Impact of automated peritoneal dialysis remote monitoring on hospitalization events

Chronic kidney disease (CKD) and GCC countries epidemiology

In this study, the PD population was more female, and the mean age of all participants was 51.2 years; Kuwaiti mean age was 58.4 years, while Omani patients’ mean age was 47 years. Diabetes mellitus (DM), followed by hypertension (HTN), was the major cause of CKD as shown in Table 2.

GCC faces complex challenges in managing and mitigating the record-high prevalence of lifestyle-related noncommunicable diseases.^[4,5] In 2018, the prevalence of end-stage kidney disease (ESKD) in GCC was 551 per million population (pmp), with Oman having the highest (1000/pmp) and Qatar having the lowest prevalence (347), and the median age of ESKD patients is 53.5 years for men and 51.0 years for women.^[6] By contrast, in the US, CKD is more common in women (15%) than men (12%).^[7] The age of ESKD patients in the GCC countries ranged from 33 to 61.9 years.^[8] To be specific, the median age of ESKD patients in Kuwait was 45 years^[9,10] and in Oman was 53 years.^[11] CKD in the US is more common in people aged 65 years or older (38%) than in people aged 45–64 years.^[7]

DM and HTN are the major causes of CKD in adults.^[7] Likewise, in all GCC countries, DM and HTN emerged as the most common etiologies, followed by chronic glomerulonephritis (66.7%) and unknown etiology (33.3%).^[6-9,11] According to a meta-analysis, the most common comorbidities associated with ESKD in GCC countries are DM (47.85%), HTN (77.88%), cardiovascular disease (14.51%), and vascular disease (11.7%).^[8]

PD complications pre- and post-implementation of Sharesource

Table 3 shows rates of PD complications and hospitalization before ARM compared to their rates after ARM initiation. Twenty-two percentages of group A patients had infectious complications, and peritonitis represented 50% of the infectious complications. In group B, 10% of patients had infectious complications, and peritonitis remained the most common complication. Non-infectious complications occurred in 25% of group A patients and declined to 13% in group B. Figure 2 demonstrates the causes of admission in both groups. Hospitalization rates and length were higher pre-remote monitoring; 17 cases were admitted due to PD complications. They were admitted for a period between 5 and 7 days for management of non-infectious complications compared with 11 patients post ARM initiation admitted for 3–5 days [Figure 3]. Most of the admissions in both groups were due to peritonitis, with an average admission of 10 days during pre- and post-application of ARM.

Close

Figure 2:

Information about the reasons for admission in our participants and comparing the findings during the two periods, pre- and post-remote monitoring (group A and B).

Export to PPT

Close

Figure 3:

Comparing the average length of admission days during the two periods, pre and post remote monitoring (group A and B).

Export to PPT

List of alarms, skipping, and incomplete PD therapy recorded from the platform

Over 6 months of monitoring through the Sharesource platform, we observed several alarms, which usually vary by cause and type of treatment. Some alarms were simple and managed through phone calls easily. However, other alarms indicated serious PD complications that required management in the hospital. Figure 4 lists names and the average number of alarms per month. The number of alarms during one APD exchange depends on the patient’s response to correct the problem (act or ignore) and the severity of the problem (the patient could not manage the issue without visiting the hospital).

Close

Figure 4:

Listed number of alarms per one month observed through the Sharesource platform.

Export to PPT

Table 4 and Figure 5 illustrate the number of skipped APD therapy and uncompleted therapy from January to June. The number of dialysis exchanges is calculated by multiplying the number of patients (92) by the number of days per month. For example, April has 30 days, therefore, the total exchanges is 2760. The percentage of skipped therapy varies between 7.1% and 12.8%. This percentage is not all related to patient adherence but also includes dialysis off days given by the PD team and temporary withholding of PD due to surgical or other medical reasons. There was an average of 6–10 APD therapies/month disconnected before completing full therapy. Patients were asked verbally for reasons behind this action, and they reported one or more of three main reasons: Disturbance due to continuous machine alarm (LOW DRAIN VOLUME), intolerable abdominal pain during the draining phase, or personal reasons.

Close

Figure 5:

Illustration of the number of skipped APD therapy and incomplete treatment by our participants during a six-month period. APD: Automated peritoneal dialysis.

Export to PPT

Relationship between early detection of complications through remote monitoring and reduced admission rate

Table 5 details issues that happened during APD exchanges and were managed early. Some of these problems were managed only by advice over the phone, such as low drain volume alarms caused by the patient’s sleeping position and skipping or disconnecting of APD therapy before completion. On the other hand, other problems required examination and management at the PD clinic, such as low drain volume and low UF due to constipation, catheter blockage, and uncontrolled BP due to non-adherence to medication, salt and/or fluid intake, and APD therapy regimen, and due to inadequate PD prescription. Some issues require both a PD clinic assessment and admission. Early detection of such issues facilitated early management and reduced admission days. For example, low UF due to volume depletion required admission for 1–3 days for correction of volume status and adjustment of APD prescription. Figures 6 and 7, taken from the Sharesource system, compare UF status with volume depletion and after 1 month of treatment for an old patient suffering from poor fluid intake.

Close

Figure 6:

Taken from platform screen for one of elderly patient suffered from dehydration and poor oral intake. It showed UF status when patient was dehydrated. UF: Ultrafiltration.

Export to PPT

Close

Figure 7:

Taken from platform, showed the status of UF of the elderly patient post treatment. UF: Ultrafiltration.

Export to PPT

Similarly, patients presented with frequent alarms of low drain volume and poor UF due to catheter migration required hospitalization between 3 and 5 days for surgical correction of catheter position using a laparoscopic technique and evaluation of therapy after adjustment, as illustrated in Photo 1a and 1b.

