Proving the Business Case for the Internet of Things

Hitachi uses AI to predict heart failure readmissions

Steve Rogerson
January 4, 2018
 
Japanese technology giant Hitachi is using artificial intelligence (AI) to help US-based Partners Connected Health predict with high accuracy the risk of hospital readmissions within 30 days for patients with heart failure.
 
The AI technology helps select appropriate patients to participate in a readmission prevention programme following hospital discharge, and can explain the reason why patients were identified as being at high risk. The 30-day readmission rate is regarded as one of the important indicators in hospital management, and can carry significant penalties for hospitals via the US Centers for Medicare & Medicaid (CMS) as part of the Affordable Care Act.
 
This technology is an example of explainable AI, a term defined as enabling machines to explain their decisions and actions to human users, and enabling them to understand, appropriately trust and effectively manage AI tools, while maintaining a high level of prediction accuracy.
 
"Traditional machine learning can help us predict events but, as end-users, we can't tell why the machine is predicting something a certain way," said Kamal Jethwani, senior director at Partners Connected Health Innovation. "With this innovation, doctors and nurses using the algorithm will be able to tell exactly why a certain patient is at high risk for hospital admission, and what they can do about it. We want to enable our providers to act on this information, which is a step beyond the state-of-the-art today, in terms of machine learning algorithms."
 
As part of the study, the Partners Connected Health Innovation team simulated the readmission prediction programme among heart failure patients participating in the Partners CCCP connected cardiac care programme, a remote monitoring and education programme designed to improve the management of heart failure patients at risk for hospitalisation.
 
These results were compared with data from approximately 12,000 heart failure patients hospitalised and discharged from the Partners HealthCare hospital network in 2014 and 2015. The analysis showed the prediction algorithm achieved a high accuracy of approximately AUC 0.71, and can significantly reduce the number of patient readmissions. AUC (area under the curve) is a measure of prediction model performance with an ideal value range from 0 to 1.
 
As a result, approximately an additional US$7000 savings per patient per year among the cohort of CCCP patients can be expected.
 
Hitachi's AI technology uses deep learning to construct this prediction model. With conventional deep learning models, it is difficult for users to understand why the AI predicted a particular outcome. This presents a challenge for its adoption in healthcare.
 
To address this problem, Hitachi developed a technology for risk prediction, analysing the results presented by deep learning and extracting the several dozens of actionable factors for each patient from the vast amount of data collected from heart failure patients. These are elements familiar to clinicians and can support medical decision-making in clinical practice.
 
Through a standard statistical approach based on this risk prediction model, the extracted factors were used to calculate the risk of hospital readmission, and the relevance of the factors was calculated. Thus, this explainable AI technology can enhance prediction accuracy and the quality of medical decision-making.
 
Hitachi and the Partners Connected Health Innovation team will jointly conduct a prospective study, which evaluates the prediction programme by clinicians and studies how to integrate this within clinical workflows. By using this AI technology, Hitachi will provide tactics for the medical field, including for insurance and pharmaceutical companies, emergency services, and other healthcare services where prediction-based on medical data can be used.
 
Partners HealthCare, an integrated health system founded by Brigham & Women's Hospital and Massachusetts General Hospital, includes two academic medical centres, community and specialty hospitals, a managed care organisation, community health centres, a physician network, home health and long-term care services, and other health care entities. Partners is a principal teaching affiliate of Harvard Medical School.