Objective:
This article gives a bird view about how AI-based Remote Patient Monitoring models can be helpful for the HealthCare vertical for predicting 30-day readmissions of patients.
The objective or outcome of this research is a Logistic Regression Model that uncovers patients with a 2 times higher likelihood of readmission. This model allows the clinicians to focus on the patients with the greatest risk of readmission and bring down the cost of care while improving health outcomes.
Solution Approach
This solution utilizes a variety of technologies to gather biometric data as well as information on medication adherence and on Self – reported symptoms & behaviors via a remote patient monitoring system.
Challenges:
The solution must be BI and Predictive Modelling based, which can analyze
- Biometric Data
- Symptom Data &
- Behaviour Data
To uncover patterns in patient’s data to help predict the likelihood of 30-day readmissions.
Tools & Methods:
Data set of several million readings will be loaded into an in-memory columnar database & intuitive BI Dashboards are developed for action-ready analysis.
A predictive model is built on “R”, which indicated that when patients answered a certain subset of questions in a particular way, they were 2 times more likely to readmitted vs Population Average.
The model allows the Hospital users to focus on the patients with the greatest risk of readmission and bring down the cost of care while improving health outcomes.
