United StatesHealthcare and life sciences2026 – ongoingDaVita
Predicting home-dialysis discontinuation to guide clinical interventions
DaVita associated prompt intervention after AI alerts with approximately 15% lower risk of transfer to in-center hemodialysis
Problem
Patients could leave peritoneal dialysis because of medical complications or difficulties managing treatment at home. DaVita wanted care teams to recognize problems earlier.
Approach
DaVita developed its PD Loss model using machine learning and patient health records. Weekly reports identified patients at higher risk of stopping peritoneal dialysis within 90 days. The system suggested interventions, while nurses chose care activities using clinical judgment and patient discussions.
Results
Takeaway
The reported benefit depended on nurses acting on risk alerts. DaVita's internal association did not establish that the model alone caused the reduction.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Built in-house
- Goal
- Quality and risk
- Evidence
- Company's own claim
Sources
Source summaries
DaVita described weekly machine-learning alerts and nurse-led interventions. It associated timely intervention with approximately 15% lower transfer risk in internal data.
Full details are in the original sources.