Improve Medical Outcomes of Kidney Patients
Implementation Time:
9 months
Solution Provider: AI Singapore
RenalTeam is a healthcare provider for dialysis treatment. Their journey of caring for patients with end-stage renal disease began in 2012. They believe that patients with renal failure can live fulfilling lives with the support of their families and its team of caregivers. RenalTeam is now established in Singapore, Malaysia and Indonesia. Their journey with renal patients continues as they move towards creating value and outcome-focused care by harnessing the power of technology, build partnerships with their patients, develop and train their staff.
- Kidney patients have a high risk of hospitalisation
- Hospitalisation increases medical cost and mortality of patients
- Current prediction of patients’ imminent risk of hospitalisation is experience-based
How can RenalTeam systematically and accurately predict which hemodialysis patient will be hospitalised within the next week?
An AI model was developed to assist nurses:
- Logistic regression is used to ensure the explainability of the algorithm
- Statistical techniques such as upsampling of hospitalization data is used to tackle the unequal distribution of data
- Established medical knowledge is embedded into the AI model via feature engineering e.g. converting raw numerical data into meaningful medical categorical information
- Medical knowledge shared by RenalTeam are turned into useful features in the AI model e.g. creating moving averages from medical readings to capture trends in patients’ medical parameter
Outcome
- AI model is 36% more precise compared to RenalTeam’s medical team
- RenalTeam is integrating AI model into their system to act as a decision support tool for their medical team
- Senior nurses can prioritise and take a second look at patients flagged by AI model
- Nurses can then decide if intervention is needed based on their medical assessmentÂ
- Data culture instilled as patients’ dialysis data are properly stored and nurse’s feedbacks are collected to improve the AI model’s performance
PDF Document
This is the last content of this tab. If you do not see any resources above, it means the solution provider have not provided any resources. Feel free to contact Solution Provider for more information.
Implementation Time
9 months
Use Case Brochure
