Key TakeawaysCombatting Home Modality Loss: To help address the 30% of patients who leave peritoneal dialysis (PD)— a home treatment option — within two years of starting treatment, DaVita developed and deployed a predictive AI model that supports clinical teams in identifying patients needs to sustain home dialysis. Closing the Visibility Gap: DaVita’s connected cyclers and CWOW system provide care teams with consistent patient updates and data to support adjusted care plans and tailored interventions before complications arise. Seamless Transitions: When a transition off peritoneal dialysis is necessary, DaVita’s PD-to-HHD pathway enables and empowers patients to transition to home hemodialysis (HHD) in a mean of 6 days.
|
For approximately 80,000 Americans, home dialysis offers a sense of normalcy and stability amid many changes that inevitably come with a diagnosis of kidney failure. Home dialysis includes both peritoneal dialysis (PD) and home hemodialysis (HHD), differentiated options that allow patients to complete treatments with greater schedule flexibility. This not only provides a greater sense of control over treatment but also reduces the need for frequent trips to dialysis centers, giving patients time back to pursue employment, education, time with loved ones and personal hobbies.
Supporting patients with home treatments and sustaining success on therapy, like PD, requires a vigilant and proactive approach from care teams:
“Sustaining success on home dialysis requires true partnership with our patients,” Mihran Naljayan, MD, chief medical officer of clinical transformation, home and pediatric modalities for DaVita, says. “We must work hand-in-hand with patients to support their well-being and their success on their treatment of choice.”
Data from the National Institutes of Health show that home dialysis has steadily expanded during the past 10 years. In 2012, less than 9% of patients with newly diagnosed kidney failure chose home dialysis. By 2023, that statistic had increased to almost 15%.
“Impressive as that trend is, home dialysis would have expanded even more rapidly if patients were able to dialyze at home for a longer duration,” explains Eric Weinhandl, PhD, an epidemiologist at DaVita who has published dozens of studies about home dialysis. “Within two years of starting PD, more than 30% have transferred to in-center hemodialysis, whether because of medical complications or psychosocial obstacles, including a sense of burnout.”1
Complicating matters is that the transfer from PD to in-center hemodialysis is often accompanied by hospitalization.2 Weinhandl says, “Historically, we may have been too reactive, not proactive, when patients were struggling the most with home dialysis.”
Now, advances in artificial intelligence (AI) are providing clinicians with powerful new tools to support their patients. Kidney care leader DaVita has developed a predictive model that leverages machine learning to help clinical teams recognize and anticipate patient needs before they become critical problems.
The tool, known as the PD Loss predictive model, extracts hundreds of variables from a patient’s health record and creates a list of individuals each week who are at relatively high risk of discontinuing PD therapy during the next 90 days.3 This report helps care teams understand when people need extra support as well as why someone has been flagged as higher risk for transitioning off PD.
When a patient is identified, the system can suggest specific interventions, such as a medication review or a consultation with a specialist. The model doesn’t dictate care, though. Rather, it offers evidence-backed insights to inform care. Based on this feedback and interactions with the patients, nurses can tailor care activities with data-driven insights.
“When nurses initiate interventions within 5 days of a patient being identified as high-risk, we see that patients have approximately 15% lower risk of transferring to in-center hemodialysis during the next 2 to 3 months,” explains Weinhandl based on internal data from DaVita.
Keeping Care Close When Dialyzing from Home
What’s driving the availability of new technology? Data.
In the past, patients recorded treatments on paper and delivered this information to their care team when they saw them — often monthly. Now, at-home devices are helping close the gap between the home and the dialysis center.
At DaVita, the majority of patients treating on PD or HHD use connected cyclers — the machines that perform treatment — that record and integrate treatment data into their clinical profile within DaVita’s cloud-based platform, Center Without Walls™ (CWOW™).4
This gives care teams unprecedented visibility into what’s happening with a patient’s health while treating at home, tracking everything from treatment adherence to fluid removal trends.
“The physiologic and lifestyle benefits of home dialysis are readily apparent,” Dr. Naljayan says. “Historically, it has been difficult for the physician to manage the home dialysis treatment regimen in a timely manner, given once-a-month updates about blood pressure and fluid removal. Connected cyclers change that. Clinical teams now have better visibility into home treatments, equipping teams with insight on how to address issues before they escalate.”
Supporting Patients Through Their Kidney Health Journey
This data stream can also help clinical caregivers create pathways for safer, more thoughtful patient care journeys. This includes supporting DaVita teammates (employees) as they empower home dialysis patients in their “PD to HHD” pathway. This pathway helps identify PD patients who have one or more well-established risk factors for transfer to hemodialysis and provides education about HHD, a therapy that can allow the patient to continue dialyzing at home if the efficacy of PD eventually diminishes.
In a Kidney360 study published in 2022, Weinhandl reported that only 3% of patients who transfer from PD to hemodialysis select HHD. By flagging these patients early, care teams can provide enhanced education and proactive planning, whether it includes education, training, or helping the patient get a fistula (a dialysis access) placed if needed.
A pilot study of DaVita’s “PD to HHD” education process, presented at the 2025 National Kidney Foundation Spring Clinical Meeting, demonstrated a significant increase of people transitioning to HHD when PD was no longer an option. The study also highlighted that early education helped patients start HHD quickly, with a mean transition of just 6 days and the majority of patients making the transition without needing in-center hemodialysis.5
“We remain focused on supporting patients on their treatment of choice,” Dr. Naljayan adds. “Harnessing the power of predictive analytics and real-time data, we are building a more connected care ecosystem. The simple reality is that home dialysis requires effort, but that effort can unlock huge upsides. Our mission at DaVita is to share the load of that effort exerted by patients and their families.”
By combining the vast analytical power of AI with the irreplaceable expertise of clinicians, healthcare providers can create a more responsive, personalized and sustainable system of care that helps patients manage their health from their preferred environment: at home.
Service provider and modality select are choices made exclusively between the patient and nephrologist.
[1] United States Renal Data System. 2025 USRDS Annual Data Report: Epidemiology of kidney disease in the United States. Figure 3.11a, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2025.
[2] Weinhandl, Eric D.1,2; Saffer, Tonya L.3; Aragon, Michael3. Hidden Costs Associated with Conversion from Peritoneal Dialysis to Hemodialysis. Kidney360 3(5):p 883-890, May 26, 2022. | DOI: 10.34067/KID.0007692021
[3] Transplant and renal recovery losses are not included in or predicted by the model, and these modality losses account for positive loss reasons to be encouraged. This flag is a tool to assist in identifying patients who may change modality within 90 days and RNs should use their clinical license to interpret if a patient is at risk of leaving their modality of choice. The RN should tie in the nephrologist and entire care team as needed and document interventions recorded. Service provider and modality selection are choices made exclusively between the patient and nephrologist.
[4] While connected cycler data can sometimes be impacted by malfunctions or patient user error, the data is one of several valuable sources of patient treatment completion that should be analyzed alongside patient submitted treatment records and clinic visit conversations.
[5] Increase in Conversions from Peritoneal Dialysis to Home Hemodialysis: A Pilot Study of a Structured Education Process; AJKD NKF 2025 Spring Clinical Meetings Abstracts
View original content to download multimedia:
View original content to download multimedia:
View original content to download multimedia:
