A.I. Solution For Stage 4 Cancer Patients Intervention

TeraCrunch - A.I. Solution For Stage 4 Cancer Patients Intervention

A.I. Solution For Stage 4 Cancer Patients Intervention

Download Case Study

About the Client

A major hospital system in the midwest

Challenges

Studies have found that timely intervention can extend length of life and improve quality of life for stage 4 and stage 3 cancer patients. Oncologists at a major hospital faced challenges in predicting adverse events in stage 3 and 4 cancer patients. The inability to anticipate complications resulted in delayed interventions, adversely affecting patients' quality of life and longevity.

Predicting Patient Risk Scores

TeraCrunch developed a predictive analytics solution that leverages machine learning algorithms to analyze patient data. The system identifies patterns and predicts the risk of adverse events, enabling timely medical interventions.


TeraCrunch collects physiological data continuously for study participants. The data are captured by wearable devices, then obtained by TeraCrunch via API and stored in a document-based database. The data then undergo a series of transformations and aggregations and are then joined with deidentified event history data from the hospital's EMR system. At that, a predictive model is learned by TeraCrunch's analytics engine and serialized to perform real-time predictions as new physiological inputs are obtained from the wearable devices.


TeraCrunch hosts the predictive model and provides predicted risk scores across a set of adverse events, including visits to the emergency department, unscheduled visits to the clinic, and death, among others. The predictions are provided in tabular format and absorbed back into the EMR system to allow for alerts in the typical workflow of the clinic.


The predictive models are retrained at regular intervals as more participants enter the study and longer time series are captured for the physiological measures. See figure below for a high-level flow

Outcomes

 
  • Reduced Emergency Dept. Complications 30% reduction in emergency complications among participants due to early intervention strategies.
  • Higher Survival rates Improved patient survival rates by an average of 6 months.
  • Better quality of life Enhanced quality of life for 70% of the patients monitored by the system.
  • Decreased Readmission Decreased hospital readmission rates by 20%, reducing healthcare costs.

Conclusion

TeraCrunch's solution empowered oncologists with actionable insights, allowing for preemptive care that significantly improved patient outcomes and healthcare efficiency.



Previous
Previous

Enhancing Value Based Care: Early detection of at-risk patients using AI & Analytics

Next
Next

Propensity to Pay (PPP) Predictive Analytics Solution to Optimize Collections