivors of the 12-month follow-up period to assess the associations in technique failure rates and the time on PD modality in a group of patients who had equivalent opportunities to experience a technique failure. This subanalysis of survivors excluded all attrition (patients who died, received a transplant, or where discharged from the provider’s clinic network).

Results

Patient Characteristics

In a population of 36,577 patients on PD treated at a large dialysis provider during the analysis period, 11,079 adult patients were treated by PD for 30 days without being hospitalized during baseline and completed the online registration, creating an RTM profile. Among this cohort, we included data from 6343 patients at 931 clinics, and excluded 4736 patients due to only incomplete/missing data on covariates used in the adjustment of the analysis (Figure 3, Table 1). Among eligible patients (n=6343), 65% never entered treatment data (nonusers), 11% entered one to 15 treatment records (moderate users), and 25% entered >15 treatment records (frequent users) during a 30 day baseline period after registration. Patients’ mean age ranged from 54 to 58 years old between groups (Table 2). Frequent users tended to more often be of a white race, non-Hispanic ethnicity, educated, and had a shorter dialysis vintage.

Figure 3.: Patient f low d iagram. BMI, body mass index.

Table 1. - Comparison of baseline characteristics among participants included versus patients excluded due to only missing data for adjustment of event analysis

Cohort	Included	Excluded Due to Missing Variables for Adjustment	Included versus Excluded P Value
Patient count	6343	4736
Demographics
Age (yr)	56.9±15.2	57.2±14.7	0.31
Males (%)	57	57	0.93
Black race (%)	23	25	0.003
White race (%)	72	68	<0.001
Other race (%)	5	7	0.008
Hispanic ethnicity (%)	10	11	0.022
Dialysis vintage (d)	690	472	<0.001
Alcohol use (%)	57	57	0.93
Urbanicity (%)
Metropolitan	81	83	0.028
Micropolitan	11	10	0.15
Rural	8	7	0.13
Education (%)
College or beyond	58	58	0.97
High school or equivalent	33	32	0.65
Less than high school or equivalent	9	10	0.51
Comorbidities (%)
Congestive heart failure	11	11	0.67
Diabetes	54	54	0.98
Ischemic heart disease	13	11	0.028
Laboratory
Albumin (g/dl)	3.55±0.45	3.58±0.47	0.002
Residual kidney function (ml/min)	4.26±3.32	4.67±4.69	<0.001
Weekly Kt/V	2.43±1.05	2.39±1.04	0.11

Comparisons between groups were made using t tests. ± denotes SD for continuous variables.

Table 2. - Baseline characteristics of participants

Group	Nonusers	Moderate Users	Frequent Users
Patient count	4093	673	1577
Demographics
Age (yr)	57.6±14.7	53.9±15.6 a	57.9±14.6
Males (%)	58	54	58
Black race (%)	24	23	18 a
White race (%)	71	72	76 a
Other race (%)	5	5	6
Hispanic ethnicity (%)	11	9 b	6 a
Dialysis vintage (d)	732±803	553±655 a	570±732 a
Alcohol use (%)	24	26	25
Urbanicity (%)
Metropolitan	81	79	81
Micropolitan	11	13	12
Rural	8	8	8
Education (%)
College or beyond	56	64 a	60 c
High school or equivalent	33	29 c	33
Less than high school or equivalent	10	7 b	7 a
Comorbidities (%)
Congestive heart failure	11	10	11
Diabetes	55	53	54
Ischemic heart disease	12	12	14
Laboratory
Albumin (g/dl)	3.51±0.44	3.62±0.44 a	3.62±0.44 a
Residual kidney function (ml/min)	4.26±3.32	4.18±3.21	4.43±3.40
Weekly Kt/V	2.42±1.08	2.40±0.72	2.47±1.12

Comparisons between groups were made using t tests.

^aP<0.001.

^bP<0.01.

^cP<0.05.

