Reduce emergency department readmissions by 15% in 6 months*.

*results may vary

HBI Spotlight Predictive Analytics Platform more accurately stratifies patients to reduce hospital readmissions.

Using custom machine learned algorithms, HBI Solutions’ predictive analytics software compiles a full risk profile for every patient, starting on admission and updating in real-time throughout the stay, continuously monitoring changing risk for readmission for that visit, as well as adverse inpatient events and mortality.

This tool is truly unique and powerful. I use it to monitor our performance on several key indicators and inform organizational planning. I can compare our readmission rate to the state average to understand where to concentrate efforts, or see our population’s future risks to decide if we need to build up or reduce service lines.

William Wood, MD

Internist and VP, Medical Affairs, St. Joseph Healthcare

Call us at 650.285.2404 or complete the form below to learn more.

St. Joseph Healthcare reduces readmissions

St. Joseph Healthcare, a community health system located in Bangor, Maine, is a pioneer of preventative and wellness services. Care managers from inpatient and ambulatory centers work together to reduce readmissions and ED revisits by employing HBI Spotlight Analytics Platform, offered through their statewide health information exchange, since 2014.

St. Joseph’s ACO population is consistently below statewide risk-adjusted targets, and care managers also report higher job satisfaction levels.


better ED 30-day return rate than the risk-adjusted target


better IP 30-day readmission rate than the risk-adjusted target

What makes HBI Solutions’ predictive analytics strategy different?

  • HBI accepts any and all data, including social determinants
  • Custom machine learning is performed on your specific data
  • Natural Language Processing picks up key features and facts in unstructured formats
  • Risk features and influence on risk are transparent to the user through interactive dashboards
  • Risk models are being used on over 40 million lives
Landing Page Screenshot V2
Screenshot does not contain real patient names.