Predictive Analytics

Contact:

Artificial intelligence allows high-risk patients to be identified before problems become aggravated. The cost-effective care of high-risk groups is possible if preventive services are targeted correctly.

NHG’s predictive analytics’ risk tools utilize the methods of machine learning to ensure the early identification of high-risk individuals. The tools search among variables to best predict adverse patient events. When high-risk patients are identified early on, preventive services can be targeted sensibly. Predicting service requirements is beneficial for allocating resources and the management of patient flows.

Our risk tools are integrated into our client’s information systems. They are highly automated and easy to use. The tools provide real-time support at the patient level for the day-to-day decision-making of experts. The tools can also be used for the continuous screening of the treated population to identify high-risk individuals.

The tools can also be used for the continuous screening of the treated population to identify high-risk individuals.

Contact:

Yrjänä Hynninen D.Sc. (Tech.)

Development Manager
yrjana.hynninen@nhg.fi

The research project consisted of developing an analytics algorithm which predicts the transfer of home care customers to round-the-clock service housing 6–12 months in advance.

The algorithm utilizes the indicators of the RAI system, which describe the functional capacity and health of a home care customer. Machine learning methods were used to search hundreds of individual RAI variables to identify the ones which best predict the transfer to round-the-clock service housing.

The algorithm proved to be accurate and was integrated into the RAI system. The data provided by the algorithm helps home care providers to plan and target services to those in most need. Targeted services make it possible to help the elderly to live longer safely at home and while staying within the sphere of home care services.

Link to RAIsoft article