The within-group trends of RTM usage were fairly sustained over the 12-month follow-up, as compared with baseline, yet frequent users had decreases in the mean number of entries (Figure 4). On the 12th month of follow-up, frequent users on average documented ten treatments in the RTM.

Figure 4.: Sustained trend in the mean number of RTM entries in each month of the follow-up period by baseline RTM use group category. RTM, remote treatment monitoring.

Longitudinal Hospitalization Rates Associated with RTM Use

Higher RTM usage in the 30 days after the start date was found to be associated with progressively lower unadjusted hospital admission and day rates in the follow-up periods (Figure 5 ). A Poisson analysis showed the incidence rate of hospital admission was 22%, 24%, 23%, and 26% lower in frequent users of the RTM application after 3, 6, 9, and 12 months of follow-up, respectively, as compared with nonusers (Table 3). The incidence rate of a greater hospital length of stay in days was found to be 38%, 35%, 34%, and 32% lower in frequent users of the RTM application after 3, 6, 9, and 12 months of follow-up, respectively, versus nonusers (Table 4). Albeit qualitative differences were observed favoring moderate use of the RTM, there were NS differences compared with nonusers.

Figure 5.: Improvement in hospitalization rates associated with RTM use. (A) Hospital admission rate by baseline RTM use after 3, 6, 9, and 12 months of follow-up. (B) Hospital day rate by baseline RTM use after 3, 6, 9, 12 months of follow-up. PPY, per patient year.

Table 3. - Associations in adjusted hospital admission rates by baseline remote treatment monitoring use

Parameter	IRR Estimate (95% CI)
	3-Mo Admission Count	6-Mo Admission Count	9-Mo Admission Count	12-Mo Admission Count
Moderate users	0.91 (0.74 to 1.12)	0.92 (0.78 to 1.10)	0.94 (0.80 to 1.11)	0.93 (0.80 to 1.09)
Frequent users	0.78 (0.66 to 0.91)	0.76 (0.67 to 0.87)	0.77 (0.68 to 0.87)	0.74 (0.66 to 0.83)
Age	1.00 (0.10 to 1.01)	1.00 (0.10 to 1.01)	1.00 (0.10 to 1.00)	1.00 (0.10 to 1.00)
Females	1.07 (0.93 to 1.23)	1.13 (1.01 to 1.27)	1.06 (0.95 to 1.18)	1.06 (0.96 to 1.17)
Race/ethnicity
Black race	0.99 (0.85 to 1.15)	0.97 (0.85 to 1.10)	0.89 (0.79 to 1.01)	0.88 (0.79 to 0.98)
Other race	0.77 (0.56 to 1.60)	0.71 (0.54 to 0.94)	0.67 (0.51 to 0.87)	0.60 (0.47 to 0.77)
Hispanic ethnicity	1.01 (0.81 to 1.25)	0.97 (0.81 to 1.16)	0.87 (0.74 to 1.04)	0.87 (0.74 to 1.03)
Comorbidity
Congestive heart failure	1.30 (1.09 to 1.55)	1.35 (1.17 to 1.56)	1.30 (1.13 to 1.50)	1.23 (1.08 to 1.40)
Diabetes	1.24 (1.09 to 1.42)	1.24 (1.11 to 1.38)	1.22 (1.10 to 1.35)	1.24 (1.13 to 1.36)
Ischemic heart disease	1.23 (1.04 to 1.46)	1.16 (1.00 to 1.34)	1.16 (1.01 to 1.33)	1.15 (1.02 to 1.31)
Laboratory measures
Albumin	0.51 (0.44 to 0.58)	0.51 (0.46 to 0.57)	0.52 (0.46 to 0.58)	0.52 (0.47 to 0.57)
Residual kidney function	0.98 (0.95 to 1.01)	0.98 (0.96 to 1.01)	0.98 (0.95 to 1.00)	0.98 (0.96 to 1.00)
Weekly Kt/V	0.91 (0.80 to 1.04)	0.88 (0.78 to 0.99)	0.90 (0.81 to 1.00)	0.91 (0.83 to 1.00)
Dialysis vintage	1.00 (1 to 1.00)	1.00 (1.00 to 1.00)	1.00 (1.00 to 1.00)	1.00 (1.00 to 1.00)
Education
College or beyond	0.90 (0.73 to 1.11)	1.23 (0.82 to 1.18)	1.01 (0.85 to 1.20)	1.14 (0.87 to 1.20)
High school or equivalent	0.99 (0.80 to 1.23)	0.96 (0.89 to 1.30)	1.09 (0.91 to 1.30)	0.10 (0.92 to 1.28)
Alcohol dependency
Declined to answer	0.70 (0.24 to 2.03)	0.98 (0.27 to 1.66)	0.94 (0.46 to 1.94)	1.02 (0.60 to 2.02)
No	0.98 (0.85 to 1.13)	1.08 (0.84 to 1.06)	0.97 (0.87 to 1.09)	1.08 (0.84 to 1.04)
Urbanicity
Metropolitan	1.20 (0.95 to 1.52)	0.67 (1.01 to 1.49)	1.16 (0.97 to 1.40)	1.10 (0.96 to 1.35)
Micropolitan	0.91 (0.67 to 1.23)	0.94 (0.75 to 1.24)	1.00 (0.80 to 1.27)	0.93 (0.81 to 1.23)

