CUSTOMER STORY | HEALTHCARE
Grupo SURA predicts neonatal complications 6 months ahead: 1,800 baby lives saved every year
In health insurance, a small group of chronic patients drives most of the costs. In under 3 months, Grupo SURA built predictive models on 8 years of clinical history to anticipate severe cases.
Baby lives saved every year
1,800
Newborns in intensive care
-20%
Accuracy, dialysis-stage diabetes
>98%
OVERVIEW
The challenge.
Being able to predict severe cases means better care and better cost control.
A few patients, most of the costs
In health insurance, a small group of chronic patients drives most of the costs.
Severe cases detected late
Severe diabetic, cardiac or birth-related complications need to be anticipated before they worsen.
Data in silos
Patient data was scattered between the health ministry, hospitals and insurers, with no overall view.
SOLUTION
What was done.
In under 3 months, a team of 4 built predictive models on 8 years of clinical history.
Model
4 predictive models, built on 8 years of clinical history (23,000 variables).
Data
A dataset of 4.5 million rows by 24,000 variables.
Deploy
The results were embedded in an analytics platform powering new targeted care programs.
RESULTS
What changed.
The models predict diabetes, cardiac and birth complications ahead of time.
Diabetes
More than 98% accuracy predicting dialysis-stage diabetes.
Cardiac complications
More than 97% accuracy predicting cardiac complications within 6 months.
Birth complications
Predicted 6 months in advance: 20% fewer newborns in intensive care, and 1,800 baby lives saved every year.
SEE FOR YOURSELF
Test it on your own data.
Run TIMi on your own data with a free trial, or talk to us about your case.