Close

Photo 1:

Compares two abdomen X-ray images pre and post catheter reposition for patient presented with continues alarm of low drain and low UF. It showed catheter migration and patient required surgical intervention for correction of PD Catheter position. (a) The red arrow in photo 1a shows that the PD catheter has migrated to the left upper quadrant of the abdomen and requires reposition. (b) The red arrow in photo 1b shows the PD catheter shadow in the pelvic region and it indicates that the catheter is back to the correct position after surgical intervention. UF: Ultrafiltration, PD: Peritoneal dialysis.

Export to PPT

Unfortunately, most infectious complications could not be detected early through Sharesource, but the patient attended the emergency department (ED) and was then admitted. It could be explained by patient awareness about the signs and symptoms of peritonitis and having to attend the ED for treatment, as they were instructed during PD training, or due to intolerable abdominal pain.

PD before and after ARM

The present study reports several complications detected in the early stage through Sharesource that prevented hospitalization or shortened the length of stay. One of the issues dealt with at PD Clinics was alterations in the Ultrafiltration pattern. This could have been the result of patient-related factors such as patient adherence, sleeping position, volume depletion, and/or constipation, and PD catheter-related factors such as catheter blockage and/or catheter migration.^[12-16] Another issue that was managed in PD Clinics was uncontrolled BP related to patient non-adherence to antihypertensive drugs, diet, and fluid restriction, and APD therapy regimen. Other issues that did not require admission in most instances included constipation, inadequate PD prescription, and PD catheter malfunction. However, a few cases required admission and surgical intervention, like catheter migration. On the other hand, infectious complications could not be detected through Sharesource, and patients required admission. In fact, hospitalization due to infectious complications, in particular, peritonitis, remained the same as it was before ARM initiation. Home-based dialysis may lead to patients’ inability to receive adequate care if they are not regularly monitored, as in-center dialysis. Patients with PD are usually seen once every 1–3 months to evaluate PD performance. This evaluation includes physical assessment and achievement of PD, such as UF and dialysis clearance. Nevertheless, patients cannot be monitored during other days at home for adherence, fluid management, and early detection of complications, which usually increases the risk of late management and hospitalization.^[17-19]

Baxter Company in the United Kingdom introduced a new system called Claria Sharesource for people on APD in 2016. The program automates the transmission of dialysis data to PD units and enables clinicians to remotely modify dialysis therapy cycler programs.^[3] In 2018, Sharesource was implemented in most (GCC) countries including Oman and Kuwait.

APD patients on remote monitoring reported long-term advantages such as treatment adherence improvement, early problem detection and resolution, fewer hospital admissions for technical and clinical issues, which saves money, and improved patient quality of life.^[20-22] Published literature also showed that remote monitoring improves ultrafiltration profile, BP control, and early catheter dysfunction detection rate, and allows timely treatment modification.^[13,16-18] For example, RPM allows medical professionals to monitor and analyze short-term ultrafiltration failures while regularly reviewing the record base and looking for trends in long-term follow-up changes.^[23]

Several studies support our findings.^[17-29] Sanabria et al.^[17] conducted a retrospective cohort study that included 360 adult PD patients from Colombia, with APD-RPM accounting for 18%, and 82% were managed with APD without RPM. Authors found APD therapy with RPM associated with significant reductions in hospitalization rate and fewer hospitalization days per patient-year compared with APD without RPM.^[17] Likewise, in 2020, Manani et al.^[27] conducted a retrospective cohort study to compare the outcomes of two groups of APD patients who were followed up with either standard care or ARM for 6 months. After analyzing ED visits, hospitalizations, peritonitis, volume depletion, and dropout rates, it was determined that ARM enhanced clinical outcomes for PD patients by lowering admission rates related to volume overload and ESI. However, no difference was observed in hospitalization rates related to peritonitis and other causes, particularly for those with higher comorbidity scores.^[27] Bubieńczyk et al.^[18] presented cases that showed RPM helped spot changes in ultrafiltration in various clinical situations. Peritonitis and some cases caused a decline in the UF pattern a few days before reporting to the hospital and admission, when the Sharesource platform was checked.^[18] The study by Yeter et al.,^[25] analyzed the effect of ARM-APD on treatment adherence, dialysis adequacy, and change in BP control, sleep quality, and health-related quality of life during the 6 months of follow-up. It found a significant increase in ultrafiltration amounts with ARM-APD and the daily antihypertensive pill needed, and alarms received from the device decreased at the 6^th month of the switch. Moreover, there was no significant change in sleep quality and health-related quality of life within 6 months.^[24]

Study limitation

A six-month duration may not be long enough to wholly evaluate the effectiveness of remote monitoring and judge its role in detecting different types of PD complications and reducing hospitalization rates. We also did not have a concurrent group without remote monitoring for comparison. We instead compared the pre-application of Sharesource period to the post-application period for the same group of patients. It is recommended in the future to conduct studies that take into consideration the two limitations. However, our study recruited participants with different backgrounds from four big public hospitals in two different GCC countries, Oman and Kuwait. Data from the platform were collected by dedicated and qualified PD nurses, which can make a difference in the accuracy of study results. Finally, the number of participants is quite good to make a clear judgment on the effectiveness of this application in reducing hospitalization rates, even with 6 months of monitoring.

ARM has an interactive interface that allows the PD team to make prescription changes via a remote connection and reduces the need for frequent patient visits to the PD clinic. Daily monitoring makes it possible to detect problems early and correct potential situations of inadequate dialysis. Moreover, it detects patients’ adherence to prescribed therapy and offers safer and better-quality treatment. It is also associated with lower hospitalization rates and length of hospital stay.