Reference groups: Nonusers, males, white race, no high-school education, yes to alcohol dependency, rural urbanicity. Hospital admission rates are adjusted by all parameters listed in the table. IRR, incidence rate ratio.

Table 4. - Associations in adjusted hospital day rates by baseline remote treatment monitoring use

Parameter	IRR Estimate (95% CI)
	3-Mo Day Count		6-Mo Day Count		9-Mo Day Count		12-Mo Day Count
Moderate users	0.89 (0.56 to 1.43)		0.95 (0.69 to 1.31)		0.89 (0.66 to 1.19)		0.88 (0.68 to 1.15)
Frequent users	0.62 (0.42 to 0.90)		0.65 (0.51 to 0.85)		0.66 (0.52 to 0.82)		0.68 (0.55 to 0.83)
Age	1.01 (0.99 to 1.02)		1.01 (0.10 to 1.01)		1.01 (0.10 to 1.01)		1.01 (1.00 to 1.01)
Females	1.06 (0.78 to 1.43)		1.13 (0.91 to 1.40)		0.99 (0.83 to 1.19)		0.98 (0.83 to 1.15)
Race/ethnicity
Black race	0.65 (0.44 to 0.94)		0.77 (0.60 to 0.10)		0.73 (0.59 to 0.92)		0.73 (0.60 to 0.90)
White race	0.54 (0.24 to 1.18)		0.57 (0.33 to 0.98)		0.49 (0.30 to 0.81)		0.45 (0.28 to 0.72)
Hispanic ethnicity	1.03 (0.65 to 1.65)		0.97 (0.70 to 1.36)		0.89 (0.66 to 1.20)		0.90 (0.68 to 1.18)
Comorbidity
Congestive heart failure	1.31 (0.89 to 1.94)		1.33 (1.02 to 1.74)		1.31 (1.03 to 1.67)		1.25 (1.00 to 1.56)
Diabetes	1.39 (1.03 to 1.87)		1.45 (1.18 to 1.79)		1.34 (1.12 to 1.61)		1.31 (1.11 to 1.54)
Ischemic heart disease	1.11 (0.75 to 1.63)		1.08 (0.82 to 1.41)		1.07 (0.84 to 1.35)		1.07 (0.86 to 1.33)
Laboratory measures
Albumin	0.52 (0.38 to 0.71)		0.52 (0.42 to 0.46)		0.48 (0.40 to 0.58)		0.48 (0.41 to 0.57)
Residual kidney function	0.94 (0.88 to 1.01)		0.97 (0.92 to 1.02)		0.94 (0.91 to 0.98)		0.95 (0.92 to 0.98)
Weekly Kt/V	0.96 (0.75 to 1.23)		0.86 (0.69 to 1.08)		1.01 (0.92 to 1.10)		0.99 (0.90 to 1.10)
Dialysis vintage	1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)
Education
College or beyond	1.29 (0.78 to 2.15)		1.29 (0.90 to 1.85)		1.31 (0.95 to 1.80)		1.30 (0.98 to 1.74)
High school or equivalent	1.15 (0.67 to 1.95)		1.22 (0.84 to 1.76)		1.28 (0.92 to 1.78)		1.23 (0.91 to 1.66)
Alcohol dependency
Declined to answer	0.35 (0.01 to 10.94)		0.39 (0.04 to 3.77)		0.54 (0.10 to 2.92)		0.93 (0.29 to 3.04)
No	0.89 (0.65 to 1.23)		0.90 (0.72 to 1.12)		0.99 (0.81 to 1.20)		0.98 (0.82 to 1.18)
Urbanicity
Metropolitan	1.26 (0.73 to 2.17)		1.29 (0.88 to 1.88)		1.27 (0.90 to 1.77)		1.16 (0.86 to 1.56)
Micropolitan	1.11 (0.57 to 2.18)		1.15 (0.72 to 1.84)		1.19 (0.79 to 1.78)		1.16 (0.81 to 1.67)

