Case Studies

Reducing cost through greater efficiency in patient care

Estimated reading time: 2 minutes

Client: World-renowned healthcare provider

Challenge: Operating on thin margins, a major U.S. hospital needed to ensure efficiency in patient care

Solution: A highly accurate machine learning algorithm that predicts patient emergency room visits


Predicting unplanned medical visits

An internationally recognized medical network that serves over four million patients annually in the United States, this client participates in a variety of programs to better serve their patients and the nation’s healthcare system. The Medicare Shared Savings Program (MSSP), incentivizes healthcare providers to deliver quality, cost-effective care for patients that rely on government insurance. While these incentives encourage hospitals to participate, the program operates on thin margins which are only financially viable when they can deliver their services in an effective and efficient manner.

By examining their claims data, the client recognized that a large portion of Medicare expenses were incurred during unplanned emergency room visits. They speculated that if they reduced this number, the MSSP could become a central part of their care efforts. Due to the siloed nature of the data, simple statistics could not predict which patients were likely to have an unplanned medical visit.


Creating a deep learning model to analyze historical data

Infinia ML created a supervised deep learning model using the client’s electronic medical records and claims data to analyze historical unplanned hospital visits and detect patterns. We calculated probabilities for 31 different diagnostic categories, each with 70-99% accuracy based on patient condition.

We projected probabilities down to the patient level and identified the most likely causes of the future unplanned emergency room visits. Using this information, physicians can proactively intervene and prevent costly future medical charges.


Healthier patient outcomes and Medicare Shared Savings Program profit

Proactively intervening to reduce the number of unplanned hospitalizations led to healthier patient outcomes and helped our client record its first ever profit in the Medicare Shared Savings Program. Infinia ML’s solution allows the client to continue providing high quality care to patients, while ensuring that operation costs are met.

Applying these techniques elsewhere

  • Identify business operation trends
  • Predict successful ventures or the likelihood of a successful product launch

Related Case Studies