HealtheDataLab: Improving population health with a secure, next-generation research tool

AI and Data, Case Study

January 25, 2019

Discussions around the power of data and analytics in the advancement of healthcare have seen no shortage of excitement and expectations from both providers and patients. With this, the invaluable role of data scientists has also been brought to the spotlight. However, their work includes a less appealing element that is both exhausting and time/resource-consuming: organising and preparing the data before it can be used for modelling.

Dr William Feaster, chief health information officer at Children’s Hospital of Orange County (CHOC) in California, United States, explains how the organisation is using Cerner’s latest intelligence tools not only to predict patients’ risk of readmission, but also to inform clinical decisions and improve outcomes. Additionally, the tools release data scientists from the heavy lifting work of migrating and constantly maintaining the data, facilitating the process so it can be more quickly analysed and turned into insights that can then be fed back into clinical workflows.

By combining and utilising different types of clinical, patient, social and open source data, the team at CHOC, including a PhD data scientist, can build intelligent models which anticipate patient outcomes based on a multitude of medical and social factors.

“We believe that we have created an algorithm for readmissions that is more predictive than any other published model. It’s currently focused on paediatrics, but it can probably be extended to adults and 30-day readmissions as well”, Dr Feaster states.