Reference groups: nonusers, males, other race, no high-school education, yes to alcohol dependency, rural urbanicity. Hospital days rates are adjusted by all parameters listed in the table. IRR, incidence rate ratio.

Longitudinal Technique Failure Rates Associated with RTM Use

We observed that higher RTM usage in the 30 days after the start date was associated with lower rates of sustained PD technique failure (i.e., required >6 consecutive weeks of treatment with HD) in all follow-up periods (Figure 6). A Cox analysis showed the adjusted risk of sustained PD technique failure at 3, 6, 9, and 12 months of follow-up was 33%, 31%, 31%, and 27% lower, respectively, in frequent users of the RTM versus nonusers (Table 5). The Kaplan–Meier estimate for PD duration days without sustained PD technique failure identified that frequent users remained on PD longer compared with nonusers (log-rank test frequent users P=0.024; Figure 7A). No significant differences were found in sustained technique failure rates in moderate users compared with nonusers.

Figure 6.: Improvement in PD technique failure rates associated with baseline RTM use after 3, 6, 9, and 12 months of follow-up.

Table 5. - Associations in adjusted sustained peritoneal dialysis technique failure rates by baseline remote treatment monitoring use

Parameter	Hazard Ratio (95% CI)
	3-mo PD Technique Failure		6-mo PD Technique Failure		9-mo PD Technique Failure		12-mo PD Technique Failure
Moderate users	0.66 (0.40 to 1.07)		0.76 (0.54 to 1.07)		0.85 (0.64 to 1.13)		0.87 (0.67 to 1.12)
Frequent users	0.67 (0.48 to 0.94)		0.69 (0.54 to 0.88)		0.69 (0.56 to 0.85)		0.73 (0.60 to 0.88)
Age	1.00 (0.99 to 1.01)		1.00 (0.99 to 1.01)		1.00 (0.99 to 1.00)		1.00 (0.99 to 1.00)
Females	0.90 (0.67 to 1.20)		0.91 (0.74 to 1.13)		0.83 (0.68 to 0.10)		0.82 (0.69 to 0.98)
Race/ethnicity
Black race	0.78 (0.56 to 1.09)		0.92 (0.72 to 1.16)		0.91 (0.74 to 1.12)		0.91 (0.75 to 1.10)
Other race	0.39 (0.16 to 0.95)		0.52 (0.29 to 0.92)		0.61 (0.39 to 0.97)		0.56 (0.36 to 0.87)
Hispanic ethnicity	0.78 (0.49 to 1.25)		0.88 (0.63 to 1.24)		0.97 (0.74 to 1.29)		0.97 (0.75 to 1.25)
Comorbidity
Congestive heart failure	1.27 (0.87 to 1.86)		1.24 (0.94 to 1.65)		1.19 (0.92 to 1.53)		1.13 (0.89 to 1.43)
Diabetes	1.10 (0.84 to 1.45)		1.11 (0.91 to 1.35)		1.06 (0.90 to 1.26)		1.11 (0.95 to 1.30)
Ischemic heart disease	0.92 (0.62 to 1.36)		1.01 (0.76 to 1.34)		1.06 (0.83 to 1.36)		0.99 (0.79 to 1.24)
Laboratory measures
Albumin	0.64 (0.48 to 0.85)		0.64 (0.52 to 0.79)		0.69 (0.57 to 0.84)		0.68 (0.57 to 0.81)
Residual kidney function	1.03 (0.99 to 1.08)		1.02 (0.97 to 1.06)		1.01 (0.97 to 1.05)		1.01 (0.98 to 1.05)
