Prediction

PREDICTION

Despite we live longer than ever, ample people develop chronic diseases early and have lower than expected life expectancy compared to the population as a whole. In an era of population ageing and an emphasis on equal opportunities for citizens to live a healthy long live, there is a need for personalized health services especially at old age where differences between people is far more prevalent than at young age.

Currently municipalities in Denmark invite every citizen aged 75-80 years old for an information meeting to prevent ill health and disability. In addition, from age 80 onwards, all are offered annual, preventive home visits. Between age 65-79, municipality offers preventive home visits only for those at high risk. However, there is no established strategy on how to accurately identify older persons at risk of declining health and in need of services. Here there is an opportunity to make use of register data and apply machine learning methods to establish an innovative, data-driven, population-based prevention strategy to identify older persons at high risk of disease and disability, and to circumvent the need of long-term care for those would benefit from tailor-made interventions.

The necessary machine learning methods have been developed and are being applied on a dataset consisting of all aged 65 years and over living in Denmark during the period of 1997-2017. To this end we combine predictors using register data available within Statistics Denmark, and outcomes defined by the municipality as the type and volume of personal care that is provided.

It is expected that this rigorous scientific analysis of health and registry data will improve our ability to identify citizens at risk of declining health, facilitating a shift from the traditional, demand-driven care towards a novel, data-driven prevention strategy. As the developed algorithms are flexible and scalable, this innovation will be applicable to other populations within Denmark and beyond.

Stay tuned and subscribe for updates.

For further information, contact us at info@dataforgood.science.