MexicoHealthcare and life sciences2023 – ongoingChristus Muguerza
Prediction model for acute pain after surgery
Christus Muguerza's pain model predicted 77% of acute pain episodes in surgical patients
Problem
Patient information at Christus Muguerza's hospitals was isolated and processed by hand, which made monitoring and analysis hard. Doctors could not anticipate pain in complex cases and treated it only after the patient reported it.
Approach
The health system built a predictive model on Google Cloud to anticipate the needs of postsurgical patients in real time. Dialogflow interprets the patient's natural-language message to find which part of the body hurts. Vertex AI handles the machine learning, and Cloud Functions calculate the risk of each incident, assign a team of doctors, notify the departments involved and pull data from the electronic record.
Results
Takeaway
Linking an electronic medical record to a prediction model let the hospital act on pain before the patient reported it. The company did not say how it measured the prediction rate.
Categories
- Industry
- Healthcare and life sciences
- Business function
- Production and quality
- Innovation type
- Process innovationProduct and service innovation
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Google Cloud's customer story describes how Christus Muguerza, a private health system in northern Mexico with 14 hospitals and over 7,000 workers, built a model to predict and prevent acute pain after surgery. It reports that 77% of acute pain episodes can now be predicted. The page carries no date; the Internet Archive first saved it on April 10, 2023.
Full details are in the original sources.