Weekly Kt/V	0.58 (0.44 to 0.77)		0.58 (0.46 to 0.73)		0.67 (0.55 to 0.82)		0.67 (0.56 to 0.81)
Dialysis vintage	1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)
Urbanicity
Metropolitan	0.80 (0.53 to 1.21)		0.93 (0.67 to 1.30)		0.90 (0.68 to 1.19)		0.92 (0.70 to 1.19)
Micropolitan	0.63 (0.35 to 1.12)		0.74 (0.48 to 1.15)		0.75 (0.51 to 1.09)		0.81 (0.57 to 1.14)
Education
College or beyond	0.67 (0.45 to 1.00)		0.85 (0.61 to 1.18)		0.94 (0.71 to 1.26)		0.95 (0.73 to 1.23)
High school or equivalent	0.74 (0.49 to 1.13)		0.93 (0.66 to 1.30)		0.99 (0.74 to 1.33)		0.96 (0.73 to 1.27)
Alcohol dependency
Declined to answer	0.00 (0.00 to 1.18E+19)		0.00 (0.00 to 1.87E+14)		0.00 (0.00 to 1.02E+19)		0.00 (0.00 to 3.05E+17)
No	0.84 (0.63 to 1.11)		0.82 (0.63 to 1.02)		0.85 (0.71 to 1.02)		0.81 (0.69 to 0.96)

Reference groups: nonusers, males, white race, no high-school education, yes to alcohol dependency, rural urbanicity. Sustained PD technique failure rates are adjusted by all parameters listed in the table. PD, peritoneal dialysis.

Figure 7.: Longer time on PD modality associates with higher RTM use. (A) Kaplan Meier curve plot to assess the associations in the time on PD modality over the 12 months follow-up for non-users, moderate-users and frequent-users. Frequent-users of the RTM remained on a PD modality 13 days longer in comparison to the non-users group. (B) Kaplan-Meier plot for survivors of the 12-month follow-up period to assess the associations in technique failure rates and the time on PD modality in a group of patients who had equivalent opportunities to experience a technique failure. Frequent-users of the RTM in the survivor subgroup remained on a PD modality 11 days longer than non-users.

A Cox analysis of a subgroup of patients who survived the entire 12-month follow-up period and continued to be treated in the providers clinics confirmed higher RTM usage was associated with 26% and 21% lower adjusted risk of sustained PD technique failure in frequent users versus nonusers at the 9 and 12 month follow-up periods (Table 6). However, we did not find significant differences between moderate users and nonusers of the RTM. In this subgroup of survivors, we also observed the adjusted risk of sustained technique failure was 51% and 42% lower in moderate users of the RTM at 3 and 6 months of follow-up, respectively, compared with nonusers, yet no significant differences were found at later time points. The Kaplan–Meier estimate of PD duration days without sustained PD technique failure showed that frequent users of the RTM in the survivor subgroup remained on a PD modality longer than nonusers (log-rank test frequent users P<0.001; Figure 7B).

Table 6. - Associations in adjusted sustained peritoneal dialysis technique failure rates by baseline remote treatment monitoring use among survivors of the 12-mo follow-up

Parameter	Hazard Ratio (95% CI)
	3-Mo Technique Failure		6-Mo Technique Failure		9-Mo Technique Failure		12-Mo Technique Failure
Moderate users	0.49 (0.24 to 1.00)		0.58 (0.35 to 0.96)		0.81 (0.57 to 1.16)		0.83 (0.59 to 1.15)
Frequent users	0.75 (0.49 to 1.13)		0.74 (0.55 to 1.02)		0.74 (0.57 to 0.96)		0.79 (0.63 to 1.00)
Age	0.10 (0.98 to 1.01)		0.10 (0.99 to 1.01)		0.99 (0.99 to 1.00)		0.99 (0.98 to 0.10)
Females	0.89 (0.62 to 1.28)		0.84 (0.63 to 1.10)		0.74 (0.58 to 0.93)		0.76 (0.61 to 0.94)
Race/ethnicity
Black race	0.72 (0.47 to 1.11)		0.84 (0.62 to 1.14)		0.86 (0.66 to 1.11)		0.84 (0.66 to 1.06)
Other race	0.35 (0.11 to 1.12)		0.46 (0.21 to 0.97)		0.51 (0.28 to 0.94)		0.46 (0.26 to 0.82)
Hispanic ethnicity	0.89 (0.51 to 1.54)		0.81 (0.52 to 1.25)		0.96 (0.68 to 1.35)		0.10 (0.73 to 1.36)
Comorbidity
Congestive heart failure	1.12 (0.65 to 1.90)		1.04 (0.70 to 1.56)		1.09 (0.78 to 1.52)		0.99 (0.72 to 1.36)
Diabetes	1.00 (0.71 to 1.42)		1.16 (0.90 to 1.50)		1.08 (0.87 to 1.33)		1.11 (0.91 to 1.35)
Ischemic heart disease	0.93 (0.55 to 1.55)		0.98 (0.67 to 1.42)		1.08 (0.79 to 1.46)		1.08 (0.81 to 1.43)
Laboratory measures
Albumin	0.56 (0.38 to 0.81)		0.59 (0.44 to 0.78)		0.65 (0.51 to 0.82)		0.64 (0.52 to 0.80)
Residual kidney function	1.03 (0.99 to 1.08)		1.02 (0.97 to 1.07)		1.02 (0.97 to 1.06)		1.02 (0.98 to 1.06)
Weekly Kt/V	0.57 (0.41 to 0.82)		0.57 (0.43 to 0.77)		0.70 (0.55 to 0.89)		0.68 (0.54 to 0.84)
Dialysis vintage	1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)		1.00 (1.00 to 1.00)
Urbanicity
Metropolitan	0.70 (0.42 to 1.18)		0.84 (0.56 to 1.28)		0.79 (0.56 to 1.12)		0.80 (0.58 to 1.10)
Micropolitan	0.46 (0.22 to 0.99)		0.70 (0.40 to 1.21)		0.71 (0.45 to 1.12)		0.77 (0.51 to 1.16)
College or beyond	0.69 (0.41 to 1.17)		0.93 (0.60 to 1.42)		1.01 (0.70 to 1.46)		1.05 (0.75 to 1.47)
High school or equivalent	0.78 (0.45 to 1.34)		1.07 (0.69 to 1.66)		1.07 (0.74 to 1.56)		1.06 (0.75 to 1.50)
Alcohol dependency
Declined to answer	0.00 (0.00 to 1E+10)		0.00 (0.00 to 1E+10)		0.00 (0.00 to 1E+10)		0.00 (0.00 to 1E+10)
No	0.84 (0.58 to 1.22)		0.86 (0.66 to 1.14)		0.85 (0.67 to 1.06)		0.78 (0.64 to 0.96)

Reference groups: nonusers, males, white race, no high-school education, yes to alcohol dependency, rural urbanicity. Sustained PD technique failure rates are adjusted by all parameters listed in the table. PD, peritoneal dialysis.

Discussion

In a large population of patients on PD who registered online for the PatientHub RTM application, we found higher RTM use was associated with lower hospitalization and sustained technique failure rates. Hospital admission and day rates were observed to be 22% and 38% lower within 3 months, respectively, and continued to decrease in frequent users of the RTM over the 12 months of follow-up compared with nonusers. The risk of sustained PD technique failure (i.e., required >6 consecutive weeks of HD) was about 30% lower during follow-up for frequent users versus nonusers of the RTM. Assessment of a subgroup of survivors of the 12-month follow-up period found consistent trends, but significant differences were only observed at the 9- and 12-month follow-up periods in frequent users versus nonusers. Consistent trends were seen with respect to moderate users of the RTM versus nonusers, albeit outcomes did not significantly differ with the exception of sustained PD technique failure rates in the subgroup of survivors at the 3- and 6-month follow-up periods. These results further substantiate prior findings suggesting that use of other RTM systems in patients on PD may reduce hospital rates (¹⁴,¹⁵), and reveal frequent RTM use may also have the potential to increase sustained use of PD as a modality.

The associations between RTM use in patients on PD and hard outcomes have been reported in a limited number of small cross-sectional analyses. A study of 63 patients who had an RTM system incorporated in their PD cycler found that the incidence of hospital admissions was 39% (incidence rate ratio, 0.61; 95% CI, 0.39 to 0.95) lower and hospital days was 54% (incidence rate ratio, 0.46; 95% CI, 0.23 to 0.92) lower compared with 63 matched patients who did not have an RTM system in their cycler (¹⁴). Another study of 269 patients on PD who received an intervention of daily RTM of BP and weight coupled with video conferencing telehealth care showed that the adjusted risk of hospital admission was 46% (odds ratio, 0.54; 95% CI, 0.33 to 0.89) lower and hospital days was 54% (odds ratio, 0.46; 95% CI, 0.26 to 0.81) lower compared with before the intervention (¹⁵). It has been estimated that adoption of RTM systems to monitor PD treatments may also yield some economic benefits to healthcare systems in various countries (¹³,¹⁵).

The influence of RTM on technique failure rates has not been reported previously in PD, yet the improvements we found are consistent with observations from RTM in the home HD population (¹⁶). Notably, technique failure rates in all groups of our analysis were lower than many reports in the literature (¹⁷,¹⁸), which could be in part due to our definition of a sustained technique failure event that required >6 consecutive weeks of HD and appropriately did not count technique failures as composite outcomes including death. These findings could also be in part representative of patients included in our analysis being a highly select, healthier population compared to the overall PD population. Despite the differences in our technique failure rates with some reports, they are relatively consistent with technique failure rates reported in Japan and Asia that exclude mortality events from the definition (¹⁸).

RTM has been suggested to have the potential to improve patient-care team communications, timely interventions, and patient outcomes in PD for more than a decade (¹⁹). Despite this, the dialysis population has been known to have barriers to access the internet along with inadequate understanding of online systems (^2021–22). It is estimated that about 35%–90% of patients on dialysis use the internet (^2021–22). In our analysis we included 17% of patients from the overall active PD population based on inclusionary restrictions to have created an online account, age ≥18 years, and body mass index <65 kg/m². Prior reports suggest similar adoption with 18% of patients on PD in Columbia using a different remote monitoring system incorporated via connected health sensors in the cycler (¹⁴). Given patients on PD are typically younger, it would be expected that there might be a larger proportion of patients on PD with access to the internet. If this is a correct assumption, RTM might have the potential to be used in a larger proportion of the PD population and, as smartphone and computer technology advances and becomes more universally affordable, it could become an option for treatment monitoring in the majority of patients.

Our analysis design that included a group of patients on PD who universally had internet access to create an RTM account, along with temporal assessments of outcomes adjusted for confounding variables related to demographics, urbanicity, education, comorbidities, and laboratories, adds to the strength of these findings. The relative age of patients did not differ in patients on PD who were nonusers versus frequent users, which was expected given the inclusion of only patients with internet access and the ability to use online applications. The majority of the analysis population (65%) never used the RTM to enter clinical information within 30 days of creating an account online. RTM-use groups tended to have a sustained higher or lower pattern of RTM use over time compared with baseline, albeit the frequent-user group had some temporal decreases in entries. Both nonusers and moderate users of the RTM may be specific patient types that could be encouraged and trained to become more active in their care monitoring, and if frequent use is adopted as an adjunct, it may have the potential to yield improvements in outcomes.

Given nonusers of RTM were more commonly of a black race, Hispanic, with less education, and a longer dialysis vintage compared to frequent users; it may be important to target interventions to increase RTM use in patients with these attributes who initially elect to capture their treatment data using an RTM. However, further studies would be needed to test if targeting interventions based on patient profiles would be effective.

The RTM evaluated in this study is an application-based patient and clinician portal that we qualitatively believe required relatively minimal resources to construct, deploy, and maintain, as compared with modem-based systems integrated into cyclers. The flexible ability of the RTM to allow patients to enter records when it works the best for them during/around their treatment, combined with the ability for providers to review patients’ entries around their daily workflows, may be advantageous attributes of this type of connected health technology. These characteristics of the application-based RTM platform appear to allow it to be rapidly beneficial for the smaller percentage of patients with internet access who adopt RTM and allows for scalability over time.

Although this analysis has many strengths, there are some limitations including the inclusion a subgroup of the PD population who had access to the internet and registered for the RTM online. Therefore, these findings are not anticipated to be generalizable to patients on PD without access to the internet. We excluded patients with missing data on covariates to provide groups with equivalent adjustments for comparisons, yet this made our population have a higher representation of patients with a white race and longer dialysis vintage, among other distinctions. Neither the RTM nor the electronic medical record captured data on interventions performed due to findings from RTM entries, so we are not able to assess if interventions are being performed in a more timely manner before monthly clinic visits. Also, we cannot rule out that the favorable associations of higher RTM use might be due to patients being more engaged and having a higher health literacy. However, given the inclusion requirements of the design and the adjustments for education, these are not clear confounders. It is possible that more use of the RTM could be a driver influencing engagement and literacy, thereby teaching the patient to become more of a partner in their care. However, this concept would require further investigations. Our findings suggest frequent use of an RTM application associates with less hospital admissions, shorter hospital length of stay, and lower rates of sustained technique failure requiring HD exposure for >6 weeks. It appears prudent for PD care teams and providers to consider adopting RTM applications to better engage patients in their care, recognize and manage potential complications in a timely manner, improve the sustainability on the modality, and improve patient outcomes.

Disclosures

S. Chaudhuri, H. Han, J. Larkin, F. Maddux, M. Mendoza, and L. Usvyat are employees of Fresenius Medical Care in the Global Medical Office. D. Chatoth, D. Maddux, C. Muchiutti, and J. Ryter are employees of Fresenius Medical Care North America. D. Maddux, F. Maddux, and L. Usvyat have share options/ownership in Fresenius Medical Care. F. Maddux has directorships in American National Bank & Trust and is chairman of Pacific Care Renal Foundation 501(c)(3) nonprofit.

Funding

Analysis was supported by Fresenius Medical Care.